WEBVTT

00:00:00.000 --> 00:00:01.780
<v Chris Gammell>This is The Amp Hour Podcast.

00:00:03.220 --> 00:00:05.480
<v Chris Gammell>Release May 13th, 2025.

00:00:07.820 --> 00:00:09.360
<v Chris Gammell>Episode 693.

00:00:11.240 --> 00:00:14.280
<v Chris Gammell>Small-scale electronics manufacturing with Colin O'Flynn.

00:00:30.000 --> 00:00:31.700
<v Chris Gammell>Welcome to the Amp Hour.

00:00:31.900 --> 00:00:33.400
<v Chris Gammell>I'm Chris Gammell of Contextual Electronics.

00:00:33.720 --> 00:00:37.640
<v Chris Gammell>And I'm Colin O'Flynn of Dalhousie University and also New A technology.

00:00:38.080 --> 00:00:39.680
<v Chris Gammell>Welcome back, Colin, for the third time.

00:00:40.200 --> 00:00:40.480
<v Chris Gammell>It's been...

00:00:40.480 --> 00:00:40.900
<v Chris Gammell>Thank you.

00:00:40.900 --> 00:00:43.020
<v Chris Gammell>Yeah, we're on like a five-year cadence here.

00:00:43.360 --> 00:00:46.820
<v Chris Gammell>You and I chat a little bit more often than that, but it's good to have you back.

00:00:47.460 --> 00:00:49.340
<v Chris Gammell>Yeah, a recorded five-year conversation.

00:00:49.580 --> 00:00:49.880
<v Chris Gammell>That's key.

00:00:50.360 --> 00:00:50.640
<v Speaker ?>Yeah.

00:00:51.960 --> 00:00:57.620
<v Chris Gammell>Well, you know, if we just keep this up, then it's like a little time capsule for each year we operate, right?

00:00:58.200 --> 00:00:59.840
<v Colin O'Flynn>Yeah, time capsule for AI.

00:01:00.000 --> 00:01:04.280
<v Chris Gammell>Right, to be able to better recreate our history, our future.

00:01:04.800 --> 00:01:08.200
<v Chris Gammell>Chris and Colin sound progressively older every five years.

00:01:08.900 --> 00:01:12.860
<v Chris Gammell>Both in temperament and in vocal tambulations.

00:01:14.300 --> 00:01:16.320
<v Chris Gammell>Complaining about more basic things.

00:01:16.540 --> 00:01:17.080
<v Chris Gammell>Yeah, right.

00:01:18.100 --> 00:01:19.300
<v Chris Gammell>Don't you hate soup?

00:01:19.640 --> 00:01:20.420
<v Chris Gammell>Ah, soup.

00:01:21.100 --> 00:01:22.480
<v Chris Gammell>I guess we'd like more soup.

00:01:22.560 --> 00:01:24.000
<v Chris Gammell>Anyways, how you been?

00:01:24.160 --> 00:01:27.680
<v Chris Gammell>So you're kind of back in academia a little bit more, huh?

00:01:27.680 --> 00:01:29.720
<v Chris Gammell>I think the past two times...

00:01:29.720 --> 00:01:30.080
<v Chris Gammell>Well, let's see.

00:01:30.120 --> 00:01:32.660
<v Chris Gammell>We have you on episode 239.

00:01:33.100 --> 00:01:34.480
<v Chris Gammell>And we'll obviously link these in.

00:01:35.040 --> 00:01:36.040
<v Chris Gammell>And that was in 2015.

00:01:36.220 --> 00:01:37.880
<v Chris Gammell>And then 552, which is 2021.

00:01:38.940 --> 00:01:44.040
<v Chris Gammell>And I guess you were just starting at Dalhousie in 2021, right?

00:01:44.080 --> 00:01:45.100
<v Chris Gammell>That's like when you just started there?

00:01:45.100 --> 00:01:45.660
<v Colin O'Flynn>Yeah.

00:01:45.840 --> 00:01:47.440
<v Colin O'Flynn>So, and actually, it's kind of a...

00:01:47.440 --> 00:01:49.420
<v Colin O'Flynn>There's a story in there, too.

00:01:49.480 --> 00:01:53.540
<v Colin O'Flynn>Because in between, I was not at the university for a couple years.

00:01:54.100 --> 00:01:55.960
<v Colin O'Flynn>I had started back at the university.

00:01:55.960 --> 00:02:00.200
<v Colin O'Flynn>And then when Chip Whisperer and New A.E. was sort of taking off, I ended up leaving.

00:02:00.900 --> 00:02:04.480
<v Colin O'Flynn>Also over the pandemic, you know, things were a lot trickier to teach.

00:02:04.480 --> 00:02:06.740
<v Colin O'Flynn>So it felt like I needed to dedicate full time.

00:02:07.340 --> 00:02:10.640
<v Colin O'Flynn>Which at the time, you know, it's often you're deciding at the time of like, okay, that's

00:02:10.640 --> 00:02:11.360
<v Colin O'Flynn>the end of academia.

00:02:12.920 --> 00:02:14.160
<v Chris Gammell>But I came back.

00:02:14.160 --> 00:02:16.200
<v Chris Gammell>Is it like an in or out kind of thing usually, too?

00:02:16.280 --> 00:02:17.460
<v Chris Gammell>I mean, I guess I don't really know.

00:02:17.580 --> 00:02:18.280
<v Chris Gammell>Like, is it...

00:02:18.280 --> 00:02:19.880
<v Chris Gammell>Was it hard to switch in and out like that?

00:02:20.820 --> 00:02:21.660
<v Colin O'Flynn>It is.

00:02:21.960 --> 00:02:22.420
<v Colin O'Flynn>It depends.

00:02:22.540 --> 00:02:26.060
<v Colin O'Flynn>So I kind of got lucky with the timing that they were growing the computer engineering

00:02:26.060 --> 00:02:26.460
<v Colin O'Flynn>program.

00:02:26.620 --> 00:02:28.500
<v Colin O'Flynn>So this program didn't exist the first time.

00:02:28.600 --> 00:02:32.460
<v Colin O'Flynn>The first time I was actually connected more to like an entrepreneurship program.

00:02:32.460 --> 00:02:37.240
<v Colin O'Flynn>And they were growing computer engineering with the company or still able to publish.

00:02:37.320 --> 00:02:38.000
<v Colin O'Flynn>So I still publishing.

00:02:38.100 --> 00:02:39.240
<v Colin O'Flynn>That's normally the problem, right?

00:02:39.280 --> 00:02:45.460
<v Colin O'Flynn>As you go to industry, you don't have much time for publishing and you're kind of left

00:02:45.460 --> 00:02:45.920
<v Colin O'Flynn>behind.

00:02:47.500 --> 00:02:48.980
<v Colin O'Flynn>So it becomes harder to come back.

00:02:49.220 --> 00:02:49.500
<v Colin O'Flynn>So...

00:02:49.500 --> 00:02:51.600
<v Colin O'Flynn>But they promised me this time I can't leave and come back.

00:02:51.660 --> 00:02:55.460
<v Colin O'Flynn>They said there's a limit to how many times you do that.

00:02:56.160 --> 00:02:56.320
<v Chris Gammell>Right.

00:02:56.560 --> 00:02:56.800
<v Chris Gammell>Right.

00:02:57.320 --> 00:02:58.560
<v Chris Gammell>They told me that, too.

00:02:58.560 --> 00:03:03.660
<v Chris Gammell>But it was more in an academic, probationary regard.

00:03:04.420 --> 00:03:04.540
<v Chris Gammell>Yeah.

00:03:04.660 --> 00:03:06.940
<v Chris Gammell>Chris, if you don't get those grades up, you can't come back.

00:03:08.920 --> 00:03:09.760
<v Chris Gammell>Everybody gets one.

00:03:10.520 --> 00:03:10.880
<v Chris Gammell>Yeah.

00:03:11.100 --> 00:03:11.520
<v Chris Gammell>Well, that's cool.

00:03:11.640 --> 00:03:13.260
<v Chris Gammell>So what is it...

00:03:13.260 --> 00:03:14.580
<v Chris Gammell>What are you publishing about these days?

00:03:15.320 --> 00:03:17.640
<v Colin O'Flynn>So it's basically continuing on with...

00:03:17.640 --> 00:03:21.440
<v Colin O'Flynn>Right now, side channel analysis was some of my more recent work.

00:03:21.580 --> 00:03:23.120
<v Colin O'Flynn>I mean, which was previous work, too.

00:03:23.120 --> 00:03:26.500
<v Colin O'Flynn>But especially looking at side channel over different ports.

00:03:26.800 --> 00:03:29.440
<v Colin O'Flynn>So one more interesting thing was like using...

00:03:30.520 --> 00:03:31.900
<v Colin O'Flynn>Basically using jitter measurements.

00:03:32.240 --> 00:03:39.720
<v Colin O'Flynn>So as data flows through devices, the time delay varies with voltage, which means you can

00:03:39.720 --> 00:03:45.660
<v Colin O'Flynn>actually get a really accurate measurement of power on a device, which then gives you information

00:03:45.660 --> 00:03:46.980
<v Colin O'Flynn>about secrets.

00:03:46.980 --> 00:03:52.540
<v Colin O'Flynn>And so what I did is I ran that through the JTAG port, which is funny because this is the

00:03:52.540 --> 00:03:53.540
<v Colin O'Flynn>oldest attack.

00:03:53.780 --> 00:03:53.960
<v Colin O'Flynn>Like, right?

00:03:54.040 --> 00:03:56.140
<v Colin O'Flynn>Joe Grant's been doing JTAGulator for years.

00:03:56.300 --> 00:03:56.460
<v Chris Gammell>Yeah.

00:03:56.560 --> 00:03:56.720
<v Chris Gammell>Right.

00:03:57.500 --> 00:03:59.360
<v Colin O'Flynn>Everyone thinks JTAG is locked down.

00:03:59.440 --> 00:04:02.960
<v Colin O'Flynn>But yeah, if you run it in like a boundary scan mode, right?

00:04:02.980 --> 00:04:05.480
<v Colin O'Flynn>And you just do bypass, where it's just shifting bits through.

00:04:06.940 --> 00:04:11.740
<v Colin O'Flynn>You can basically shift through bits at pretty high, like sometimes a few hundred megahertz.

00:04:11.740 --> 00:04:13.640
<v Colin O'Flynn>And it basically gives you a pretty good scope.

00:04:14.120 --> 00:04:14.480
<v Colin O'Flynn>Yeah.

00:04:15.360 --> 00:04:16.060
<v Colin O'Flynn>Which is...

00:04:16.060 --> 00:04:17.860
<v Chris Gammell>And what is the differential?

00:04:18.220 --> 00:04:20.020
<v Chris Gammell>Like, what's the jitter amount?

00:04:20.260 --> 00:04:23.240
<v Chris Gammell>Is it like nanoseconds, picoseconds that you need to be able to detect?

00:04:24.100 --> 00:04:25.380
<v Colin O'Flynn>It's pretty small.

00:04:25.600 --> 00:04:30.500
<v Colin O'Flynn>Although the thing is that I did another version of the attack where what I do is it's called

00:04:30.500 --> 00:04:31.400
<v Colin O'Flynn>like fault sensitivity.

00:04:31.400 --> 00:04:36.780
<v Colin O'Flynn>But I basically clock in a bit right at the clock edge, right?

00:04:36.840 --> 00:04:40.220
<v Colin O'Flynn>And so depending on how that clock edge shifts, it's going to see a one or a zero.

00:04:40.220 --> 00:04:46.760
<v Colin O'Flynn>Um, and so you actually only get one bit of output data, you know, did a clock in a one

00:04:46.760 --> 00:04:47.200
<v Colin O'Flynn>or a zero.

00:04:47.540 --> 00:04:50.700
<v Colin O'Flynn>Um, so it's kind of like one bit GPS receivers type thing.

00:04:51.500 --> 00:04:55.320
<v Colin O'Flynn>The SNR is worse, but you get enough data and it still works.

00:04:55.560 --> 00:04:57.620
<v Chris Gammell>So averages over time sort of thing or...

00:04:57.620 --> 00:04:58.140
<v Colin O'Flynn>Yeah, exactly.

00:04:58.380 --> 00:05:01.540
<v Chris Gammell>So, um, there's, yeah, there's a lot of pretty interesting stuff you can do with that.

00:05:01.700 --> 00:05:04.280
<v Chris Gammell>Do you need then like a super specialized scope?

00:05:04.340 --> 00:05:06.700
<v Chris Gammell>Is it like a scope on top of something like that chip whisperer?

00:05:06.700 --> 00:05:10.740
<v Chris Gammell>Or like, what is the, what is the measurement methodology for doing that sort of thing then?

00:05:11.420 --> 00:05:11.560
<v Colin O'Flynn>Yeah.

00:05:11.620 --> 00:05:15.740
<v Colin O'Flynn>So for this one, it was actually just a old, you know, old school, but RF mixer.

00:05:16.080 --> 00:05:21.740
<v Colin O'Flynn>Um, so they, they also give you phase information as well as frequency if you mix the two signals

00:05:21.740 --> 00:05:22.120
<v Colin O'Flynn>together.

00:05:22.260 --> 00:05:27.940
<v Colin O'Flynn>So basically that's what was used is just a little, yeah, cheap mixer.

00:05:27.940 --> 00:05:30.460
<v Colin O'Flynn>So all the designs are available if someone wants to run it.

00:05:30.520 --> 00:05:34.160
<v Colin O'Flynn>And then I added it in, um, FPGA code to the chip whisperer.

00:05:34.160 --> 00:05:34.880
<v Chris Gammell>Yeah, that's great.

00:05:35.140 --> 00:05:37.320
<v Chris Gammell>And then can you maybe run people down again?

00:05:37.440 --> 00:05:41.460
<v Chris Gammell>You know, every five years is good about like kind of how a side channel generally works

00:05:41.460 --> 00:05:42.820
<v Chris Gammell>in a broad scenario.

00:05:43.560 --> 00:05:43.920
<v Colin O'Flynn>Yeah.

00:05:44.040 --> 00:05:45.340
<v Colin O'Flynn>I feel like I, sorry, I kind of dove into it.

00:05:45.360 --> 00:05:46.900
<v Colin O'Flynn>I don't know if this is the right intro, right?

00:05:46.980 --> 00:05:50.140
<v Colin O'Flynn>But, um, I should have started with what is side channel.

00:05:50.640 --> 00:05:55.380
<v Colin O'Flynn>I feel like more people know it now, but basically the idea is that you have information that

00:05:55.380 --> 00:05:58.600
<v Colin O'Flynn>influences some measurement you could do on the device.

00:05:58.600 --> 00:06:04.060
<v Colin O'Flynn>Um, which by that, I mean, specifically we're often looking at power analysis, which is,

00:06:04.140 --> 00:06:09.280
<v Colin O'Flynn>you know, as a device is running different instructions, take different amounts of power.

00:06:09.280 --> 00:06:14.040
<v Colin O'Flynn>So if your code does like a load or an ad or something, this will look different in a

00:06:14.040 --> 00:06:14.620
<v Colin O'Flynn>power trace.

00:06:15.440 --> 00:06:22.080
<v Colin O'Flynn>And it even gets more interesting, which is depending on the data transferred, it can vary

00:06:22.080 --> 00:06:22.760
<v Colin O'Flynn>a bit as well.

00:06:22.760 --> 00:06:30.200
<v Colin O'Flynn>So the, um, if you set a bit of a bunch of ones over a data bus, it takes more power out

00:06:30.200 --> 00:06:33.740
<v Colin O'Flynn>of a positive rail than if you send a bunch of zeros.

00:06:34.100 --> 00:06:37.800
<v Colin O'Flynn>Um, which seems kind of crazy that it's, you know, that you can actually measure that.

00:06:37.920 --> 00:06:41.360
<v Colin O'Flynn>I almost didn't believe it at first when someone said that, and then I was like, okay, let's

00:06:41.360 --> 00:06:41.840
<v Colin O'Flynn>test this.

00:06:42.620 --> 00:06:45.240
<v Colin O'Flynn>Um, but yeah, it actually works because it's charging right there.

00:06:45.240 --> 00:06:48.580
<v Colin O'Flynn>The, the, the bus internally is basically capacitance.

00:06:48.580 --> 00:06:54.380
<v Colin O'Flynn>Um, and so you're charging a capacitor to set it to a high or, you know, just charging,

00:06:54.460 --> 00:06:55.100
<v Colin O'Flynn>just set it to a low.

00:06:55.120 --> 00:06:56.380
<v Colin O'Flynn>And that comes from different power rails.

00:06:56.580 --> 00:06:59.180
<v Colin O'Flynn>So if you guys run one, you do see this.

00:06:59.500 --> 00:07:02.300
<v Chris Gammell>And I remember, so I got to go to a training that Colin ran as well.

00:07:02.300 --> 00:07:06.340
<v Chris Gammell>And I remember sitting there and kind of like watching what looked like a scope trace on

00:07:06.340 --> 00:07:07.540
<v Chris Gammell>like a 3.3 volt rail.

00:07:07.700 --> 00:07:10.580
<v Chris Gammell>And then you saw, it was kind of like floating on top of that rail.

00:07:10.580 --> 00:07:12.620
<v Chris Gammell>You saw like action happening, whatever.

00:07:12.620 --> 00:07:17.440
<v Chris Gammell>And then eventually though, you were using that to pull out encryption keys, which was kind

00:07:17.440 --> 00:07:18.600
<v Chris Gammell>of the wacky thing to me.

00:07:18.600 --> 00:07:22.940
<v Chris Gammell>Like you were doing kind of, uh, cross-referencing based on different actions that were happening.

00:07:23.040 --> 00:07:26.480
<v Chris Gammell>I don't remember the exact part of that, but that's what's in my mind at least.

00:07:27.040 --> 00:07:27.140
<v Colin O'Flynn>Yeah.

00:07:27.260 --> 00:07:30.260
<v Colin O'Flynn>No, I mean, that's pretty, that's basically, you know, a high level, what happens?

00:07:30.420 --> 00:07:32.520
<v Colin O'Flynn>The, the idea is right in that power trace.

00:07:32.700 --> 00:07:39.100
<v Colin O'Flynn>If I know at what point in time, for example, it's loaded an encryption key and it XORs that

00:07:39.100 --> 00:07:41.360
<v Colin O'Flynn>encryption key with some input data.

00:07:41.960 --> 00:07:45.920
<v Colin O'Flynn>Um, I know the input data because it's not secret.

00:07:45.920 --> 00:07:50.120
<v Colin O'Flynn>Like that would be the, you know, data coming in over the air, for example, I don't know

00:07:50.120 --> 00:07:56.740
<v Colin O'Flynn>the encryption key, but if I can use the power trace to say, Hey, there's, um, you know,

00:07:56.740 --> 00:08:03.340
<v Colin O'Flynn>it looks like there's more ones at this time after it did the XOR, um, or less ones.

00:08:03.340 --> 00:08:04.840
<v Colin O'Flynn>So we have this linear relationship.

00:08:05.080 --> 00:08:09.380
<v Colin O'Flynn>You could say, well, I know I put in, for example, zero, so there's zero one.

00:08:09.480 --> 00:08:11.000
<v Colin O'Flynn>So all those ones came from the key.

00:08:11.620 --> 00:08:12.600
<v Colin O'Flynn>Um, right.

00:08:12.600 --> 00:08:16.240
<v Colin O'Flynn>So by varying what the input data is, you can kind of, you can imagine that you could sort

00:08:16.240 --> 00:08:18.100
<v Colin O'Flynn>of slowly puzzle out in a way, right.

00:08:18.140 --> 00:08:23.340
<v Colin O'Flynn>What the, what the key is, because it's like, well, if I know I put in this and I see a lot

00:08:23.340 --> 00:08:27.440
<v Colin O'Flynn>of ones or low ones, um, and the only thing that's fixed is that key.

00:08:27.900 --> 00:08:30.080
<v Colin O'Flynn>Then I can kind of figure out what it is.

00:08:30.360 --> 00:08:30.920
<v Chris Gammell>I'm still not sure.

00:08:31.000 --> 00:08:31.960
<v Chris Gammell>I quite remember.

00:08:31.960 --> 00:08:38.380
<v Chris Gammell>I remember XOR, but, uh, so basically though, you're, it's, it's kind of, uh, probing with

00:08:38.380 --> 00:08:44.580
<v Chris Gammell>no, so like known inputs, unknown kind of reference data, and then known outputs cause you're measuring

00:08:44.580 --> 00:08:44.740
<v Chris Gammell>it.

00:08:44.820 --> 00:08:45.940
<v Chris Gammell>So that's kind of the idea is you're back.

00:08:46.020 --> 00:08:47.380
<v Chris Gammell>So you're back calculating that key.

00:08:47.500 --> 00:08:48.240
<v Chris Gammell>That's kind of the idea.

00:08:48.780 --> 00:08:49.280
<v Colin O'Flynn>Yeah, exactly.

00:08:49.500 --> 00:08:50.720
<v Colin O'Flynn>So, so that's the thing you basically have.

00:08:50.800 --> 00:08:54.820
<v Colin O'Flynn>And this works cause you have like, um, you know, the, the algorithms, they were not secret.

00:08:55.460 --> 00:08:58.860
<v Colin O'Flynn>Um, so you can build a pretty accurate system model.

00:08:58.860 --> 00:09:04.440
<v Colin O'Flynn>So the, the part that's kind of crazy is that if you do a test where all you do is you put

00:09:04.440 --> 00:09:09.320
<v Colin O'Flynn>different data on the data bus internally and measure the power at that point in time, it's

00:09:09.320 --> 00:09:11.720
<v Colin O'Flynn>a super, it's like a shockingly linear relationship.

00:09:11.720 --> 00:09:17.300
<v Colin O'Flynn>If I was, um, yeah, I'm always amazed when I do this demo and it's like, oh, that, you

00:09:17.300 --> 00:09:22.800
<v Colin O'Flynn>know, this sounds like a simple, uh, it's almost to protect against this.

00:09:22.880 --> 00:09:23.000
<v Chris Gammell>Huh?

00:09:23.200 --> 00:09:24.060
<v Chris Gammell>I mean, that's the other thing too.

00:09:24.120 --> 00:09:27.600
<v Chris Gammell>That's interesting is like, I remember seeing some marketing material that maybe even referenced

00:09:27.600 --> 00:09:29.780
<v Chris Gammell>the chip whisper, but it was like from chip companies.

00:09:29.820 --> 00:09:30.900
<v Chris Gammell>Now chip companies are like looking at it.

00:09:30.920 --> 00:09:32.180
<v Chris Gammell>They're like, oh, we should fix this.

00:09:32.180 --> 00:09:35.020
<v Chris Gammell>And we have countermeasures to try and make it better.

00:09:35.020 --> 00:09:38.640
<v Chris Gammell>But it, like you said, it's, it's physics at the end of the day, it's like capacitance

00:09:38.640 --> 00:09:39.700
<v Chris Gammell>and charge.

00:09:40.000 --> 00:09:40.320
<v Chris Gammell>Yeah, exactly.

00:09:40.520 --> 00:09:44.520
<v Colin O'Flynn>So, but I mean, and there is, there's a lot of ways to fix it so they can, you know, more

00:09:44.520 --> 00:09:45.840
<v Colin O'Flynn>complicated in the hardware.

00:09:46.120 --> 00:09:50.500
<v Colin O'Flynn>You can do dual rail and things where you basically have two different lines swinging high and

00:09:50.500 --> 00:09:50.760
<v Colin O'Flynn>low.

00:09:50.960 --> 00:09:56.080
<v Colin O'Flynn>So if you had a differential, right, the there's still a tiny bit of a glitch when it switches,

00:09:56.080 --> 00:09:59.540
<v Colin O'Flynn>but it's much, much, much more difficult to detect.

00:09:59.540 --> 00:10:04.500
<v Colin O'Flynn>Um, you can also change the algorithms a bit, like still using AES, but you can imagine,

00:10:04.640 --> 00:10:04.820
<v Colin O'Flynn>right?

00:10:04.860 --> 00:10:10.980
<v Colin O'Flynn>If I X or it in random data every time, um, and the algorithm is designed such that I

00:10:10.980 --> 00:10:14.260
<v Colin O'Flynn>can X or that data at the end to remove the effect of it.

00:10:14.280 --> 00:10:19.740
<v Colin O'Flynn>Like I've, I've complicated the attacker's life because they no longer know what's happening

00:10:19.740 --> 00:10:20.540
<v Colin O'Flynn>internally.

00:10:20.980 --> 00:10:21.620
<v Colin O'Flynn>Got it.

00:10:21.720 --> 00:10:21.920
<v Colin O'Flynn>Yeah.

00:10:22.220 --> 00:10:25.600
<v Colin O'Flynn>But I mean, the, the bigger thing, cause this, this often comes up and, you know, people

00:10:25.600 --> 00:10:27.160
<v Colin O'Flynn>say, okay, what should I do?

00:10:27.160 --> 00:10:30.580
<v Colin O'Flynn>Should I use super advanced crypto to avoid this?

00:10:30.640 --> 00:10:32.480
<v Colin O'Flynn>But it's like, what doesn't matter?

00:10:32.600 --> 00:10:36.800
<v Colin O'Flynn>The, the best choice, because, you know, as we might talk about later too, there's a million

00:10:36.800 --> 00:10:37.660
<v Colin O'Flynn>different attacks.

00:10:37.760 --> 00:10:42.920
<v Colin O'Flynn>And if you have one key for your 10 million devices out there, right?

00:10:42.960 --> 00:10:43.880
<v Colin O'Flynn>Like that's right.

00:10:43.960 --> 00:10:45.440
<v Chris Gammell>That's, that's the problem for sure.

00:10:45.520 --> 00:10:45.660
<v Colin O'Flynn>Yeah.

00:10:45.740 --> 00:10:45.860
<v Chris Gammell>Right.

00:10:45.960 --> 00:10:46.180
<v Chris Gammell>Exactly.

00:10:46.260 --> 00:10:47.160
<v Colin O'Flynn>Like that's the problem.

00:10:47.160 --> 00:10:52.340
<v Colin O'Flynn>If someone does a crazy power analysis attack and recovers the key from one device, like

00:10:52.340 --> 00:10:53.380
<v Colin O'Flynn>you shouldn't care.

00:10:53.500 --> 00:10:55.020
<v Chris Gammell>That controls your whole fleet of devices.

00:10:55.020 --> 00:10:55.420
<v Chris Gammell>Yeah.

00:10:55.420 --> 00:10:57.000
<v Chris Gammell>That's, yeah, that's right.

00:10:57.060 --> 00:11:02.480
<v Colin O'Flynn>Like you've made a bad decision early on and you're just trying to fix it through something

00:11:02.480 --> 00:11:02.760
<v Chris Gammell>else.

00:11:03.400 --> 00:11:04.160
<v Chris Gammell>Right, right, right, right.

00:11:04.500 --> 00:11:04.700
<v Chris Gammell>Okay.

00:11:04.720 --> 00:11:07.060
<v Chris Gammell>So, and you had mentioned the power rails, you mentioned JTAG.

00:11:07.120 --> 00:11:11.640
<v Chris Gammell>Are there other like interesting, like, I guess you, I, you kind of have power in a

00:11:11.640 --> 00:11:12.020
<v Chris Gammell>couple of places.

00:11:12.080 --> 00:11:14.660
<v Chris Gammell>You might have the battery, you might have the USB port, the five volts and USB.

00:11:15.000 --> 00:11:18.620
<v Chris Gammell>Are there other like ports and other ways to do this sort of thing?

00:11:18.620 --> 00:11:19.020
<v Chris Gammell>Yeah.

00:11:19.020 --> 00:11:19.420
<v Colin O'Flynn>Yeah.

00:11:19.520 --> 00:11:24.060
<v Colin O'Flynn>I mean, and that's kind of the, the interesting thing is that there's, so with this, this

00:11:24.060 --> 00:11:28.540
<v Colin O'Flynn>one doing this phase modulation stuff, it was kind of showing, you know, digital ports in

00:11:28.540 --> 00:11:28.800
<v Colin O'Flynn>general.

00:11:28.940 --> 00:11:33.940
<v Colin O'Flynn>So JTAG is one of the more accessible, but the other one, I, I didn't implement it, but

00:11:33.940 --> 00:11:39.020
<v Colin O'Flynn>sort of looked at was, you know, like an SD card or SDIO interface because that often

00:11:39.020 --> 00:11:40.220
<v Colin O'Flynn>is user.

00:11:40.820 --> 00:11:41.220
<v Colin O'Flynn>Right.

00:11:41.220 --> 00:11:43.700
<v Colin O'Flynn>Like some front panel interface has something in there.

00:11:43.700 --> 00:11:43.960
<v Chris Gammell>Right.

00:11:44.080 --> 00:11:44.260
<v Colin O'Flynn>Yeah.

00:11:44.320 --> 00:11:44.820
<v Colin O'Flynn>Yeah, exactly.

00:11:45.020 --> 00:11:49.640
<v Colin O'Flynn>And like, and they're pretty, it's high speed, which is good for being able to sample a lot

00:11:49.640 --> 00:11:50.080
<v Colin O'Flynn>of data.

00:11:50.740 --> 00:11:55.420
<v Colin O'Flynn>Um, so any of those interfaces usually encrypted through there or would that not even matter?

00:11:55.840 --> 00:12:00.620
<v Colin O'Flynn>Uh, it wouldn't because like really what you're doing is you're, you're basically using for

00:12:00.620 --> 00:12:01.360
<v Colin O'Flynn>the attack to work.

00:12:01.460 --> 00:12:06.400
<v Colin O'Flynn>What you want is a clock ideally that comes out from the target device.

00:12:06.400 --> 00:12:09.800
<v Colin O'Flynn>Cause that clock has small phase shifts encoded in it.

00:12:09.800 --> 00:12:11.480
<v Colin O'Flynn>Um, right.

00:12:11.560 --> 00:12:15.560
<v Colin O'Flynn>So if you get a clock and like the SDIO is good because it has a clock rate that runs

00:12:15.560 --> 00:12:16.300
<v Colin O'Flynn>quite a bit.

00:12:16.760 --> 00:12:23.620
<v Colin O'Flynn>Um, and you can kind of trick the system because you can send busy command busy states back.

00:12:23.740 --> 00:12:26.900
<v Colin O'Flynn>So if it does a read, you know, you send, you say, oh, actually I'm busy.

00:12:27.000 --> 00:12:28.280
<v Colin O'Flynn>Please continue to read for me.

00:12:28.760 --> 00:12:33.680
<v Colin O'Flynn>Um, and it'll just keep basically giving you a nice clock and every clock edge has some

00:12:33.680 --> 00:12:37.400
<v Colin O'Flynn>data encoded that, um, may do that.

00:12:37.400 --> 00:12:39.220
<v Colin O'Flynn>And then, you know, there's a similar thing.

00:12:39.720 --> 00:12:44.520
<v Colin O'Flynn>And one of the interesting things about, I don't know, one of the points of this paper

00:12:44.520 --> 00:12:50.060
<v Colin O'Flynn>actually that I had tried to make was not that, you know, this is a cool attack, but it was

00:12:50.060 --> 00:12:52.440
<v Colin O'Flynn>actually referenced in some of the earliest work.

00:12:52.440 --> 00:12:56.040
<v Colin O'Flynn>Um, you know, some people explored it and then sort of put it away.

00:12:56.440 --> 00:12:58.540
<v Colin O'Flynn>So like a lot of things in life, right.

00:12:59.300 --> 00:13:00.660
<v Colin O'Flynn>It's sort of been sitting there.

00:13:01.260 --> 00:13:06.380
<v Colin O'Flynn>Um, you know, from like 2003, I think the, the first paper I saw referencing this sort

00:13:06.380 --> 00:13:08.300
<v Colin O'Flynn>of angle modulation, they called it.

00:13:08.600 --> 00:13:08.700
<v Chris Gammell>Yeah.

00:13:08.800 --> 00:13:12.020
<v Chris Gammell>So does it mean that there's also like other clocked things that you could target like

00:13:12.020 --> 00:13:15.880
<v Chris Gammell>displays and even like RF front ends and stuff like that?

00:13:15.940 --> 00:13:16.060
<v Chris Gammell>Or.

00:13:16.060 --> 00:13:16.780
<v Colin O'Flynn>Well, yeah, exactly.

00:13:16.880 --> 00:13:21.800
<v Colin O'Flynn>And that's, that's kind of the, the idea is, and after I had worked on this paper, um,

00:13:21.800 --> 00:13:26.360
<v Colin O'Flynn>someone else totally independently published, you know, basically using the phase modulation,

00:13:26.360 --> 00:13:31.360
<v Colin O'Flynn>they were measuring over RF, um, the phase modulation, they were showing that has the

00:13:31.360 --> 00:13:31.680
<v Colin O'Flynn>advantage.

00:13:31.740 --> 00:13:37.060
<v Colin O'Flynn>Of course, you can, you know, you get longer distances, um, which was a bit of a spinoff

00:13:37.060 --> 00:13:40.320
<v Colin O'Flynn>of someone else had previously shown that you could do this over just RF.

00:13:40.320 --> 00:13:47.540
<v Colin O'Flynn>Um, what they actually did is they used a, I think it was a 802 15 four chip.

00:13:48.320 --> 00:13:49.560
<v Colin O'Flynn>Um, but the idea is right.

00:13:49.620 --> 00:13:54.620
<v Colin O'Flynn>As you, that chip is transmitting, you have it, it's running crypto at the same time.

00:13:54.620 --> 00:13:57.420
<v Colin O'Flynn>And like this ends up modulating onto the RF carrier.

00:13:57.900 --> 00:14:00.040
<v Colin O'Flynn>Oh, that was pretty wild.

00:14:00.160 --> 00:14:02.800
<v Colin O'Flynn>You know, it's, it's like kind of one, I think most of them are one die.

00:14:02.880 --> 00:14:03.060
<v Colin O'Flynn>Right.

00:14:03.140 --> 00:14:05.680
<v Colin O'Flynn>And so it's, you're going to vary.

00:14:05.800 --> 00:14:06.380
<v Colin O'Flynn>And even if not.

00:14:06.380 --> 00:14:07.200
<v Chris Gammell>It depends on the chip.

00:14:07.200 --> 00:14:13.080
<v Chris Gammell>Like, I think the TI one is, has a dual radio, like two, two, like, like chiplets or something

00:14:13.080 --> 00:14:13.480
<v Chris Gammell>like that.

00:14:13.480 --> 00:14:13.680
<v Chris Gammell>Okay.

00:14:13.740 --> 00:14:15.460
<v Chris Gammell>But I think the Nordic one is, is yeah.

00:14:15.460 --> 00:14:16.760
<v Chris Gammell>Like single die.

00:14:17.140 --> 00:14:18.940
<v Chris Gammell>Like the chiplets.

00:14:19.100 --> 00:14:22.880
<v Colin O'Flynn>I think there's enough coupling was the thing because like, you know, most of the radios

00:14:22.880 --> 00:14:26.860
<v Colin O'Flynn>are designed to say, you know, all they care about is do I meet, am I going to meet EM

00:14:26.860 --> 00:14:27.700
<v Colin O'Flynn>requirements?

00:14:28.100 --> 00:14:28.280
<v Chris Gammell>Right.

00:14:28.420 --> 00:14:28.640
<v Chris Gammell>Right.

00:14:28.740 --> 00:14:28.980
<v Chris Gammell>Right.

00:14:28.980 --> 00:14:29.180
<v Colin O'Flynn>Right.

00:14:29.240 --> 00:14:33.140
<v Colin O'Flynn>So like if they see a little bit of spread because of CPU noise, they're going to say,

00:14:33.180 --> 00:14:33.680
<v Colin O'Flynn>yeah, it's fine.

00:14:34.160 --> 00:14:34.360
<v Colin O'Flynn>Yeah.

00:14:34.360 --> 00:14:35.120
<v Colin O'Flynn>I don't care about that.

00:14:35.120 --> 00:14:40.300
<v Colin O'Flynn>But yeah, someone comes along to attack it and they say, oh, well that, you know, that

00:14:40.300 --> 00:14:45.100
<v Colin O'Flynn>phase spread is actually related to data being processed on the CPU at the same time.

00:14:45.260 --> 00:14:48.980
<v Colin O'Flynn>And yeah, which someone showed that quite a while ago.

00:14:49.000 --> 00:14:51.100
<v Colin O'Flynn>And there's been, as I said, some follow on on that.

00:14:51.160 --> 00:14:55.060
<v Colin O'Flynn>And then this kind of phase modulation over RF was proving that as well.

00:14:55.440 --> 00:14:55.540
<v Chris Gammell>Yeah.

00:14:55.540 --> 00:14:59.360
<v Chris Gammell>It sounds like you get to buy some fun toys then that also, I guess it's, you mentioned

00:14:59.360 --> 00:15:03.000
<v Chris Gammell>some of it's simple toys, but I mean, when it starts getting into the RF space, I imagine

00:15:03.000 --> 00:15:09.300
<v Chris Gammell>that starts moving, moving you up the, the purchase list of all the fancy RF measurement

00:15:09.300 --> 00:15:09.580
<v Chris Gammell>equipment.

00:15:09.960 --> 00:15:10.100
<v Chris Gammell>Yeah.

00:15:10.280 --> 00:15:11.200
<v Colin O'Flynn>I'm lucky there though.

00:15:11.320 --> 00:15:17.460
<v Colin O'Flynn>I, you know, actually I sort of co-located with a professor at Dalhousie that does Hans

00:15:17.460 --> 00:15:18.000
<v Colin O'Flynn>RF, right?

00:15:18.040 --> 00:15:21.160
<v Colin O'Flynn>So he has like a 40 gigahertz spectrum analyzer.

00:15:21.160 --> 00:15:21.360
<v Colin O'Flynn>He's got the good stuff.

00:15:21.540 --> 00:15:21.680
<v Colin O'Flynn>Yeah.

00:15:21.680 --> 00:15:22.020
<v Colin O'Flynn>Yeah, exactly.

00:15:22.200 --> 00:15:22.340
<v Colin O'Flynn>Right.

00:15:22.420 --> 00:15:23.420
<v Colin O'Flynn>Like anechoic chamber.

00:15:23.660 --> 00:15:26.880
<v Colin O'Flynn>So, so I like half of the, half of what I'm doing is trying to think, okay, well we have

00:15:26.880 --> 00:15:31.220
<v Colin O'Flynn>access to this, like, let's make sure we use it for, for doing this stuff.

00:15:31.540 --> 00:15:33.680
<v Colin O'Flynn>So this works wasn't, but yeah, exactly.

00:15:33.800 --> 00:15:37.600
<v Colin O'Flynn>I really want to, that's like the next stuff I want to look at is a little more on EM, like,

00:15:37.700 --> 00:15:40.680
<v Colin O'Flynn>you know, long, medium, long range, which I think is super interesting.

00:15:41.040 --> 00:15:41.140
<v Chris Gammell>Yeah.

00:15:41.140 --> 00:15:44.860
<v Chris Gammell>I mean, it's really, it's interesting, you know, looking at your, the stuff you've been

00:15:44.860 --> 00:15:45.180
<v Chris Gammell>doing.

00:15:45.420 --> 00:15:50.260
<v Chris Gammell>I feel like I keep coming across it in my work where I'm like, oh, this is what Colin was

00:15:50.260 --> 00:15:50.740
<v Chris Gammell>talking about.

00:15:50.740 --> 00:15:55.400
<v Chris Gammell>Like, as I get more into the space and just, you know, think more things that need to be

00:15:55.400 --> 00:15:59.960
<v Chris Gammell>encrypted, more, you know, putting secure elements onto things, all the trust zone stuff, like

00:15:59.960 --> 00:16:03.900
<v Chris Gammell>things that I know that you've, you've played with, you've, you've tested and stuff in the

00:16:03.900 --> 00:16:04.220
<v Chris Gammell>past.

00:16:04.220 --> 00:16:06.220
<v Chris Gammell>And it's like, oh, this is, this is why it's here.

00:16:06.360 --> 00:16:09.580
<v Chris Gammell>So it's good to have, just to also contextualize for people that are like listening to like,

00:16:09.640 --> 00:16:11.020
<v Chris Gammell>well, why do I care about security?

00:16:11.080 --> 00:16:14.020
<v Chris Gammell>It's like, well, maybe you don't right now, but at some point you might, you know, like

00:16:14.020 --> 00:16:15.580
<v Chris Gammell>yeah, it's going to come.

00:16:15.720 --> 00:16:16.260
<v Chris Gammell>That's the thing.

00:16:16.260 --> 00:16:21.100
<v Chris Gammell>I mean, broad strokes across the industry, are things steady state?

00:16:21.200 --> 00:16:21.700
<v Chris Gammell>Are they improving?

00:16:22.000 --> 00:16:24.280
<v Chris Gammell>Are they, you know, like, are we seeing more focus on this?

00:16:24.460 --> 00:16:25.640
<v Chris Gammell>Less focus doesn't really matter.

00:16:25.780 --> 00:16:26.140
<v Chris Gammell>I don't know.

00:16:26.140 --> 00:16:29.480
<v Chris Gammell>I don't know like where, where the state of the hardware security industry is.

00:16:30.020 --> 00:16:30.120
<v Colin O'Flynn>Yeah.

00:16:30.180 --> 00:16:34.760
<v Colin O'Flynn>I mean, there's definitely more focus on, especially, you know, the, the thing I find interesting

00:16:34.760 --> 00:16:39.900
<v Colin O'Flynn>is when I sort of started this, I don't know, like public chips that talked about side

00:16:39.900 --> 00:16:42.960
<v Colin O'Flynn>channel fault injection resistance were super low.

00:16:42.960 --> 00:16:49.140
<v Colin O'Flynn>So, um, you know, there is one or two vaguely mentioning a bit now, like the, the, the

00:16:49.140 --> 00:16:51.900
<v Colin O'Flynn>Raspberry Pi, that new device, the 2350.

00:16:51.900 --> 00:16:52.480
<v Colin O'Flynn>Oh, 2350.

00:16:52.840 --> 00:16:53.040
<v Chris Gammell>Yeah.

00:16:53.040 --> 00:16:53.200
<v Chris Gammell>Yeah.

00:16:53.200 --> 00:16:53.940
<v Colin O'Flynn>Like that.

00:16:54.460 --> 00:16:58.920
<v Colin O'Flynn>A, they went really far out of their way to add a bunch of more fault injection, but

00:16:58.920 --> 00:17:02.840
<v Colin O'Flynn>countermeasures and, and even just tools that, you know, you can use as a developer.

00:17:03.640 --> 00:17:05.080
<v Colin O'Flynn>Um, they do.

00:17:05.240 --> 00:17:08.920
<v Chris Gammell>When they were on the show too, they were also promoting how they ran the contests and they

00:17:08.920 --> 00:17:12.320
<v Chris Gammell>were promoting, you know, trying to get people to test it and push the limits of it.

00:17:12.320 --> 00:17:16.980
<v Chris Gammell>Like that, that security focus was definitely a thing they wanted to talk about.

00:17:17.040 --> 00:17:17.160
<v Chris Gammell>Right.

00:17:17.540 --> 00:17:17.800
<v Colin O'Flynn>Yeah.

00:17:17.800 --> 00:17:20.380
<v Colin O'Flynn>And that would be like, um, was unheard of.

00:17:20.400 --> 00:17:23.040
<v Colin O'Flynn>I mean, they, and they really pushed it harder than anyone did.

00:17:23.500 --> 00:17:25.000
<v Colin O'Flynn>Um, but you know, there's a few others.

00:17:25.000 --> 00:17:28.720
<v Colin O'Flynn>So like we've been involved a bit with the open Titan project, which was an open source

00:17:28.720 --> 00:17:35.340
<v Colin O'Flynn>route of trust, um, that Google was heavily involved with, um, and is basically going to

00:17:35.340 --> 00:17:38.900
<v Colin O'Flynn>be using that as part of their, um, some shipping hardware.

00:17:38.900 --> 00:17:42.200
<v Colin O'Flynn>So there's some, you know, blog posts or PR where they talk about where they're going

00:17:42.200 --> 00:17:43.040
<v Colin O'Flynn>to end up shipping it.

00:17:43.560 --> 00:17:48.400
<v Colin O'Flynn>Um, but yeah, so like there's, there's been a big push, right.

00:17:48.500 --> 00:17:53.860
<v Colin O'Flynn>For not just security, but like actually having more transparent security, I think is the,

00:17:53.860 --> 00:17:57.760
<v Colin O'Flynn>the thing that is more helpful for everyone.

00:17:57.760 --> 00:17:57.880
<v Colin O'Flynn>Yeah.

00:17:58.840 --> 00:17:59.160
<v Chris Gammell>Yeah.

00:17:59.220 --> 00:18:01.700
<v Chris Gammell>So you like, traditionally it was like, well, no one's going to find out about this.

00:18:01.700 --> 00:18:03.060
<v Chris Gammell>Cause like, how would they possibly get information?

00:18:03.060 --> 00:18:07.280
<v Chris Gammell>And now it's like, I ask a GPT to give me all the information.

00:18:07.280 --> 00:18:08.260
<v Chris Gammell>It's like, here you go, sir.

00:18:08.740 --> 00:18:09.000
<v Chris Gammell>Yeah.

00:18:09.680 --> 00:18:09.940
<v Chris Gammell>Yeah.

00:18:10.020 --> 00:18:10.180
<v Chris Gammell>Yeah.

00:18:10.180 --> 00:18:10.780
<v Chris Gammell>No, very true.

00:18:10.880 --> 00:18:10.980
<v Chris Gammell>Right.

00:18:11.040 --> 00:18:12.200
<v Chris Gammell>And like, yeah.

00:18:13.440 --> 00:18:17.720
<v Chris Gammell>Can you remind me or can remind me and the audience about like root of trust?

00:18:17.760 --> 00:18:22.420
<v Chris Gammell>I remember Laura from, uh, Laura Abbott was on the show.

00:18:22.620 --> 00:18:25.920
<v Chris Gammell>She's at that server company that I always forget the name of.

00:18:26.340 --> 00:18:27.240
<v Chris Gammell>They were also working.

00:18:27.800 --> 00:18:28.120
<v Chris Gammell>Yeah.

00:18:28.300 --> 00:18:28.440
<v Chris Gammell>Yeah.

00:18:28.480 --> 00:18:28.740
<v Chris Gammell>Okay.

00:18:28.860 --> 00:18:29.040
<v Chris Gammell>Yep.

00:18:29.300 --> 00:18:31.520
<v Colin O'Flynn>I was like, that seems like the only cool server company.

00:18:31.520 --> 00:18:34.640
<v Chris Gammell>I don't know who else you have pretty much, uh, them and impair impair.

00:18:34.760 --> 00:18:34.860
<v Chris Gammell>Okay.

00:18:34.880 --> 00:18:35.060
<v Chris Gammell>Yeah.

00:18:35.060 --> 00:18:35.360
<v Chris Gammell>Okay.

00:18:35.440 --> 00:18:36.140
<v Chris Gammell>I take it back.

00:18:36.200 --> 00:18:36.520
<v Chris Gammell>I'm sorry.

00:18:36.640 --> 00:18:36.800
<v Chris Gammell>I'm fair.

00:18:37.260 --> 00:18:37.500
<v Chris Gammell>Yeah.

00:18:37.700 --> 00:18:37.840
<v Chris Gammell>Yeah.

00:18:37.840 --> 00:18:38.840
<v Chris Gammell>Those, those two together though.

00:18:38.880 --> 00:18:39.060
<v Chris Gammell>Yeah.

00:18:39.300 --> 00:18:41.020
<v Chris Gammell>Those, those, and then like what Dell.

00:18:42.340 --> 00:18:43.340
<v Chris Gammell>So cool.

00:18:43.840 --> 00:18:45.000
<v Chris Gammell>Dude, you're getting a Dell.

00:18:46.100 --> 00:18:47.720
<v Colin O'Flynn>I'm sure they're doing cool stuff right there.

00:18:47.880 --> 00:18:51.080
<v Chris Gammell>I'm sure they are, but it's just like, you know, they're stalwart in the industry.

00:18:51.220 --> 00:18:53.160
<v Chris Gammell>It doesn't have the, uh, it doesn't have the same right.

00:18:53.380 --> 00:18:53.580
<v Colin O'Flynn>Cool.

00:18:53.720 --> 00:18:57.560
<v Chris Gammell>I mean, Dell might have a podcast, but is it, is it as good?

00:18:57.860 --> 00:18:58.600
<v Chris Gammell>I don't know.

00:18:58.680 --> 00:19:00.360
<v Colin O'Flynn>You know, you just wait.

00:19:00.360 --> 00:19:01.680
<v Colin O'Flynn>Mm-hmm.

00:19:02.680 --> 00:19:03.020
<v Chris Gammell>Um, yeah.

00:19:03.040 --> 00:19:06.140
<v Chris Gammell>So could you remind people what, so like how the root of trust stuff fits in as well?

00:19:06.160 --> 00:19:10.200
<v Chris Gammell>Cause that's like a, that's like a security element that's on the server board.

00:19:10.300 --> 00:19:10.620
<v Chris Gammell>Is that right?

00:19:10.840 --> 00:19:11.160
<v Colin O'Flynn>Yeah.

00:19:11.200 --> 00:19:12.400
<v Colin O'Flynn>So it can be on anything.

00:19:12.400 --> 00:19:16.040
<v Colin O'Flynn>I mean, the idea of the root of trust is this is what's booting your system.

00:19:16.640 --> 00:19:17.000
<v Colin O'Flynn>Right.

00:19:17.080 --> 00:19:21.660
<v Colin O'Flynn>So, um, the normal use case for it, and I'm, I'm going to kind of generalize cause someone's

00:19:21.660 --> 00:19:23.640
<v Colin O'Flynn>going to definitely be like, that's not actually true.

00:19:23.900 --> 00:19:26.580
<v Colin O'Flynn>Um, I don't actually know all the industry definitions.

00:19:26.580 --> 00:19:26.700
<v Colin O'Flynn>Don't worry.

00:19:26.760 --> 00:19:28.060
<v Colin O'Flynn>Colin's just explaining it for me.

00:19:28.060 --> 00:19:28.560
<v Chris Gammell>I'm a dummy.

00:19:28.700 --> 00:19:29.940
<v Chris Gammell>Just we'll, you know, Henry.

00:19:30.140 --> 00:19:30.780
<v Chris Gammell>I have to have that.

00:19:30.880 --> 00:19:32.940
<v Colin O'Flynn>See, this is, this is one of the downsides.

00:19:33.620 --> 00:19:37.360
<v Colin O'Flynn>You know, I don't know if it's feels like you're complaining as a hipster now, but now

00:19:37.360 --> 00:19:40.780
<v Colin O'Flynn>that there's more people involved before you could totally get away with like, it was totally

00:19:40.780 --> 00:19:41.400
<v Chris Gammell>cool before.

00:19:41.520 --> 00:19:41.760
<v Chris Gammell>Yeah.

00:19:41.760 --> 00:19:42.020
<v Chris Gammell>Yeah.

00:19:42.020 --> 00:19:46.300
<v Colin O'Flynn>Being like more vague or being, you know, being a bit wrong cause no one knew, but now lots

00:19:46.300 --> 00:19:46.880
<v Colin O'Flynn>of people know.

00:19:48.260 --> 00:19:48.940
<v Colin O'Flynn>Anyway, but yeah.

00:19:48.940 --> 00:19:51.080
<v Colin O'Flynn>So the root of trust, the idea is this is booting your system.

00:19:51.220 --> 00:19:53.620
<v Colin O'Flynn>So you have an untrusted system.

00:19:53.940 --> 00:19:55.320
<v Colin O'Flynn>How do you boot it up?

00:19:55.760 --> 00:19:58.800
<v Colin O'Flynn>And so the idea is you have this one super secure device.

00:19:59.020 --> 00:20:02.160
<v Colin O'Flynn>That's like checking every stage of the firmware.

00:20:02.160 --> 00:20:06.360
<v Colin O'Flynn>Um, so devices will have this internally, right?

00:20:06.360 --> 00:20:11.740
<v Colin O'Flynn>So some devices will say, you know, they have the first stage bootloader is in ROM and normally

00:20:11.740 --> 00:20:12.860
<v Colin O'Flynn>can never be changed.

00:20:12.860 --> 00:20:13.160
<v Colin O'Flynn>Right.

00:20:13.180 --> 00:20:15.920
<v Colin O'Flynn>And then that boots some external bootloader that boots the next one.

00:20:15.920 --> 00:20:18.240
<v Colin O'Flynn>So it's like bootstrapping your whole system securely.

00:20:18.920 --> 00:20:23.200
<v Colin O'Flynn>Um, it normally also can then handle stuff in the future.

00:20:23.200 --> 00:20:27.120
<v Colin O'Flynn>Like if you need to sign something or you're getting a firmware update, right?

00:20:27.120 --> 00:20:33.040
<v Colin O'Flynn>The root of trust is going to manage that and make sure you don't, you know, get a bad

00:20:33.040 --> 00:20:35.520
<v Colin O'Flynn>firmware update or malicious really, not just bad.

00:20:35.900 --> 00:20:35.960
<v Chris Gammell>Right.

00:20:36.680 --> 00:20:43.480
<v Chris Gammell>So is it like the idea is me as a user, I log into a server built by Google, whatever, multiple

00:20:43.480 --> 00:20:45.140
<v Chris Gammell>layers of software on top of it, whatever.

00:20:45.140 --> 00:20:48.500
<v Chris Gammell>And I'm trying to say, how do I know this is actually the server it is?

00:20:48.500 --> 00:20:51.340
<v Chris Gammell>And I could dive, keep diving down, down, down, down, down.

00:20:51.360 --> 00:20:53.820
<v Chris Gammell>And then at some point this is the chip that says it.

00:20:53.880 --> 00:20:54.560
<v Chris Gammell>Is that kind of the idea?

00:20:54.700 --> 00:20:54.980
<v Colin O'Flynn>Exactly.

00:20:55.060 --> 00:20:55.260
<v Colin O'Flynn>Yes.

00:20:55.280 --> 00:20:57.180
<v Colin O'Flynn>This is your like total, right.

00:20:57.240 --> 00:20:58.220
<v Colin O'Flynn>This is your root of trust.

00:20:58.280 --> 00:20:59.520
<v Colin O'Flynn>This is how you trust everything.

00:21:00.100 --> 00:21:05.280
<v Colin O'Flynn>And like the idea of open Titan was they had kind of said, you know, well, look, how do

00:21:05.280 --> 00:21:06.380
<v Colin O'Flynn>you trust the root of trust?

00:21:06.820 --> 00:21:07.140
<v Colin O'Flynn>Right.

00:21:07.260 --> 00:21:07.460
<v Colin O'Flynn>Right.

00:21:08.140 --> 00:21:08.880
<v Chris Gammell>Who says that?

00:21:09.020 --> 00:21:09.840
<v Chris Gammell>Who watches the watchman?

00:21:10.100 --> 00:21:10.420
<v Colin O'Flynn>Exactly.

00:21:10.660 --> 00:21:12.960
<v Colin O'Flynn>And like, and you know, this was also around like the Snowden time.

00:21:13.040 --> 00:21:14.200
<v Colin O'Flynn>I think they were starting this project.

00:21:14.200 --> 00:21:15.380
<v Colin O'Flynn>So it was a real question.

00:21:15.380 --> 00:21:19.360
<v Colin O'Flynn>Like, you know, what if the government forces Google to tamper with your server?

00:21:19.620 --> 00:21:23.920
<v Colin O'Flynn>You know, they, they, Google wants you to believe that they have no tampering.

00:21:24.060 --> 00:21:28.360
<v Colin O'Flynn>So, so they kind of said like, well, if it's an open source chip, then you can in theory

00:21:28.360 --> 00:21:32.060
<v Colin O'Flynn>do, you know, attestation that the design is what's running.

00:21:32.280 --> 00:21:35.980
<v Colin O'Flynn>There's some tricky bits in there, but the general idea, right.

00:21:35.980 --> 00:21:37.240
<v Colin O'Flynn>It gets you all the way up.

00:21:38.240 --> 00:21:43.180
<v Colin O'Flynn>But I mean, and people use them like for little, you know, Chromebooks, phones have these like

00:21:43.180 --> 00:21:43.640
<v Colin O'Flynn>got it.

00:21:43.720 --> 00:21:43.900
<v Colin O'Flynn>Yeah.

00:21:44.020 --> 00:21:45.420
<v Chris Gammell>Kind of everything will use this.

00:21:45.420 --> 00:21:49.800
<v Chris Gammell>And so it's just there to, I mean, it's got all this stuff to prove itself.

00:21:49.920 --> 00:21:53.260
<v Chris Gammell>I'm looking at a data sheet, which Colin also is seeing my screen.

00:21:53.340 --> 00:21:57.060
<v Chris Gammell>So at least we can have a shared context here, but it has a lot of stuff in it that I don't

00:21:57.660 --> 00:21:58.980
<v Chris Gammell>know why it would need it.

00:21:59.040 --> 00:22:02.280
<v Chris Gammell>But I guess, I mean, random number generator.

00:22:02.420 --> 00:22:02.780
<v Chris Gammell>Okay.

00:22:02.960 --> 00:22:03.560
<v Chris Gammell>Key manager.

00:22:03.760 --> 00:22:03.980
<v Chris Gammell>Sure.

00:22:04.100 --> 00:22:05.080
<v Chris Gammell>Like power monitoring.

00:22:05.080 --> 00:22:09.860
<v Chris Gammell>Like it's, it's just able to like look internally and say, everything's cool boss.

00:22:10.000 --> 00:22:10.600
<v Chris Gammell>That's kind of the idea.

00:22:10.860 --> 00:22:11.240
<v Colin O'Flynn>Right.

00:22:11.320 --> 00:22:15.960
<v Colin O'Flynn>So the thing is the, um, and you know, you have to remember too, they're pretty somewhat

00:22:15.960 --> 00:22:19.520
<v Colin O'Flynn>generic, but, um, they're often booting stuff.

00:22:19.520 --> 00:22:22.860
<v Colin O'Flynn>So like they might need to read from an external spy and validate it.

00:22:22.860 --> 00:22:29.300
<v Colin O'Flynn>And then, you know, they can emulate that spy to a system, um, so that someone doesn't

00:22:29.300 --> 00:22:33.540
<v Colin O'Flynn>do like a attack where they switch, right.

00:22:33.540 --> 00:22:38.440
<v Colin O'Flynn>What the spy image is there's the classic sort of attack where the, the system looks

00:22:38.440 --> 00:22:38.660
<v Colin O'Flynn>right.

00:22:38.700 --> 00:22:42.320
<v Colin O'Flynn>That it validates some firmware image and says, okay, that firmware image looks good.

00:22:42.900 --> 00:22:45.600
<v Colin O'Flynn>Um, so that's when it does the time of check.

00:22:45.700 --> 00:22:48.420
<v Colin O'Flynn>And then when it goes to use it, it like reloads the spy flash.

00:22:48.420 --> 00:22:50.100
<v Colin O'Flynn>Cause it doesn't have everything in memory.

00:22:50.100 --> 00:22:51.420
<v Colin O'Flynn>It didn't have enough space in memory.

00:22:51.780 --> 00:22:54.240
<v Colin O'Flynn>Um, and the attacker switches it at that moment in time.

00:22:54.240 --> 00:22:54.740
<v Colin O'Flynn>I see.

00:22:55.100 --> 00:22:55.360
<v Chris Gammell>Right.

00:22:55.720 --> 00:22:57.420
<v Chris Gammell>It's like the side of hand, but with memory.

00:22:57.720 --> 00:22:58.080
<v Colin O'Flynn>Exactly.

00:22:58.200 --> 00:22:59.480
<v Colin O'Flynn>It's like, that's like, that was a common.

00:22:59.560 --> 00:23:02.820
<v Colin O'Flynn>I mean, it's less, more people are doing validation for the whole thing.

00:23:02.820 --> 00:23:02.980
<v Colin O'Flynn>Yeah.

00:23:03.080 --> 00:23:04.740
<v Colin O'Flynn>Avoiding that, but it's still an issue.

00:23:04.740 --> 00:23:08.480
<v Colin O'Flynn>So, so if you have a root of trust, right, it's actually forcing that.

00:23:08.480 --> 00:23:12.320
<v Colin O'Flynn>Like, there's just no, there's nowhere to switch it because if you switch it, the root

00:23:12.320 --> 00:23:15.660
<v Colin O'Flynn>of trust is going to see that or not even possibly not even read it.

00:23:15.680 --> 00:23:16.800
<v Colin O'Flynn>It might just cash everything.

00:23:17.220 --> 00:23:20.480
<v Chris Gammell>I guess, I mean, Laura was on the show and prior to that, I think it maybe heard it one

00:23:20.480 --> 00:23:23.280
<v Chris Gammell>or two times, but I just didn't really understand it.

00:23:23.320 --> 00:23:27.700
<v Chris Gammell>I think the only thing I had heard about it under was the, um, when there was like that

00:23:27.700 --> 00:23:29.940
<v Chris Gammell>scare about there being a chip on one of the servers.

00:23:29.940 --> 00:23:35.580
<v Chris Gammell>And then I remember Joe Fitz made that, that tiny sticker with like the little pencil tip.

00:23:35.860 --> 00:23:37.600
<v Chris Gammell>I want to believe or whatever.

00:23:37.880 --> 00:23:38.780
<v Chris Gammell>I don't remember what it was.

00:23:38.800 --> 00:23:45.900
<v Chris Gammell>It was like, it was like someone had apparently found a malicious chip on a server PCB.

00:23:46.300 --> 00:23:46.500
<v Chris Gammell>Right.

00:23:47.080 --> 00:23:47.480
<v Chris Gammell>Yeah.

00:23:48.020 --> 00:23:49.340
<v Colin O'Flynn>Which, and I don't even know.

00:23:49.420 --> 00:23:51.900
<v Colin O'Flynn>And I think the chip that you use, like unvalidated, right.

00:23:51.900 --> 00:23:54.760
<v Chris Gammell>It was like, it was like a lot of fear mongering and that sort of thing.

00:23:54.760 --> 00:23:55.000
<v Chris Gammell>Yeah.

00:23:55.120 --> 00:23:59.860
<v Colin O'Flynn>Well, and I remember the big thing was the chip they used in the photo was, uh, the, I think

00:23:59.860 --> 00:24:01.580
<v Colin O'Flynn>it was just like a ball in or something.

00:24:01.580 --> 00:24:01.800
<v Colin O'Flynn>Right.

00:24:01.800 --> 00:24:02.920
<v Colin O'Flynn>It was like an art, it was like totally.

00:24:03.420 --> 00:24:06.320
<v Colin O'Flynn>And I don't know if that's what they were saying was delicious or they were just like,

00:24:06.360 --> 00:24:07.360
<v Colin O'Flynn>here's a cool looking chip.

00:24:07.480 --> 00:24:08.380
<v Colin O'Flynn>Let's put that on the tip.

00:24:08.380 --> 00:24:08.480
<v Colin O'Flynn>Yeah.

00:24:08.480 --> 00:24:08.680
<v Colin O'Flynn>Right.

00:24:08.780 --> 00:24:08.980
<v Colin O'Flynn>Right.

00:24:09.560 --> 00:24:09.760
<v Chris Gammell>Right.

00:24:09.940 --> 00:24:11.060
<v Chris Gammell>You never know with us.

00:24:11.540 --> 00:24:15.140
<v Chris Gammell>To a journalist, it's just a black of black blob on a board sort of thing.

00:24:15.640 --> 00:24:15.960
<v Chris Gammell>Yeah.

00:24:15.960 --> 00:24:16.100
<v Chris Gammell>Yeah.

00:24:16.100 --> 00:24:16.420
<v Chris Gammell>Exactly.

00:24:16.640 --> 00:24:16.980
<v Chris Gammell>Yeah.

00:24:17.280 --> 00:24:17.460
<v Chris Gammell>Yeah.

00:24:17.460 --> 00:24:20.820
<v Chris Gammell>Um, but yeah.

00:24:20.940 --> 00:24:21.180
<v Chris Gammell>Okay.

00:24:21.380 --> 00:24:28.660
<v Chris Gammell>How did we get here talking about the, the open Titan and the state of the industry, state

00:24:28.660 --> 00:24:31.800
<v Colin O'Flynn>of the industry, industry is looking better, I think is the summary.

00:24:31.980 --> 00:24:32.220
<v Chris Gammell>Okay.

00:24:32.340 --> 00:24:33.700
<v Chris Gammell>So there's just more focus on it.

00:24:33.700 --> 00:24:37.040
<v Chris Gammell>It sounds like, and it's like, there's more, uh, people more focused.

00:24:37.120 --> 00:24:38.160
<v Colin O'Flynn>And I think personally, right.

00:24:38.160 --> 00:24:41.280
<v Colin O'Flynn>Like, I think a lot of people don't think the open solution is good.

00:24:41.860 --> 00:24:43.480
<v Colin O'Flynn>Um, yeah, which I think is nice.

00:24:43.480 --> 00:24:46.620
<v Colin O'Flynn>Like, as I said, the Reservoir Pi scene, what they did is really cool.

00:24:46.620 --> 00:24:51.440
<v Colin O'Flynn>Um, you know, seeing stuff like open Titan and, and they're sort of different levels,

00:24:51.440 --> 00:24:56.280
<v Colin O'Flynn>but it's, it's not just that there's, you know, venture capitalists saying, let's make

00:24:56.280 --> 00:25:01.780
<v Colin O'Flynn>another proprietary solution or AI does X, Y, Z, right.

00:25:01.840 --> 00:25:06.260
<v Colin O'Flynn>That's like, here's an actual thing that could actually help people seems reasonably secure.

00:25:06.880 --> 00:25:10.960
<v Chris Gammell>You know, one thing that you always kind of bring up is like these attack vectors and

00:25:10.960 --> 00:25:18.260
<v Chris Gammell>like, uh, how people think about it, but how should people listening, think about it for

00:25:18.260 --> 00:25:21.620
<v Chris Gammell>their own products as well of like, Oh, this is what someone is likely to do.

00:25:21.620 --> 00:25:24.580
<v Chris Gammell>And then I guess what some, something I always think about is like, Oh, well they, they need

00:25:24.580 --> 00:25:27.540
<v Chris Gammell>to have physical access sometimes, but not always, you know?

00:25:27.540 --> 00:25:30.880
<v Chris Gammell>So like the likelihood of certain things, is there like a good guideline for that sort

00:25:30.880 --> 00:25:31.180
<v Chris Gammell>of thing?

00:25:31.300 --> 00:25:34.700
<v Chris Gammell>So if I'm designing a product, what do I, what do I have to care about?

00:25:34.700 --> 00:25:39.120
<v Chris Gammell>Because there's no like test other than like maybe hiring a security firm because there's

00:25:39.120 --> 00:25:40.060
<v Chris Gammell>just general guidelines.

00:25:40.660 --> 00:25:40.780
<v Colin O'Flynn>Yeah.

00:25:40.880 --> 00:25:44.160
<v Colin O'Flynn>And there is, and to be honest, and this is where I'm bad again, this is where people

00:25:44.160 --> 00:25:49.480
<v Colin O'Flynn>listening will know and correct me, uh, because there is some more recent industry guidelines

00:25:49.480 --> 00:25:50.540
<v Colin O'Flynn>that I've seen released.

00:25:50.640 --> 00:25:51.520
<v Colin O'Flynn>I haven't even read them yet.

00:25:51.600 --> 00:25:51.780
<v Chris Gammell>Oh yeah.

00:25:52.200 --> 00:25:58.980
<v Colin O'Flynn>Um, but you know, to, to go back to an older one, like arm had their, um, PSA platform security

00:25:58.980 --> 00:25:59.800
<v Colin O'Flynn>architecture, right.

00:25:59.800 --> 00:26:07.560
<v Colin O'Flynn>Um, which was a level and they basically had several levels of, um, if you want to get

00:26:07.560 --> 00:26:09.960
<v Colin O'Flynn>your device secure, here's like what you can do.

00:26:10.020 --> 00:26:13.880
<v Colin O'Flynn>So the highest level involved actual side channel and fault injection testing.

00:26:14.360 --> 00:26:19.560
<v Colin O'Flynn>Um, I liked this one cause the lower levels were just kind of, um, checklist.

00:26:19.780 --> 00:26:21.040
<v Colin O'Flynn>Like, so you could do it yourself.

00:26:21.500 --> 00:26:24.060
<v Colin O'Flynn>I don't, you might have to pay money to get an actual PSA.

00:26:24.140 --> 00:26:28.960
<v Colin O'Flynn>I don't actually know what the system is for commercial, but, um, they did have a bunch

00:26:28.960 --> 00:26:31.900
<v Colin O'Flynn>of stuff you could download that would say, Hey, and you know, and it would be reasonable

00:26:31.900 --> 00:26:32.180
<v Colin O'Flynn>things.

00:26:32.240 --> 00:26:33.120
<v Colin O'Flynn>I liked it because they're big.

00:26:33.180 --> 00:26:33.340
<v Colin O'Flynn>Okay.

00:26:33.340 --> 00:26:35.280
<v Colin O'Flynn>Is the J take port locked unlocked, right?

00:26:35.360 --> 00:26:37.740
<v Colin O'Flynn>Like things that people actually mess up all the time.

00:26:37.920 --> 00:26:38.280
<v Colin O'Flynn>Right.

00:26:38.460 --> 00:26:38.800
<v Colin O'Flynn>Totally.

00:26:39.220 --> 00:26:43.600
<v Colin O'Flynn>So, so you could get like a different, and they had some, a few example threat models,

00:26:43.600 --> 00:26:49.260
<v Colin O'Flynn>um, where it's like, okay, here's if you're making a smart meter, I think was one.

00:26:49.260 --> 00:26:52.480
<v Chris Gammell>And yeah, I'm looking at this page right now.

00:26:52.540 --> 00:26:55.060
<v Chris Gammell>It says smart meter asset tracker network camera.

00:26:55.300 --> 00:26:55.700
<v Chris Gammell>Yeah.

00:26:56.240 --> 00:26:57.960
<v Chris Gammell>TM, TMSAs, right?

00:26:57.960 --> 00:26:58.720
<v Chris Gammell>Yeah.

00:26:58.880 --> 00:27:01.420
<v Colin O'Flynn>Threat model security analysis documents, I think.

00:27:01.580 --> 00:27:04.560
<v Colin O'Flynn>So, so that's kind of showing you, okay, if you have this doc here, you know, and it

00:27:04.560 --> 00:27:08.260
<v Colin O'Flynn>goes over like, okay, well, would people have physical access to it or not?

00:27:08.260 --> 00:27:10.820
<v Chris Gammell>Or, um, stuff like that.

00:27:11.060 --> 00:27:11.320
<v Chris Gammell>So.

00:27:12.240 --> 00:27:17.860
<v Chris Gammell>Not lost on me that I'm downloading a binary, uh, from, uh, a thing about like threat models.

00:27:18.320 --> 00:27:18.760
<v Chris Gammell>Yeah.

00:27:19.640 --> 00:27:21.600
<v Chris Gammell>Excel file on your, on your computer.

00:27:21.780 --> 00:27:22.420
<v Chris Gammell>No big deal.

00:27:22.720 --> 00:27:23.380
<v Chris Gammell>Macro enabled.

00:27:23.460 --> 00:27:23.900
<v Chris Gammell>Of course.

00:27:24.200 --> 00:27:25.040
<v Chris Gammell>Why would it happen?

00:27:25.180 --> 00:27:25.420
<v Chris Gammell>Right.

00:27:25.660 --> 00:27:25.980
<v Chris Gammell>Right.

00:27:25.980 --> 00:27:26.240
<v Chris Gammell>Yeah.

00:27:26.640 --> 00:27:28.940
<v Chris Gammell>And then that's how I got hacked by VBA.

00:27:29.440 --> 00:27:29.800
<v Chris Gammell>Yeah.

00:27:29.800 --> 00:27:29.960
<v Chris Gammell>Yeah.

00:27:29.960 --> 00:27:33.100
<v Chris Gammell>Yeah.

00:27:33.100 --> 00:27:38.520
<v Chris Gammell>I mean, yeah, they definitely give you some, some stuff here and I'm sure they would definitely

00:27:38.520 --> 00:27:42.300
<v Chris Gammell>take your money if you wanted to, uh, pay them to certify it.

00:27:42.300 --> 00:27:42.480
<v Chris Gammell>Right.

00:27:42.800 --> 00:27:43.080
<v Colin O'Flynn>Yeah.

00:27:43.160 --> 00:27:46.060
<v Colin O'Flynn>But, but, but I mean, and there is, as I said, there's, I've seen some other, like there

00:27:46.060 --> 00:27:53.000
<v Colin O'Flynn>is some, uh, I think the UK government just released a kind of IOT smart device, um, guidelines

00:27:53.000 --> 00:27:53.620
<v Colin O'Flynn>as well.

00:27:54.000 --> 00:27:57.980
<v Colin O'Flynn>Um, there's been a few kind of like that and they, they're pretty, again, like I think the,

00:27:57.980 --> 00:28:01.680
<v Colin O'Flynn>the cool thing we're seeing is that things are, are reasonable.

00:28:01.680 --> 00:28:02.700
<v Colin O'Flynn>I want to say, right.

00:28:03.000 --> 00:28:03.400
<v Colin O'Flynn>Yeah.

00:28:03.780 --> 00:28:05.280
<v Colin O'Flynn>They're, they're looking at real threats.

00:28:05.280 --> 00:28:10.940
<v Chris Gammell>I know there's some like, uh, upcoming regulations in the CRA, I think is the one in the EU.

00:28:11.700 --> 00:28:15.060
<v Chris Gammell>Um, that's like a requirement for like being able to OTA stuff.

00:28:15.500 --> 00:28:20.680
<v Chris Gammell>And then there's trust mark or something like that in the U S but which is less of a, there's

00:28:20.680 --> 00:28:22.100
<v Chris Gammell>like some kind of marking as well.

00:28:22.400 --> 00:28:23.180
<v Chris Gammell>But I don't know.

00:28:23.200 --> 00:28:25.980
<v Chris Gammell>Whenever they think about these things too, they're always like, Oh, you're running like

00:28:25.980 --> 00:28:29.400
<v Chris Gammell>a Linux box and you left SSH open.

00:28:29.400 --> 00:28:30.600
<v Chris Gammell>And like, yeah.

00:28:30.760 --> 00:28:30.920
<v Chris Gammell>Yeah.

00:28:30.920 --> 00:28:34.620
<v Colin O'Flynn>Which is why the RPSA one I like, because it's, again, this is more like embedded, right?

00:28:34.660 --> 00:28:35.620
<v Colin O'Flynn>That's their whole thing.

00:28:35.840 --> 00:28:36.800
<v Colin O'Flynn>So I think that's, it is.

00:28:36.860 --> 00:28:41.540
<v Colin O'Flynn>If you're doing embedded, like it's a good, you know, as I said, at least at one or level

00:28:41.540 --> 00:28:43.240
<v Colin O'Flynn>one or two, it's like a checklist you can do.

00:28:43.320 --> 00:28:48.620
<v Colin O'Flynn>I think you can even just certify without paying from, you know, my information is years

00:28:48.620 --> 00:28:52.980
<v Colin O'Flynn>old, not checked back check whatsoever, but well, this is the empire.

00:28:53.060 --> 00:28:53.960
<v Chris Gammell>So you don't have to worry about fact.

00:28:54.040 --> 00:28:54.360
<v Chris Gammell>Perfect.

00:28:54.560 --> 00:28:55.280
<v Chris Gammell>I like it.

00:28:55.400 --> 00:28:56.540
<v Chris Gammell>Not another thing we do here.

00:28:56.820 --> 00:28:57.700
<v Chris Gammell>Don't worry about us.

00:28:57.700 --> 00:28:58.640
<v Chris Gammell>Uh, I don't know.

00:28:58.640 --> 00:28:59.800
<v Chris Gammell>I did work.

00:29:00.140 --> 00:29:05.360
<v Chris Gammell>I did have a consulting thing back in my day of like, that was, I had somewhat, the third

00:29:05.360 --> 00:29:08.380
<v Chris Gammell>party come in and like do like an audit, but like all the things they were doing, there

00:29:08.380 --> 00:29:12.700
<v Chris Gammell>was just like, Oh, are you not like a, like a raspberry pie?

00:29:12.980 --> 00:29:13.420
<v Chris Gammell>No.

00:29:13.880 --> 00:29:14.160
<v Chris Gammell>Okay.

00:29:14.240 --> 00:29:15.480
<v Chris Gammell>Then we don't know what to do with you.

00:29:16.800 --> 00:29:17.160
<v Chris Gammell>Okay.

00:29:18.060 --> 00:29:20.180
<v Chris Gammell>Then that's, that's less useful.

00:29:20.320 --> 00:29:23.900
<v Chris Gammell>You know, like, I, like, I, I think the thing is like, as like a product designer, it's like,

00:29:23.900 --> 00:29:25.260
<v Chris Gammell>I want to know the best practices.

00:29:25.260 --> 00:29:27.100
<v Chris Gammell>So things like this are good.

00:29:27.260 --> 00:29:27.660
<v Chris Gammell>I feel like.

00:29:27.840 --> 00:29:29.880
<v Chris Gammell>So what about the embedded side of things?

00:29:29.920 --> 00:29:33.660
<v Chris Gammell>I mean, like, are you focusing mostly, mostly in the embedded space and lower?

00:29:33.800 --> 00:29:35.860
<v Chris Gammell>I mean, like, are you doing, doing that kind of stuff these days?

00:29:36.660 --> 00:29:36.840
<v Colin O'Flynn>Yeah.

00:29:36.920 --> 00:29:42.920
<v Colin O'Flynn>That's, um, you know, I mean, my, my own background is less like Linux high level.

00:29:42.920 --> 00:29:48.440
<v Colin O'Flynn>Um, so most of the stuff is, you know, lower level, bare metal embedded looking at that.

00:29:49.100 --> 00:29:50.740
<v Colin O'Flynn>Um, yeah, yeah.

00:29:50.740 --> 00:29:51.880
<v Colin O'Flynn>A little architecture stuff.

00:29:51.920 --> 00:29:55.980
<v Colin O'Flynn>So, I mean, one newer thing I've worked on is, um, and I haven't worked on it so much.

00:29:56.000 --> 00:30:00.040
<v Colin O'Flynn>I worked on board designs and have been involved in it.

00:30:00.040 --> 00:30:02.880
<v Colin O'Flynn>So I feel like when I say worked on, I don't want to have any credit.

00:30:02.880 --> 00:30:04.860
<v Colin O'Flynn>Is this, um, chariot system?

00:30:04.860 --> 00:30:11.140
<v Colin O'Flynn>Um, and in particular, so design this board Sonata, which low risk, um, new age parent

00:30:11.140 --> 00:30:17.480
<v Colin O'Flynn>company, uh, had a project where they made this open source.

00:30:17.720 --> 00:30:23.600
<v Colin O'Flynn>Um, in fact, the idea is to have like an open source microcontroller kind of soft core device,

00:30:23.600 --> 00:30:23.760
<v Colin O'Flynn>right?

00:30:23.760 --> 00:30:27.620
<v Colin O'Flynn>That just feels like a microcontroller, but it's running something called cherry and cherry

00:30:27.620 --> 00:30:31.020
<v Colin O'Flynn>for IOT, which does all this compartmentalization.

00:30:31.020 --> 00:30:36.380
<v Colin O'Flynn>Um, so it, you know, rather than just saying, are you in like, if people are familiar with

00:30:36.380 --> 00:30:41.400
<v Colin O'Flynn>cortex M 33 and these devices, they have like a secure and a non-secure mode and, you know,

00:30:41.440 --> 00:30:45.460
<v Colin O'Flynn>different peripherals, you can say, okay, the spy device can only be accessed by the secure

00:30:45.460 --> 00:30:45.820
<v Colin O'Flynn>code.

00:30:46.540 --> 00:30:51.780
<v Colin O'Flynn>Um, and so cherry, it's kind of cool because you define compartments and you say, okay,

00:30:51.780 --> 00:30:55.740
<v Colin O'Flynn>the, you know, the spy driver code can access the spy peripheral and that's it.

00:30:56.000 --> 00:30:56.200
<v Chris Gammell>Huh?

00:30:56.400 --> 00:30:56.660
<v Chris Gammell>Okay.

00:30:56.860 --> 00:30:57.300
<v Colin O'Flynn>You are.

00:30:57.380 --> 00:31:00.680
<v Colin O'Flynn>And so it's like at a much more fine grain level.

00:31:00.680 --> 00:31:04.360
<v Colin O'Flynn>So it's kind of cool because you can say like, okay, the web server process can only

00:31:04.360 --> 00:31:08.640
<v Colin O'Flynn>access these certain things and like this API, stuff like that.

00:31:08.800 --> 00:31:09.160
<v Colin O'Flynn>Right.

00:31:09.220 --> 00:31:13.780
<v Colin O'Flynn>So if someone hacks it, it's just like, there's just no, you know, there's no way to read other

00:31:13.780 --> 00:31:14.100
<v Colin O'Flynn>memory.

00:31:14.240 --> 00:31:14.380
<v Colin O'Flynn>Yeah.

00:31:14.860 --> 00:31:15.140
<v Colin O'Flynn>Right.

00:31:15.300 --> 00:31:16.140
<v Colin O'Flynn>And that was the problem.

00:31:16.200 --> 00:31:20.680
<v Colin O'Flynn>You know, what I see with cortex M 33 or any of these where it's just secure or non-secure,

00:31:21.260 --> 00:31:25.000
<v Colin O'Flynn>you know, as a time crunched embedded developer, what do you end up doing?

00:31:25.080 --> 00:31:28.700
<v Colin O'Flynn>You put too much in secure because you're like, you know, I need access to everything.

00:31:28.700 --> 00:31:28.980
<v Chris Gammell>Okay.

00:31:29.100 --> 00:31:29.820
<v Chris Gammell>Secure code.

00:31:29.820 --> 00:31:31.200
<v Chris Gammell>There's no one really enforcing it.

00:31:31.200 --> 00:31:31.300
<v Chris Gammell>Right.

00:31:31.300 --> 00:31:34.520
<v Chris Gammell>It's just like, you're only really size limited based on how it's laid out.

00:31:34.600 --> 00:31:35.160
<v Chris Gammell>I think.

00:31:35.460 --> 00:31:35.700
<v Chris Gammell>No.

00:31:35.760 --> 00:31:35.900
<v Chris Gammell>Yeah.

00:31:35.900 --> 00:31:39.480
<v Colin O'Flynn>I don't think there's really like there it's up to you to decide what's secure and what's

00:31:39.480 --> 00:31:39.940
<v Colin O'Flynn>non-secure.

00:31:40.020 --> 00:31:40.140
<v Colin O'Flynn>Right.

00:31:40.180 --> 00:31:43.580
<v Colin O'Flynn>So you're supposed to go through and make sure the secure code is only the finest quality

00:31:43.580 --> 00:31:45.740
<v Chris Gammell>code that has no problems ever.

00:31:45.980 --> 00:31:49.720
<v Chris Gammell>There's that's how I get around it by never, ever having fine quality code for my stuff.

00:31:50.080 --> 00:31:50.640
<v Chris Gammell>Yeah, exactly.

00:31:51.180 --> 00:31:51.500
<v Chris Gammell>Perfect.

00:31:52.440 --> 00:31:52.800
<v Chris Gammell>Yeah.

00:31:52.920 --> 00:31:53.900
<v Chris Gammell>Or just do open source.

00:31:54.460 --> 00:31:55.060
<v Chris Gammell>Doesn't matter.

00:31:55.160 --> 00:31:55.300
<v Chris Gammell>Yeah.

00:31:55.300 --> 00:31:55.460
<v Chris Gammell>Right.

00:31:56.720 --> 00:31:57.080
<v Chris Gammell>Yeah.

00:31:57.080 --> 00:31:58.940
<v Chris Gammell>Someone will check it at some point on GitHub.

00:31:59.080 --> 00:31:59.340
<v Chris Gammell>I'm sure.

00:31:59.700 --> 00:31:59.940
<v Chris Gammell>All right.

00:32:00.000 --> 00:32:03.020
<v Chris Gammell>So this board looks like it's had a 40 pin header.

00:32:03.380 --> 00:32:04.540
<v Chris Gammell>It's got a display.

00:32:05.100 --> 00:32:05.200
<v Colin O'Flynn>Yeah.

00:32:05.280 --> 00:32:06.880
<v Colin O'Flynn>So it kind of, it's smushed together.

00:32:07.700 --> 00:32:09.160
<v Colin O'Flynn>You can fit a Raspberry Pi.

00:32:09.280 --> 00:32:10.680
<v Colin O'Flynn>I don't know if there's a photo of it with all of them on.

00:32:10.880 --> 00:32:13.860
<v Colin O'Flynn>A Raspberry Pi hat, an Arduino shield and micro click.

00:32:13.860 --> 00:32:14.860
<v Colin O'Flynn>Oh, and PMOD.

00:32:15.360 --> 00:32:16.620
<v Colin O'Flynn>So not simultaneously.

00:32:17.580 --> 00:32:17.680
<v Colin O'Flynn>Yeah.

00:32:17.760 --> 00:32:18.400
<v Colin O'Flynn>Quick as well.

00:32:18.820 --> 00:32:18.940
<v Colin O'Flynn>Yeah.

00:32:19.000 --> 00:32:19.120
<v Colin O'Flynn>Yeah.

00:32:19.160 --> 00:32:19.920
<v Colin O'Flynn>Two quick ports.

00:32:20.200 --> 00:32:21.120
<v Colin O'Flynn>So it was like, okay.

00:32:21.120 --> 00:32:21.640
<v Colin O'Flynn>Put all the things on there.

00:32:21.700 --> 00:32:22.260
<v Colin O'Flynn>All the things.

00:32:22.400 --> 00:32:22.860
<v Colin O'Flynn>All of it.

00:32:22.880 --> 00:32:26.560
<v Colin O'Flynn>Because what happened is like, we're looking at different sensors, right?

00:32:26.740 --> 00:32:28.060
<v Colin O'Flynn>People might want to use.

00:32:28.540 --> 00:32:32.540
<v Colin O'Flynn>And it's like, ah, well, there's a really nice Arduino shield that does whatever it was,

00:32:32.600 --> 00:32:33.640
<v Colin O'Flynn>you know, can or something.

00:32:34.140 --> 00:32:34.440
<v Chris Gammell>Yeah.

00:32:34.540 --> 00:32:38.280
<v Colin O'Flynn>That's like, oh, but there's a really nice Raspberry Pi hat that has all the RGB LEDs.

00:32:39.280 --> 00:32:39.460
<v Colin O'Flynn>Right.

00:32:39.840 --> 00:32:40.120
<v Colin O'Flynn>Right.

00:32:40.980 --> 00:32:43.520
<v Colin O'Flynn>So yeah, you can make them kind of work.

00:32:43.600 --> 00:32:46.040
<v Colin O'Flynn>Like it's a little bit funky, the headers, but.

00:32:46.280 --> 00:32:46.400
<v Chris Gammell>Yeah.

00:32:46.420 --> 00:32:47.940
<v Chris Gammell>You know, like the really tall headers or something.

00:32:47.940 --> 00:32:48.480
<v Chris Gammell>Yeah, exactly.

00:32:48.640 --> 00:32:52.700
<v Colin O'Flynn>Like the, you'll see the hat has like super high headers and the Arduino is really low

00:32:52.700 --> 00:32:53.000
<v Colin O'Flynn>headers.

00:32:53.300 --> 00:32:54.000
<v Chris Gammell>So yeah.

00:32:54.260 --> 00:32:54.520
<v Chris Gammell>Yeah.

00:32:54.720 --> 00:32:56.980
<v Colin O'Flynn>It'll fit over mostly, but yeah.

00:32:57.080 --> 00:32:58.500
<v Chris Gammell>So then it's got a FPGA.

00:32:58.600 --> 00:33:00.440
<v Chris Gammell>It's got an Arctic seven on there.

00:33:00.500 --> 00:33:04.160
<v Chris Gammell>It looks like, and that's where it's running the actual like Verilog.

00:33:04.260 --> 00:33:08.300
<v Chris Gammell>That is the Sonata, not Sonata, the Cherry, Cherry.

00:33:08.420 --> 00:33:08.620
<v Chris Gammell>Yeah.

00:33:08.620 --> 00:33:11.140
<v Chris Gammell>So while the Sonata core basically.

00:33:11.660 --> 00:33:15.000
<v Chris Gammell>So I guess Cherry is the RTOS and then the.

00:33:15.200 --> 00:33:16.440
<v Colin O'Flynn>So Cherry is the.

00:33:16.940 --> 00:33:17.700
<v Colin O'Flynn>Ah, yeah, it gets worse.

00:33:17.840 --> 00:33:19.000
<v Colin O'Flynn>So Ibex is the core.

00:33:19.200 --> 00:33:24.840
<v Colin O'Flynn>Ibex is this risk risk five core that low risk developed.

00:33:25.400 --> 00:33:26.640
<v Colin O'Flynn>So it's using a few things.

00:33:26.680 --> 00:33:27.820
<v Colin O'Flynn>It's used in open Titan as well.

00:33:27.820 --> 00:33:30.820
<v Colin O'Flynn>So Cherry is the extensions that give you the memory.

00:33:31.720 --> 00:33:34.940
<v Colin O'Flynn>I want to say memory safety, but compartmentalization, all the other features.

00:33:34.940 --> 00:33:42.020
<v Colin O'Flynn>Um, so yeah, so it runs kind of a, I mean, the idea of the board is that you, you can ignore

00:33:42.020 --> 00:33:43.200
<v Colin O'Flynn>the fact that it's an FPGA.

00:33:43.360 --> 00:33:47.560
<v Colin O'Flynn>So there's a, you know, prebuilt microcontroller that has an RTOS.

00:33:47.700 --> 00:33:50.160
<v Colin O'Flynn>And if you really want, you can go and change it.

00:33:50.160 --> 00:33:55.120
<v Colin O'Flynn>But, you know, I think what, what I've sort of seen is that, especially with students and

00:33:55.120 --> 00:34:01.140
<v Colin O'Flynn>then engineers that just want to develop something that's like step one, install 20 gigs of Havato.

00:34:01.680 --> 00:34:02.720
<v Colin O'Flynn>Nobody likes that.

00:34:03.240 --> 00:34:03.600
<v Chris Gammell>Yeah.

00:34:03.880 --> 00:34:04.000
<v Colin O'Flynn>Yeah.

00:34:04.000 --> 00:34:04.160
<v Colin O'Flynn>Right.

00:34:04.160 --> 00:34:07.880
<v Colin O'Flynn>So it's like step one, download a bit stream and just drag that on and you're done.

00:34:07.880 --> 00:34:08.520
<v Chris Gammell>Right.

00:34:08.660 --> 00:34:10.280
<v Chris Gammell>And now, now you presto change.

00:34:10.320 --> 00:34:12.060
<v Chris Gammell>Oh, you are a microcontroller of sorts.

00:34:12.460 --> 00:34:13.360
<v Chris Gammell>Um, exactly.

00:34:13.540 --> 00:34:13.620
<v Chris Gammell>Right.

00:34:13.660 --> 00:34:19.020
<v Colin O'Flynn>And you've got an RTOS and it's already got the HAL that has all the things set up for

00:34:19.020 --> 00:34:19.940
<v Chris Gammell>the bit stream.

00:34:20.220 --> 00:34:24.960
<v Chris Gammell>I have been really pleasantly surprised when like, I, I personally, like the way I operate

00:34:24.960 --> 00:34:28.840
<v Chris Gammell>with microcontrollers these days, I couldn't tell the difference between a risk five and

00:34:28.840 --> 00:34:29.900
<v Chris Gammell>a, you know, arm core.

00:34:29.900 --> 00:34:35.200
<v Chris Gammell>If I, aside from like, maybe if I'm looking in and seeing what, you know, build systems

00:34:35.200 --> 00:34:40.400
<v Chris Gammell>being called, like those usually have a different, like risk five, risk V or, you know, the, the

00:34:40.400 --> 00:34:44.160
<v Chris Gammell>arm, whatever, but like, otherwise I'm like, okay, tools are there.

00:34:44.300 --> 00:34:45.260
<v Chris Gammell>I don't have to care, you know?

00:34:45.760 --> 00:34:46.260
<v Colin O'Flynn>Yeah, no, sure.

00:34:46.320 --> 00:34:50.480
<v Colin O'Flynn>I mean, I, I recently started using some of the ESP 32 C six, I think.

00:34:50.540 --> 00:34:50.740
<v Colin O'Flynn>Yeah.

00:34:51.260 --> 00:34:52.000
<v Colin O'Flynn>Um, yeah.

00:34:52.080 --> 00:34:52.280
<v Colin O'Flynn>Yeah.

00:34:52.280 --> 00:34:53.780
<v Colin O'Flynn>It just kind of magically works.

00:34:53.820 --> 00:34:55.000
<v Colin O'Flynn>Like I would never know.

00:34:55.100 --> 00:34:56.300
<v Colin O'Flynn>I wouldn't have even known great.

00:34:56.420 --> 00:34:56.920
<v Colin O'Flynn>What right.

00:34:57.400 --> 00:34:58.040
<v Colin O'Flynn>The core was.

00:34:58.160 --> 00:34:59.300
<v Chris Gammell>So that's great.

00:34:59.300 --> 00:35:00.120
<v Chris Gammell>That's what I want.

00:35:00.200 --> 00:35:01.320
<v Chris Gammell>I want to, I want it to work.

00:35:01.380 --> 00:35:01.540
<v Chris Gammell>Right.

00:35:01.620 --> 00:35:05.800
<v Chris Gammell>I want it to be open for other people to develop on top of it so that I can get other, you know,

00:35:05.800 --> 00:35:06.600
<v Chris Gammell>fun things in the future.

00:35:06.600 --> 00:35:09.320
<v Chris Gammell>But like, yeah, I'm not building anything else.

00:35:09.880 --> 00:35:11.380
<v Chris Gammell>It's a custom like that.

00:35:11.380 --> 00:35:15.860
<v Chris Gammell>So, so like this Ibex score, it's just like one, it's an off the shelf effectively from

00:35:15.860 --> 00:35:16.380
<v Chris Gammell>low risk.

00:35:17.020 --> 00:35:17.700
<v Colin O'Flynn>Yeah, exactly.

00:35:17.780 --> 00:35:24.620
<v Colin O'Flynn>So that was an open source core that, um, I think, where did it spin out of, uh, Zurich

00:35:24.620 --> 00:35:26.060
<v Colin O'Flynn>research center.

00:35:26.280 --> 00:35:27.000
<v Colin O'Flynn>See, now I'm forgetting.

00:35:27.120 --> 00:35:28.880
<v Colin O'Flynn>This is where I should have fact checked anything.

00:35:29.300 --> 00:35:31.400
<v Colin O'Flynn>Um, it's fine.

00:35:31.500 --> 00:35:31.980
<v Colin O'Flynn>Do it in post.

00:35:32.300 --> 00:35:32.840
<v Colin O'Flynn>Do it in post.

00:35:33.300 --> 00:35:34.160
<v Colin O'Flynn>Um, yeah, yeah.

00:35:34.400 --> 00:35:36.840
<v Chris Gammell>They in fact did not do it in post.

00:35:37.340 --> 00:35:43.340
<v Colin O'Flynn>But yeah, so the, the, so this core is, um, was run by them for a bunch of stuff.

00:35:43.340 --> 00:35:48.880
<v Colin O'Flynn>And then there's a version of it that adds the chariot stuff on, um, to it.

00:35:48.960 --> 00:35:52.880
<v Colin O'Flynn>So, but yeah, I mean, it's basically like a risk five core that's been used.

00:35:52.940 --> 00:35:55.520
<v Colin O'Flynn>I think it's been used in a number of like tape outs and other things.

00:35:55.520 --> 00:35:56.640
<v Colin O'Flynn>So it's a pretty, yeah.

00:35:56.640 --> 00:35:56.840
<v Chris Gammell>Yeah.

00:35:57.040 --> 00:35:57.800
<v Colin O'Flynn>Well tested.

00:35:58.200 --> 00:36:02.400
<v Chris Gammell>So then who, who, so like students are using this to, to try out that compartmentalization,

00:36:02.400 --> 00:36:06.300
<v Chris Gammell>like try and kind of break it open or what's the, what's the thought of like, what's the

00:36:06.300 --> 00:36:08.180
<v Chris Gammell>reason for this overall board?

00:36:08.680 --> 00:36:09.040
<v Chris Gammell>Yeah.

00:36:09.080 --> 00:36:12.120
<v Colin O'Flynn>I mean, really it was just to, to get people, um, yeah.

00:36:12.200 --> 00:36:14.020
<v Colin O'Flynn>Get people being able to use the core.

00:36:14.020 --> 00:36:15.600
<v Colin O'Flynn>So the original idea was students, right.

00:36:15.600 --> 00:36:20.720
<v Colin O'Flynn>If you had, um, an undergrad class, which is why it sort of had this Arduino, right.

00:36:20.760 --> 00:36:24.140
<v Colin O'Flynn>All these shield embedded engineers, I think would be like, ah, whatever.

00:36:24.280 --> 00:36:25.140
<v Colin O'Flynn>I don't care about that.

00:36:25.260 --> 00:36:25.420
<v Chris Gammell>Yeah.

00:36:25.460 --> 00:36:26.640
<v Chris Gammell>I use them all the time, man.

00:36:27.040 --> 00:36:27.360
<v Chris Gammell>Yeah.

00:36:27.540 --> 00:36:27.780
<v Chris Gammell>Yeah.

00:36:27.780 --> 00:36:28.360
<v Colin O'Flynn>So that's the thing.

00:36:28.520 --> 00:36:29.480
<v Chris Gammell>It's what you can buy.

00:36:29.560 --> 00:36:32.000
<v Chris Gammell>So like, that's really the, you know, whatever it's quick.

00:36:32.160 --> 00:36:32.440
<v Colin O'Flynn>So yeah.

00:36:32.520 --> 00:36:33.400
<v Colin O'Flynn>But that was the idea.

00:36:33.460 --> 00:36:34.900
<v Colin O'Flynn>It's like partially students.

00:36:34.900 --> 00:36:35.840
<v Colin O'Flynn>So get it academic.

00:36:36.140 --> 00:36:38.200
<v Colin O'Flynn>Um, you know, the hope was to have lower costs.

00:36:38.380 --> 00:36:40.820
<v Colin O'Flynn>The FPGAs get expensive, unfortunately.

00:36:40.820 --> 00:36:45.900
<v Colin O'Flynn>So yeah, especially at lower volumes, like it's, you know, we're not, well, at least

00:36:45.900 --> 00:36:49.300
<v Colin O'Flynn>you can get them these days aside from the, yeah, that is an amazing thing.

00:36:49.380 --> 00:36:54.840
<v Colin O'Flynn>Like, uh, you know, we could just get them under a reasonable 14 weekly times or sometimes

00:36:54.840 --> 00:36:55.820
<v Colin O'Flynn>if they're in stock.

00:36:56.360 --> 00:36:57.080
<v Colin O'Flynn>Oh, thank you.

00:36:57.200 --> 00:36:57.700
<v Colin O'Flynn>Stuff like that.

00:36:58.260 --> 00:36:59.320
<v Colin O'Flynn>One plus a year.

00:37:00.000 --> 00:37:00.820
<v Colin O'Flynn>Hope you have enough.

00:37:00.860 --> 00:37:05.380
<v Colin O'Flynn>And I think it's like, I think our last quotes are now cancelable or changeable again.

00:37:05.700 --> 00:37:08.000
<v Colin O'Flynn>It used to be worth committing like non-cancelable.

00:37:08.380 --> 00:37:08.540
<v Colin O'Flynn>NCNR.

00:37:08.880 --> 00:37:09.180
<v Colin O'Flynn>Yeah.

00:37:09.180 --> 00:37:09.420
<v Colin O'Flynn>Yeah.

00:37:09.420 --> 00:37:11.620
<v Colin O'Flynn>Like 52 plus weekly times.

00:37:11.960 --> 00:37:13.100
<v Colin O'Flynn>How many FPS do you need?

00:37:13.120 --> 00:37:16.460
<v Chris Gammell>You will, you will pay us up front and you will like it and we may deliver them to you

00:37:16.460 --> 00:37:16.860
<v Chris Gammell>someday.

00:37:17.380 --> 00:37:17.700
<v Colin O'Flynn>Right.

00:37:17.800 --> 00:37:18.080
<v Colin O'Flynn>Yeah.

00:37:18.180 --> 00:37:19.760
<v Colin O'Flynn>That was the best part.

00:37:20.780 --> 00:37:21.320
<v Colin O'Flynn>So yeah.

00:37:21.420 --> 00:37:22.160
<v Colin O'Flynn>So it is a lot easier.

00:37:22.360 --> 00:37:24.220
<v Colin O'Flynn>That's pretty, pretty nice.

00:37:24.960 --> 00:37:25.200
<v Colin O'Flynn>Yeah.

00:37:25.320 --> 00:37:25.860
<v Chris Gammell>That's great.

00:37:26.220 --> 00:37:26.360
<v Colin O'Flynn>Okay.

00:37:26.420 --> 00:37:30.820
<v Colin O'Flynn>Are you involved in much hardware, like, uh, production these days where you're seeing

00:37:30.820 --> 00:37:31.580
<v Colin O'Flynn>that side of it or?

00:37:31.900 --> 00:37:34.180
<v Chris Gammell>Uh, no, no, not enough volume that to matter.

00:37:34.440 --> 00:37:39.100
<v Chris Gammell>Um, obviously we have other things going on with hardware production down in these.

00:37:39.100 --> 00:37:41.240
<v Chris Gammell>United States that you don't have to deal with, which is nice.

00:37:41.240 --> 00:37:41.520
<v Chris Gammell>Yeah.

00:37:41.920 --> 00:37:44.400
<v Chris Gammell>We kind of have to, but yeah, right.

00:37:44.400 --> 00:37:44.760
<v Chris Gammell>Of course.

00:37:45.120 --> 00:37:47.500
<v Chris Gammell>Like we, uh, you're still coming from DigiKey, huh?

00:37:47.860 --> 00:37:48.140
<v Colin O'Flynn>Yeah.

00:37:48.260 --> 00:37:50.920
<v Colin O'Flynn>And I think, uh, well, we use most as a distributor.

00:37:51.120 --> 00:37:52.700
<v Colin O'Flynn>So, but they've been, I mean, they're really both.

00:37:52.700 --> 00:37:57.400
<v Colin O'Flynn>Like everyone's good to deal with or doing what they can with what they've got.

00:37:57.660 --> 00:37:58.740
<v Colin O'Flynn>So, yeah, totally.

00:37:58.860 --> 00:38:02.420
<v Colin O'Flynn>Um, and so far, and I'm going to jinx myself by saying it on a podcast.

00:38:02.420 --> 00:38:02.800
<v Colin O'Flynn>Yes, you totally are.

00:38:03.060 --> 00:38:04.420
<v Colin O'Flynn>You don't have to finish the sentence.

00:38:04.540 --> 00:38:05.140
<v Colin O'Flynn>You're already jinxed.

00:38:05.240 --> 00:38:05.500
<v Colin O'Flynn>Just so you know.

00:38:05.520 --> 00:38:06.000
<v Colin O'Flynn>Already jinxed.

00:38:06.040 --> 00:38:06.500
<v Colin O'Flynn>It's already done.

00:38:06.820 --> 00:38:06.980
<v Colin O'Flynn>Right.

00:38:06.980 --> 00:38:12.180
<v Colin O'Flynn>Is that, um, the, for some reason this whole time, the like, um, customer right.

00:38:12.180 --> 00:38:13.720
<v Colin O'Flynn>Made in Canada has been exempt.

00:38:13.720 --> 00:38:18.860
<v Colin O'Flynn>Oh, because we make everything here or add enough value or whatever it is.

00:38:18.860 --> 00:38:19.080
<v Colin O'Flynn>Right.

00:38:19.120 --> 00:38:23.480
<v Colin O'Flynn>We've always ended up, we've never gone the full outsourcing route, which.

00:38:23.480 --> 00:38:24.820
<v Colin O'Flynn>Oh, yeah.

00:38:25.180 --> 00:38:30.020
<v Colin O'Flynn>Should have originally, but it's saved us because suddenly we're not.

00:38:30.600 --> 00:38:31.040
<v Colin O'Flynn>Right.

00:38:31.120 --> 00:38:31.700
<v Chris Gammell>We're actually.

00:38:32.160 --> 00:38:32.640
<v Chris Gammell>Yeah.

00:38:32.720 --> 00:38:36.360
<v Chris Gammell>Past Colin being intransigent and just wanted to build with a fun pick and place.

00:38:36.820 --> 00:38:37.300
<v Chris Gammell>Yeah, exactly.

00:38:37.300 --> 00:38:39.340
<v Chris Gammell>They saved your modern day ass, you know?

00:38:39.580 --> 00:38:40.020
<v Colin O'Flynn>Yeah.

00:38:40.200 --> 00:38:40.940
<v Colin O'Flynn>A hundred percent.

00:38:41.020 --> 00:38:45.500
<v Colin O'Flynn>Cause we were totally like, you know, and the issue was that we were going to have to,

00:38:45.580 --> 00:38:49.760
<v Colin O'Flynn>they had limited space that was bonded that, you know, they wouldn't have to charge the

00:38:49.760 --> 00:38:51.660
<v Colin O'Flynn>tarot zone, even if we're shipping to Europe.

00:38:51.660 --> 00:38:55.660
<v Colin O'Flynn>So everyone was going to have to pay because they had to pay on import.

00:38:56.400 --> 00:38:57.400
<v Colin O'Flynn>Um, oh, wow.

00:38:58.100 --> 00:39:02.980
<v Chris Gammell>So like it's, it flows through a Mauser warehouse, but it doesn't get tacked on because the end

00:39:02.980 --> 00:39:04.020
<v Chris Gammell>destination is Canada.

00:39:04.140 --> 00:39:06.100
<v Chris Gammell>Is that, is that kind of the short, short form of that?

00:39:06.600 --> 00:39:07.220
<v Colin O'Flynn>No, no.

00:39:07.240 --> 00:39:11.700
<v Colin O'Flynn>It's like in our case, because our stuff came from Canada, stuff came from Canada, like

00:39:11.700 --> 00:39:16.820
<v Colin O'Flynn>because we were source made in Canada, it was exempt under like we're exempt under, you

00:39:16.820 --> 00:39:19.120
<v Colin O'Flynn>know, the old new NAFTA customer.

00:39:19.820 --> 00:39:20.480
<v Colin O'Flynn>Oh, okay.

00:39:20.480 --> 00:39:20.520
<v Colin O'Flynn>Yeah.

00:39:20.680 --> 00:39:25.820
<v Colin O'Flynn>System because that's always been exempt forever until a little while ago.

00:39:26.160 --> 00:39:27.760
<v Chris Gammell>Um, but because of automotive.

00:39:27.920 --> 00:39:28.160
<v Chris Gammell>Sorry.

00:39:28.200 --> 00:39:29.420
<v Chris Gammell>Is that as a seller then?

00:39:29.520 --> 00:39:30.160
<v Chris Gammell>Is that what you're saying?

00:39:30.200 --> 00:39:31.280
<v Chris Gammell>Is that why it's exempt or?

00:39:31.280 --> 00:39:31.540
<v Chris Gammell>Yeah.

00:39:31.540 --> 00:39:32.020
<v Chris Gammell>As a seller.

00:39:32.220 --> 00:39:32.740
<v Chris Gammell>As a seller.

00:39:32.980 --> 00:39:33.120
<v Chris Gammell>I see.

00:39:33.220 --> 00:39:33.600
<v Chris Gammell>As a seller.

00:39:33.840 --> 00:39:34.020
<v Chris Gammell>Okay.

00:39:34.140 --> 00:39:34.560
<v Chris Gammell>Got it.

00:39:34.580 --> 00:39:34.740
<v Chris Gammell>Got it.

00:39:34.740 --> 00:39:35.020
<v Chris Gammell>Got it.

00:39:35.040 --> 00:39:35.220
<v Chris Gammell>Okay.

00:39:36.060 --> 00:39:37.220
<v Chris Gammell>So yeah.

00:39:37.260 --> 00:39:39.460
<v Colin O'Flynn>So we kind of lucked out there a little, so we'll see.

00:39:39.560 --> 00:39:42.740
<v Colin O'Flynn>I'm always worried that's going to change and, uh, we'll see.

00:39:42.740 --> 00:39:46.800
<v Chris Gammell>But you don't just because you, but just, just to close the loop on it, because you are building

00:39:46.800 --> 00:39:52.440
<v Chris Gammell>in Canada when parts flow from like a Malaysian packaging facility through Texas at Mauser

00:39:52.440 --> 00:39:57.960
<v Chris Gammell>to you or, you know, but like originally from, you know, a Chinese fab or something like that,

00:39:57.960 --> 00:40:01.480
<v Chris Gammell>for instance, do you get any, is there any tariffs in that, in that loop or no?

00:40:01.820 --> 00:40:02.560
<v Colin O'Flynn>There can be.

00:40:02.620 --> 00:40:03.320
<v Colin O'Flynn>So that's the issue.

00:40:03.320 --> 00:40:04.200
<v Chris Gammell>Is that, okay.

00:40:04.600 --> 00:40:08.160
<v Colin O'Flynn>They have some space and I don't know if they've, you know, the problem is everything

00:40:08.160 --> 00:40:08.600
<v Colin O'Flynn>changed.

00:40:08.600 --> 00:40:12.740
<v Colin O'Flynn>So like no one wants to make drastic changes and then the tariff will change.

00:40:12.920 --> 00:40:16.340
<v Chris Gammell>But, well, I'm sure, I'm sure Mauser has been real estate shopping in Canada as well.

00:40:16.340 --> 00:40:16.580
<v Chris Gammell>Right.

00:40:16.600 --> 00:40:18.460
<v Chris Gammell>Of like some, or some offloaded site.

00:40:18.700 --> 00:40:19.140
<v Chris Gammell>So that, yeah.

00:40:19.520 --> 00:40:22.520
<v Colin O'Flynn>Well, that's, I mean, so what they had is they did have bonded areas.

00:40:22.520 --> 00:40:22.720
<v Colin O'Flynn>Right.

00:40:22.720 --> 00:40:26.480
<v Colin O'Flynn>So it's like, you didn't, you don't pay tariff because it's in the special bonded zone.

00:40:26.600 --> 00:40:26.900
<v Chris Gammell>Right.

00:40:27.120 --> 00:40:27.260
<v Chris Gammell>Right.

00:40:27.260 --> 00:40:30.900
<v Chris Gammell>Like where it's only on the final, the final delivery.

00:40:31.240 --> 00:40:31.260
<v Chris Gammell>Exactly.

00:40:31.360 --> 00:40:31.520
<v Chris Gammell>Right.

00:40:31.520 --> 00:40:34.800
<v Colin O'Flynn>So, so, but it was a physical, as far as I understood when we were talking to them,

00:40:34.840 --> 00:40:36.120
<v Colin O'Flynn>it's like a physical zone.

00:40:36.120 --> 00:40:40.720
<v Colin O'Flynn>It's like this area of the warehouses, you know, chained off because I think, you know,

00:40:40.720 --> 00:40:43.040
<v Colin O'Flynn>the bonding is a generic like customs thing.

00:40:43.100 --> 00:40:44.580
<v Colin O'Flynn>It's not specific to any tariff stuff.

00:40:44.700 --> 00:40:44.800
<v Chris Gammell>Right.

00:40:44.880 --> 00:40:45.080
<v Chris Gammell>Right.

00:40:45.760 --> 00:40:49.340
<v Chris Gammell>I'm just imagining like there's like a metal detector looking thing and the person walks

00:40:49.340 --> 00:40:51.700
<v Chris Gammell>in, but it has like a little like American flag and their shirt.

00:40:51.800 --> 00:40:54.860
<v Colin O'Flynn>And then like alarms are going not allowed.

00:40:55.940 --> 00:40:56.340
<v Colin O'Flynn>Serifed.

00:40:56.460 --> 00:40:56.780
<v Colin O'Flynn>But yeah.

00:40:56.980 --> 00:40:57.320
<v Colin O'Flynn>So yeah.

00:40:57.400 --> 00:41:01.480
<v Colin O'Flynn>So we, so anyway, I think like if you had some parts would, some parts wouldn't, it

00:41:01.480 --> 00:41:05.360
<v Colin O'Flynn>would depend if, you know, they could get them in that zone where they didn't have

00:41:05.360 --> 00:41:07.580
<v Colin O'Flynn>to apply input tariffs as they came in.

00:41:08.260 --> 00:41:14.500
<v Colin O'Flynn>But some, and like, the thing is what we're seeing is, you know, like boxes and just random

00:41:14.500 --> 00:41:15.840
<v Colin O'Flynn>stuff too, as terrorists that.

00:41:16.520 --> 00:41:17.320
<v Colin O'Flynn>Oh, I see.

00:41:17.480 --> 00:41:17.620
<v Colin O'Flynn>Yeah.

00:41:17.620 --> 00:41:21.780
<v Colin O'Flynn>It's like the supplier gets boxes from the U S because our new line, because everyone

00:41:21.780 --> 00:41:22.200
<v Colin O'Flynn>does.

00:41:22.360 --> 00:41:23.160
<v Colin O'Flynn>Um, yeah.

00:41:23.220 --> 00:41:27.680
<v Colin O'Flynn>And like there's counter tariffs and like, even I was talking to a metal supplier and they

00:41:27.680 --> 00:41:31.360
<v Colin O'Flynn>were saying that, you know, Oh, there was a West, a lot of shipping would go through

00:41:31.360 --> 00:41:33.780
<v Colin O'Flynn>the U S because the rail lines were better.

00:41:34.460 --> 00:41:34.860
<v Chris Gammell>Yeah.

00:41:35.040 --> 00:41:35.260
<v Colin O'Flynn>Right.

00:41:35.400 --> 00:41:35.800
<v Colin O'Flynn>Right.

00:41:35.800 --> 00:41:35.960
<v Colin O'Flynn>Right.

00:41:36.400 --> 00:41:42.260
<v Colin O'Flynn>That was their entire distribution chain was it crosses Canada, U S U S Canada.

00:41:42.260 --> 00:41:47.060
<v Chris Gammell>Well, in case I've made my feelings not known on this, this is all stupid.

00:41:47.380 --> 00:41:49.300
<v Chris Gammell>This is so, so stupid.

00:41:49.760 --> 00:41:52.540
<v Chris Gammell>Uh, and I will continue to say as much, but, uh, sorry, you're dealing with that.

00:41:52.580 --> 00:41:52.980
<v Chris Gammell>That sucks.

00:41:53.120 --> 00:41:56.340
<v Chris Gammell>Uh, and I'm glad at least on the selling side, it's less bad.

00:41:56.360 --> 00:41:56.660
<v Chris Gammell>Right.

00:41:56.860 --> 00:41:57.100
<v Colin O'Flynn>Yeah.

00:41:57.260 --> 00:41:58.280
<v Colin O'Flynn>I mean, I feel bad.

00:41:58.540 --> 00:41:59.920
<v Colin O'Flynn>The, and the, the kind of crazy thing.

00:41:59.920 --> 00:42:04.080
<v Colin O'Flynn>And again, I'm going to like totally jinx myself spilling the beans on this is that because

00:42:04.080 --> 00:42:08.140
<v Colin O'Flynn>of that customer stuff, I mean, are the tariffs paused or no, are they back on now?

00:42:08.200 --> 00:42:10.240
<v Colin O'Flynn>What's the like Chinese tariffs?

00:42:10.240 --> 00:42:12.520
<v Chris Gammell>I literally can't pay attention.

00:42:12.640 --> 00:42:16.080
<v Chris Gammell>I mean, I I'm very blessed that I don't have to like think about it right now, but I, if

00:42:16.080 --> 00:42:20.500
<v Chris Gammell>I pay attention, I just get angry and like, and starts muttering stupid, stupid, stupid

00:42:20.500 --> 00:42:21.000
<v Chris Gammell>over again.

00:42:21.000 --> 00:42:21.820
<v Chris Gammell>So, yeah.

00:42:22.160 --> 00:42:22.440
<v Colin O'Flynn>Yeah.

00:42:22.460 --> 00:42:26.440
<v Colin O'Flynn>So the amazing thing was that you actually, what people should do if they wanted to avoid

00:42:26.440 --> 00:42:30.880
<v Colin O'Flynn>tariffs under the current system is that you should build your stuff in Canada or Mexico

00:42:30.880 --> 00:42:33.920
<v Colin O'Flynn>because you'll import from China, right?

00:42:33.980 --> 00:42:35.820
<v Colin O'Flynn>No tariffs to either of those places.

00:42:36.200 --> 00:42:36.640
<v Colin O'Flynn>Yep.

00:42:36.640 --> 00:42:41.940
<v Colin O'Flynn>You do enough transformation in Canada or Mexico that it becomes Canadian or Mexican

00:42:41.940 --> 00:42:42.880
<v Chris Gammell>origin product.

00:42:43.160 --> 00:42:43.340
<v Chris Gammell>Right.

00:42:43.400 --> 00:42:45.760
<v Colin O'Flynn>And then it's exempt under the current customer rule.

00:42:45.980 --> 00:42:47.280
<v Chris Gammell>So, um, right.

00:42:47.380 --> 00:42:51.020
<v Chris Gammell>Which seems counter of kind of, yeah, I talked about on the show a couple, uh, whenever Dave

00:42:51.020 --> 00:42:55.820
<v Chris Gammell>and I recorded last, like my instinct in all this too, is that like, okay, just assuming

00:42:55.820 --> 00:42:58.420
<v Chris Gammell>there's tariffs, I'm going to go with the cheapest stuff I can get.

00:42:58.500 --> 00:43:00.100
<v Chris Gammell>And guess where I get it from China.

00:43:00.100 --> 00:43:05.680
<v Chris Gammell>It's like, I'm going to get like, you know, the H32 VW three parts or equivalent, and then

00:43:05.680 --> 00:43:11.440
<v Chris Gammell>just like develop on that and do, you know, like put all my brain power into the terrible

00:43:11.440 --> 00:43:11.800
<v Chris Gammell>firmware.

00:43:11.980 --> 00:43:15.320
<v Chris Gammell>I write, but like, but deal with it and just pay the tariff on top of it.

00:43:15.740 --> 00:43:21.260
<v Chris Gammell>And, uh, yeah, it's, it's very counter, like, it's just like unintended consequences abound.

00:43:21.260 --> 00:43:23.060
<v Chris Gammell>That's, that's what it really comes down to.

00:43:23.200 --> 00:43:23.300
<v Chris Gammell>Yeah.

00:43:23.480 --> 00:43:23.760
<v Chris Gammell>Wow.

00:43:23.880 --> 00:43:25.560
<v Chris Gammell>I mean, sure.

00:43:25.940 --> 00:43:27.960
<v Chris Gammell>Stop market bounces were intended, right?

00:43:28.180 --> 00:43:29.420
<v Colin O'Flynn>That's another discussion.

00:43:29.600 --> 00:43:29.740
<v Colin O'Flynn>Right.

00:43:29.880 --> 00:43:30.060
<v Chris Gammell>Right.

00:43:31.080 --> 00:43:31.400
<v Chris Gammell>Yeah.

00:43:31.520 --> 00:43:31.880
<v Chris Gammell>Yeah.

00:43:32.020 --> 00:43:35.360
<v Chris Gammell>I, as I said last time too, yeah, I just feel bad for friends at DigiKey and Mauser and

00:43:35.360 --> 00:43:37.420
<v Chris Gammell>people that have to deal with all the paperwork.

00:43:38.420 --> 00:43:38.740
<v Chris Gammell>Yeah.

00:43:39.080 --> 00:43:39.280
<v Chris Gammell>Yeah.

00:43:39.920 --> 00:43:40.120
<v Colin O'Flynn>Yeah.

00:43:40.120 --> 00:43:40.480
<v Colin O'Flynn>That's the thing.

00:43:40.520 --> 00:43:43.420
<v Colin O'Flynn>I mean, when we were talking to America, they were like amazing credit to them because

00:43:43.420 --> 00:43:47.940
<v Colin O'Flynn>they were, you know, basically just like do whatever you need to do to get it.

00:43:47.940 --> 00:43:49.940
<v Colin O'Flynn>We just got to keep things moving.

00:43:50.320 --> 00:43:50.600
<v Chris Gammell>Yeah.

00:43:50.600 --> 00:43:54.600
<v Colin O'Flynn>I mean, like, you know, we were kind of worried because the agreements you have with, um,

00:43:54.600 --> 00:43:57.560
<v Colin O'Flynn>you know, we can't say specifically, but you normally have like, okay, if we have price

00:43:57.560 --> 00:43:59.100
<v Colin O'Flynn>changes, we have to give them notice.

00:43:59.100 --> 00:43:59.400
<v Colin O'Flynn>Right.

00:43:59.460 --> 00:44:01.760
<v Colin O'Flynn>Because I changed the price by the way.

00:44:02.120 --> 00:44:06.160
<v Colin O'Flynn>Um, but they were just like, if you need to change it, change it, we'll deal with it.

00:44:06.200 --> 00:44:06.860
<v Colin O'Flynn>We'll figure it out.

00:44:06.860 --> 00:44:10.960
<v Colin O'Flynn>You know, no, everyone's kind of in this together was the feel, which was at least nice.

00:44:11.760 --> 00:44:12.080
<v Chris Gammell>Yeah.

00:44:12.580 --> 00:44:15.900
<v Chris Gammell>Well, let's talk, let's switch to talking about actually making stuff on site.

00:44:15.900 --> 00:44:19.000
<v Chris Gammell>So you're, I think the last time you were on, you were kind of started the experiment,

00:44:19.000 --> 00:44:23.020
<v Chris Gammell>but I don't know how much you had scaled it up, but like you've continued to make stuff

00:44:23.020 --> 00:44:23.840
<v Chris Gammell>locally in Canada.

00:44:23.840 --> 00:44:27.620
<v Chris Gammell>Uh, you have some pick and place capabilities at your facility.

00:44:27.620 --> 00:44:27.880
<v Chris Gammell>Right.

00:44:27.960 --> 00:44:29.340
<v Chris Gammell>So how's that all been going?

00:44:29.340 --> 00:44:35.360
<v Chris Gammell>I mean, aside from the, the modern, the modern, uh, uh, luck that has been bestowed upon you

00:44:35.360 --> 00:44:35.980
<v Chris Gammell>because of it.

00:44:36.580 --> 00:44:36.680
<v Colin O'Flynn>Yeah.

00:44:36.720 --> 00:44:39.000
<v Colin O'Flynn>I mean, that, that was not intended luck.

00:44:39.180 --> 00:44:40.740
<v Colin O'Flynn>Um, uh, pretty good.

00:44:40.800 --> 00:44:42.680
<v Chris Gammell>I mean, so this is, that's why they call it luck, Colin.

00:44:42.800 --> 00:44:43.660
<v Chris Gammell>That's why they call it luck.

00:44:43.660 --> 00:44:51.200
<v Colin O'Flynn>Um, yeah, so we, I mean, we do a bit, so we use, um, some outsourcing for PCBs, especially

00:44:51.200 --> 00:44:55.740
<v Colin O'Flynn>because you know, that tends to be more efficient, but for a lot of smaller items.

00:44:55.740 --> 00:44:58.580
<v Colin O'Flynn>And we still have the thing where we have all these target boards that have different

00:44:58.580 --> 00:44:59.200
<v Colin O'Flynn>chips mounted.

00:44:59.200 --> 00:44:59.480
<v Colin O'Flynn>Right.

00:44:59.540 --> 00:45:00.740
<v Colin O'Flynn>So it can be difficult to know.

00:45:01.560 --> 00:45:03.520
<v Colin O'Flynn>Um, we did recently upgrade.

00:45:03.520 --> 00:45:09.380
<v Colin O'Flynn>So we're using some like small desktop pick and places and then got a, um, uh, slightly

00:45:09.380 --> 00:45:11.920
<v Colin O'Flynn>larger, like floor pick and place still pretty small.

00:45:11.920 --> 00:45:13.300
<v Colin O'Flynn>Um, what was it?

00:45:13.380 --> 00:45:13.500
<v Colin O'Flynn>Okay.

00:45:14.140 --> 00:45:15.280
<v Colin O'Flynn>I need to look up what it was.

00:45:15.840 --> 00:45:16.660
<v Colin O'Flynn>What are we on the brand?

00:45:17.140 --> 00:45:17.920
<v Chris Gammell>Not a, not a Neoden.

00:45:18.260 --> 00:45:18.880
<v Chris Gammell>Are you still running?

00:45:19.000 --> 00:45:21.540
<v Chris Gammell>I remember you had done like some open PNP experiments.

00:45:21.840 --> 00:45:21.960
<v Chris Gammell>Yeah.

00:45:22.300 --> 00:45:23.800
<v Colin O'Flynn>Uh, I use that a little bit.

00:45:23.820 --> 00:45:25.300
<v Colin O'Flynn>I was using the charm high ones.

00:45:25.420 --> 00:45:26.420
<v Colin O'Flynn>Um, no, it's a Neoden.

00:45:26.540 --> 00:45:26.700
<v Colin O'Flynn>Yeah.

00:45:26.700 --> 00:45:28.600
<v Colin O'Flynn>Cause Neoden, we had bought it from the U S.

00:45:29.220 --> 00:45:31.960
<v Colin O'Flynn>Um, I think it's the Neoden nine was the one.

00:45:32.420 --> 00:45:32.680
<v Chris Gammell>Okay.

00:45:32.800 --> 00:45:33.040
<v Chris Gammell>Yeah.

00:45:33.220 --> 00:45:33.420
<v Colin O'Flynn>Yeah.

00:45:33.420 --> 00:45:33.640
<v Colin O'Flynn>Okay.

00:45:33.720 --> 00:45:34.160
<v Colin O'Flynn>That's it.

00:45:34.160 --> 00:45:36.880
<v Colin O'Flynn>So, so it's a moderately, it has feeders and stuff like that.

00:45:36.880 --> 00:45:40.380
<v Colin O'Flynn>So it's a, uh, yeah, it's pretty good.

00:45:40.400 --> 00:45:41.900
<v Colin O'Flynn>I mean, it's kind of what we need.

00:45:41.920 --> 00:45:43.500
<v Colin O'Flynn>You know, we're not doing right.

00:45:43.800 --> 00:45:46.020
<v Chris Gammell>You're like high mix, high mix, like fast.

00:45:46.200 --> 00:45:49.740
<v Chris Gammell>I remember like you've saying like, Oh, well we need to be able to kind of turn on a dime,

00:45:49.880 --> 00:45:50.980
<v Chris Gammell>especially shortage times.

00:45:50.980 --> 00:45:53.620
<v Chris Gammell>I think you would maybe had talked about that a little bit, but.

00:45:54.040 --> 00:45:54.220
<v Colin O'Flynn>Yeah.

00:45:54.320 --> 00:45:54.480
<v Colin O'Flynn>Yeah.

00:45:54.480 --> 00:45:55.320
<v Colin O'Flynn>And that, that's exactly it.

00:45:55.320 --> 00:45:58.400
<v Colin O'Flynn>And this is kind of like when people are talking about the setup, right?

00:45:58.400 --> 00:46:02.720
<v Colin O'Flynn>It's a bit tricky because I think purely if you were to just look at what's the cost

00:46:02.720 --> 00:46:08.020
<v Colin O'Flynn>of building it, like, yeah, you shouldn't build yourself, but, um, you know, many times,

00:46:08.020 --> 00:46:10.920
<v Colin O'Flynn>especially with shortages, but even outside of that, right.

00:46:10.920 --> 00:46:16.600
<v Colin O'Flynn>Having the capabilities in house has saved us, um, for me, just being able to rework

00:46:16.600 --> 00:46:20.720
<v Colin O'Flynn>stuff right on the fly, because we have, you know, someone in house that can really

00:46:20.720 --> 00:46:21.320
<v Colin O'Flynn>do that.

00:46:21.320 --> 00:46:22.220
<v Colin O'Flynn>Um, yeah.

00:46:22.220 --> 00:46:22.340
<v Chris Gammell>Yeah.

00:46:22.340 --> 00:46:26.460
<v Colin O'Flynn>Which if you're outsourcing, there's no way you could afford to have someone idle hoping

00:46:26.460 --> 00:46:27.720
<v Colin O'Flynn>they need to rework a PCB.

00:46:29.120 --> 00:46:29.480
<v Chris Gammell>Right.

00:46:29.580 --> 00:46:32.520
<v Chris Gammell>That's the only happens when you like really shout at them, like you got this wrong and

00:46:32.520 --> 00:46:33.860
<v Chris Gammell>I have the documentation to prove it.

00:46:33.960 --> 00:46:35.700
<v Chris Gammell>And I won't send boards to you anymore.

00:46:36.280 --> 00:46:36.640
<v Colin O'Flynn>Yeah.

00:46:36.840 --> 00:46:37.840
<v Colin O'Flynn>Which like those.

00:46:37.960 --> 00:46:38.100
<v Colin O'Flynn>Yeah.

00:46:38.360 --> 00:46:38.680
<v Colin O'Flynn>Yeah.

00:46:38.740 --> 00:46:40.360
<v Colin O'Flynn>And I don't know what your experience is.

00:46:40.380 --> 00:46:43.060
<v Colin O'Flynn>And I, I, this is like, I don't know if you can do polls in this.

00:46:43.220 --> 00:46:44.460
<v Colin O'Flynn>This is a very low commitment.

00:46:44.460 --> 00:46:44.940
<v Colin O'Flynn>I'm sorry.

00:46:45.080 --> 00:46:49.900
<v Colin O'Flynn>I should have researched more, but you know, we've tried to do fast turn and like it has failed

00:46:49.900 --> 00:46:55.480
<v Colin O'Flynn>almost every time I've outsourced a fast turn to, and like, I'm multiple different companies.

00:46:55.960 --> 00:46:56.480
<v Chris Gammell>Like, yeah.

00:46:56.900 --> 00:47:02.140
<v Chris Gammell>You know, what style of, uh, like what's, what's the fast turn, I guess, in your mind?

00:47:02.640 --> 00:47:09.340
<v Colin O'Flynn>Uh, well, I mean like, so twice now and a few different companies, um, I've tried to get

00:47:09.340 --> 00:47:12.080
<v Colin O'Flynn>this, we were doing this, like it was pretty high spec board, right?

00:47:12.080 --> 00:47:18.640
<v Colin O'Flynn>It was like four, four mil trace space BGA and paid for like, you know, three or four day

00:47:18.640 --> 00:47:21.800
<v Colin O'Flynn>turn and I forget, you know, uh, maybe $10,000.

00:47:22.000 --> 00:47:23.200
<v Colin O'Flynn>Like it was a, yeah, that sounds right.

00:47:23.400 --> 00:47:24.240
<v Chris Gammell>Extremely expensive, right?

00:47:24.300 --> 00:47:24.900
<v Colin O'Flynn>Like, yeah.

00:47:25.420 --> 00:47:27.560
<v Chris Gammell>But I mean, that could get a lot more expensive than that though.

00:47:27.660 --> 00:47:27.840
<v Chris Gammell>Yeah.

00:47:27.900 --> 00:47:28.060
<v Chris Gammell>Yeah.

00:47:28.080 --> 00:47:28.920
<v Chris Gammell>But for me, right.

00:47:28.980 --> 00:47:32.980
<v Colin O'Flynn>You want to do some JLPT, um, but, but it took four weeks instead.

00:47:33.540 --> 00:47:34.600
<v Chris Gammell>Oh, really?

00:47:34.840 --> 00:47:35.160
<v Chris Gammell>Yeah.

00:47:35.260 --> 00:47:36.880
<v Chris Gammell>And it just kept, what was the failure mode?

00:47:36.900 --> 00:47:37.780
<v Chris Gammell>Was it files?

00:47:37.880 --> 00:47:38.540
<v Chris Gammell>Was it orders?

00:47:38.640 --> 00:47:39.380
<v Chris Gammell>Was it, I don't know.

00:47:39.380 --> 00:47:41.080
<v Colin O'Flynn>They never told me it would go to inspection.

00:47:41.080 --> 00:47:43.200
<v Colin O'Flynn>And then I was like, I don't know.

00:47:43.240 --> 00:47:47.520
<v Colin O'Flynn>I don't, cause the thing is when I ordered that at the same time, I ordered a JLPT board just

00:47:47.520 --> 00:47:50.960
<v Colin O'Flynn>cause I'm like, I'm going to, you know, keep a few insurance policy.

00:47:51.120 --> 00:47:51.280
<v Chris Gammell>Yeah.

00:47:51.380 --> 00:47:51.540
<v Chris Gammell>Yeah.

00:47:51.560 --> 00:47:53.960
<v Colin O'Flynn>Well, and I thought it was like, well, the fast train one's going to come, we're going

00:47:53.960 --> 00:47:56.740
<v Colin O'Flynn>to test it and then we'll have more JLPT ones to build up.

00:47:57.300 --> 00:47:58.280
<v Colin O'Flynn>Um, right.

00:47:58.340 --> 00:48:04.480
<v Colin O'Flynn>Just cause like then, but we, we kind of needed the quick or ideally wanted the quick early

00:48:04.480 --> 00:48:04.900
<v Colin O'Flynn>test.

00:48:05.160 --> 00:48:06.420
<v Colin O'Flynn>Um, yeah.

00:48:06.420 --> 00:48:11.040
<v Colin O'Flynn>So, but the JLPT ones came in and use that and I just kind of, we just didn't pay or disputed

00:48:11.040 --> 00:48:14.540
<v Chris Gammell>the one, the four weeks versus four days.

00:48:14.680 --> 00:48:15.160
<v Chris Gammell>That's pretty bad.

00:48:15.180 --> 00:48:15.760
<v Chris Gammell>That's pretty bad.

00:48:15.840 --> 00:48:16.040
<v Chris Gammell>Yeah.

00:48:16.040 --> 00:48:16.200
<v Chris Gammell>Yeah.

00:48:16.420 --> 00:48:16.700
<v Chris Gammell>Yeah.

00:48:16.720 --> 00:48:17.320
<v Colin O'Flynn>It was just crazy.

00:48:17.440 --> 00:48:23.360
<v Chris Gammell>And like, and it's like, you know, that was one of the worst ones, but pretty extreme

00:48:23.360 --> 00:48:23.920
<v Chris Gammell>have.

00:48:24.120 --> 00:48:25.040
<v Chris Gammell>How far away were they?

00:48:25.160 --> 00:48:27.480
<v Chris Gammell>Could you have gone to like shout at them in person?

00:48:27.600 --> 00:48:29.220
<v Chris Gammell>I mean, like, I feel like they're in the U S right.

00:48:29.280 --> 00:48:32.200
<v Colin O'Flynn>This is all, I don't think Canada had the spec we needed.

00:48:32.260 --> 00:48:36.940
<v Colin O'Flynn>So it was like, ah, uh, I don't know if I should stay because there's not that many.

00:48:37.020 --> 00:48:37.440
<v Colin O'Flynn>No, that's fine.

00:48:37.480 --> 00:48:41.320
<v Colin O'Flynn>We don't have to name a shame here this time, but maybe if it happens again, how about five

00:48:41.320 --> 00:48:41.620
<v Chris Gammell>years?

00:48:41.620 --> 00:48:42.860
<v Chris Gammell>If you've tried it again?

00:48:43.320 --> 00:48:43.640
<v Colin O'Flynn>Yeah.

00:48:43.640 --> 00:48:45.940
<v Colin O'Flynn>I feel like they, I don't know, you know, at the end it was fine.

00:48:45.940 --> 00:48:47.800
<v Colin O'Flynn>There was no charges.

00:48:47.800 --> 00:48:50.100
<v Colin O'Flynn>So things happen.

00:48:50.320 --> 00:48:50.360
<v Colin O'Flynn>Right.

00:48:50.360 --> 00:48:51.460
<v Colin O'Flynn>Like, I mean, that's the problem.

00:48:51.900 --> 00:48:55.640
<v Colin O'Flynn>It's like, was it a, you know, someone they hired that didn't work out?

00:48:55.720 --> 00:48:57.640
<v Colin O'Flynn>Was it, you know, shipping error?

00:48:57.720 --> 00:48:59.700
<v Colin O'Flynn>There was a shipping error on top of one of those.

00:48:59.940 --> 00:49:00.300
<v Chris Gammell>Yeah.

00:49:00.300 --> 00:49:00.380
<v Chris Gammell>Yeah.

00:49:00.540 --> 00:49:00.760
<v Colin O'Flynn>Right.

00:49:00.860 --> 00:49:04.740
<v Chris Gammell>So it's like, like we balk at the, the cost of, you know, 10 K whatever, but like, what

00:49:04.740 --> 00:49:06.900
<v Chris Gammell>a terrible business to be in.

00:49:06.940 --> 00:49:09.060
<v Chris Gammell>Like, it's just, it sucks.

00:49:09.300 --> 00:49:10.980
<v Chris Gammell>You know, manufacturing is tough.

00:49:11.040 --> 00:49:14.500
<v Chris Gammell>If you control every aspect that you can and it still sucks.

00:49:14.500 --> 00:49:18.140
<v Chris Gammell>And it's like, then if you're doing it for other people and there's, you know, like things

00:49:18.140 --> 00:49:18.900
<v Chris Gammell>happen, right.

00:49:18.920 --> 00:49:19.360
<v Chris Gammell>Like you said.

00:49:19.520 --> 00:49:19.640
<v Chris Gammell>So.

00:49:19.980 --> 00:49:20.120
<v Colin O'Flynn>Yeah.

00:49:20.240 --> 00:49:20.520
<v Colin O'Flynn>Yeah.

00:49:20.520 --> 00:49:20.860
<v Colin O'Flynn>And totally.

00:49:20.940 --> 00:49:25.820
<v Colin O'Flynn>And that's like, you know, another thing recently we did, um, where getting some metal

00:49:25.820 --> 00:49:27.940
<v Colin O'Flynn>parts fabbed and like first batch was awesome.

00:49:27.940 --> 00:49:31.600
<v Colin O'Flynn>And the second batch was just shipping or an, and it was a bit of a rush thing too.

00:49:31.600 --> 00:49:31.880
<v Colin O'Flynn>Right.

00:49:31.880 --> 00:49:32.820
<v Chris Gammell>And it's like, yeah, yeah.

00:49:33.040 --> 00:49:33.780
<v Colin O'Flynn>It's a shipping thing.

00:49:33.780 --> 00:49:35.200
<v Colin O'Flynn>Like, you know, that can happen.

00:49:35.400 --> 00:49:36.300
<v Colin O'Flynn>Packages disappear.

00:49:36.500 --> 00:49:37.320
<v Colin O'Flynn>Things get smashed.

00:49:38.000 --> 00:49:38.320
<v Chris Gammell>Yep.

00:49:38.620 --> 00:49:38.800
<v Chris Gammell>Yep.

00:49:39.200 --> 00:49:41.340
<v Colin O'Flynn>I'm like, yeah, I am.

00:49:41.600 --> 00:49:44.040
<v Chris Gammell>I'm currently on day six of waiting for a two day delivery from DigiKey.

00:49:44.860 --> 00:49:45.220
<v Chris Gammell>Yeah.

00:49:45.220 --> 00:49:45.680
<v Chris Gammell>It's not DigiKey.

00:49:45.780 --> 00:49:46.640
<v Chris Gammell>It's UPS's fault.

00:49:46.680 --> 00:49:47.360
<v Chris Gammell>It's not DigiKey's fault.

00:49:47.580 --> 00:49:47.780
<v Chris Gammell>Yeah.

00:49:48.060 --> 00:49:50.460
<v Chris Gammell>It's just like, okay, well, what am I going to do?

00:49:50.500 --> 00:49:53.080
<v Chris Gammell>It's like, it's in the ether until it shows up on my doorstep.

00:49:53.140 --> 00:49:55.900
<v Chris Gammell>And I was like, oh, I guess I could pretend I could talk to someone at UPS.

00:49:56.120 --> 00:49:56.460
<v Chris Gammell>Like, no.

00:49:57.320 --> 00:49:57.540
<v Chris Gammell>Yeah.

00:49:57.740 --> 00:49:59.440
<v Chris Gammell>Somewhere in rural North Carolina.

00:49:59.900 --> 00:50:00.220
<v Chris Gammell>Okay.

00:50:01.020 --> 00:50:01.380
<v Chris Gammell>Yeah.

00:50:02.720 --> 00:50:03.120
<v Chris Gammell>Yeah.

00:50:04.080 --> 00:50:04.480
<v Chris Gammell>Yeah.

00:50:05.120 --> 00:50:10.340
<v Chris Gammell>Well, so one of the key things that I remember talking about with the onsite manufacturing

00:50:10.340 --> 00:50:14.280
<v Chris Gammell>too, is that like, you kind of were hands on with it at the time and you were like really

00:50:14.280 --> 00:50:18.680
<v Chris Gammell>into it, but maybe you've handed it off for the everyday operations.

00:50:18.780 --> 00:50:19.820
<v Chris Gammell>Are you still doing?

00:50:20.400 --> 00:50:21.540
<v Colin O'Flynn>And I forget one exactly.

00:50:21.700 --> 00:50:24.940
<v Colin O'Flynn>We talked before who was with us, but we had someone now.

00:50:25.040 --> 00:50:29.380
<v Colin O'Flynn>So we had kind of a, you know, basically doing everything production.

00:50:29.380 --> 00:50:34.860
<v Colin O'Flynn>Um, so she's running, you know, combination of running stuff, but doing, you know, things

00:50:34.860 --> 00:50:40.060
<v Colin O'Flynn>need tweaking, actually doing, um, soldering stuff and anything like that as well.

00:50:40.120 --> 00:50:41.000
<v Colin O'Flynn>So that's been key.

00:50:41.000 --> 00:50:46.320
<v Colin O'Flynn>Um, again, having that in house and, and with me stepping away though, someone else is taking

00:50:46.320 --> 00:50:51.920
<v Colin O'Flynn>on a lot more of the day to day, you know, looking at designs, looking at, oh, cool.

00:50:51.920 --> 00:50:56.340
<v Colin O'Flynn>You know, overview of what's the, what's like, what are the lead times looking like?

00:50:56.420 --> 00:50:57.800
<v Colin O'Flynn>What's the current status of all?

00:50:57.960 --> 00:50:58.080
<v Colin O'Flynn>Right.

00:50:58.160 --> 00:51:01.900
<v Chris Gammell>Like sourcing and the boards and stuff like that.

00:51:02.040 --> 00:51:02.180
<v Chris Gammell>Yeah.

00:51:02.240 --> 00:51:02.480
<v Colin O'Flynn>Yeah.

00:51:02.480 --> 00:51:04.180
<v Colin O'Flynn>So that's, that's the, the, the dream.

00:51:04.320 --> 00:51:09.180
<v Colin O'Flynn>I mean, you know, we're still, to be honest, it took a while, like a shockingly long time

00:51:09.180 --> 00:51:16.020
<v Colin O'Flynn>after all the COVID things for us to like, it was only this year that we've, or maybe

00:51:16.020 --> 00:51:21.580
<v Colin O'Flynn>end of last year, I'd say we were caught up and like, really running and yeah, back, which,

00:51:21.580 --> 00:51:23.900
<v Colin O'Flynn>you know, part of that is just because we're pretty small.

00:51:24.120 --> 00:51:25.300
<v Colin O'Flynn>So sure.

00:51:25.560 --> 00:51:29.960
<v Colin O'Flynn>There's so many projects that sort of got just, you know, pushed aside because like, well,

00:51:30.020 --> 00:51:30.960
<v Colin O'Flynn>we're not going to get chips.

00:51:31.100 --> 00:51:33.900
<v Colin O'Flynn>So that's not a, let's not worry about that.

00:51:34.360 --> 00:51:34.720
<v Chris Gammell>Yeah.

00:51:34.900 --> 00:51:35.080
<v Chris Gammell>Yeah.

00:51:35.080 --> 00:51:35.260
<v Chris Gammell>Yeah.

00:51:35.260 --> 00:51:40.740
<v Chris Gammell>And then when you're selling this stuff too, I mean, like you're selling, uh, to a wide

00:51:40.740 --> 00:51:41.800
<v Chris Gammell>variety of folks as well.

00:51:41.800 --> 00:51:41.940
<v Chris Gammell>Right.

00:51:41.940 --> 00:51:47.300
<v Chris Gammell>I mean, like you're selling stuff to other test houses and students and things like that.

00:51:47.300 --> 00:51:47.420
<v Chris Gammell>Right.

00:51:47.420 --> 00:51:49.160
<v Chris Gammell>I mean, you guys have a pretty broad portfolio.

00:51:49.900 --> 00:51:50.540
<v Colin O'Flynn>Yeah, exactly.

00:51:50.540 --> 00:51:55.020
<v Colin O'Flynn>So we're selling, um, and that's always been the goal is, you know, we want to, uh, people,

00:51:55.340 --> 00:52:00.500
<v Colin O'Flynn>students doing research, um, on the, the other side, we want engineers using this professionally.

00:52:00.500 --> 00:52:05.440
<v Colin O'Flynn>You know, we want people doing this as part of evaluation security evaluations.

00:52:06.360 --> 00:52:11.340
<v Colin O'Flynn>Um, so it does range, you know, the goal has always been to have this range of, and it's

00:52:11.340 --> 00:52:12.320
<v Colin O'Flynn>gone up over time, right?

00:52:12.340 --> 00:52:17.420
<v Colin O'Flynn>But like a hundred bucks to 10,000 or something like that for products that people could buy

00:52:17.420 --> 00:52:20.320
<v Chris Gammell>chip whispers still being made chip shouters still being made.

00:52:20.440 --> 00:52:21.260
<v Chris Gammell>That's, I guess what you were talking.

00:52:21.360 --> 00:52:24.800
<v Chris Gammell>I think last time you were on here was when ship chip shouter, that's the one that's high

00:52:24.800 --> 00:52:25.480
<v Chris Gammell>voltage, right?

00:52:25.860 --> 00:52:26.160
<v Colin O'Flynn>Yeah.

00:52:26.300 --> 00:52:26.640
<v Colin O'Flynn>Oh yeah.

00:52:26.660 --> 00:52:26.880
<v Colin O'Flynn>Okay.

00:52:26.880 --> 00:52:28.060
<v Colin O'Flynn>So that was a little while ago.

00:52:28.060 --> 00:52:30.180
<v Colin O'Flynn>Um, yeah, I guess that's something good.

00:52:30.320 --> 00:52:30.780
<v Colin O'Flynn>That's yeah.

00:52:30.780 --> 00:52:33.600
<v Colin O'Flynn>I guess time, time happens.

00:52:34.060 --> 00:52:35.080
<v Colin O'Flynn>Um, it sure does.

00:52:35.220 --> 00:52:36.100
<v Colin O'Flynn>And marches on man.

00:52:36.660 --> 00:52:37.000
<v Colin O'Flynn>Yeah.

00:52:37.900 --> 00:52:38.400
<v Chris Gammell>Let's see.

00:52:38.580 --> 00:52:41.420
<v Chris Gammell>When was it in August of 21?

00:52:41.540 --> 00:52:44.180
<v Chris Gammell>So I had, I had, I had a baby in the house.

00:52:44.480 --> 00:52:45.160
<v Chris Gammell>My first baby.

00:52:45.640 --> 00:52:45.740
<v Chris Gammell>Yeah.

00:52:45.800 --> 00:52:47.760
<v Colin O'Flynn>And I saw our GSA was on the list.

00:52:47.920 --> 00:52:49.300
<v Colin O'Flynn>So we're repeating some stuff.

00:52:49.840 --> 00:52:50.200
<v Colin O'Flynn>People are.

00:52:50.280 --> 00:52:50.680
<v Chris Gammell>We are.

00:52:50.780 --> 00:52:51.180
<v Chris Gammell>All right.

00:52:51.280 --> 00:52:51.460
<v Chris Gammell>Yeah.

00:52:51.460 --> 00:52:52.380
<v Chris Gammell>You asked the same question.

00:52:52.540 --> 00:52:55.780
<v Chris Gammell>It's probably a full recap of what it was, but you know what?

00:52:55.780 --> 00:52:56.640
<v Chris Gammell>I was underslept then.

00:52:56.680 --> 00:52:58.040
<v Chris Gammell>I'm underslept now, you know, whatever.

00:52:58.060 --> 00:52:58.360
<v Chris Gammell>Perfect.

00:52:59.040 --> 00:53:02.600
<v Chris Gammell>I could, I could, you know, I could do is just read off the notes from the last time.

00:53:03.080 --> 00:53:04.880
<v Chris Gammell>Shout out to, to good notes, right?

00:53:05.220 --> 00:53:05.460
<v Chris Gammell>Yeah.

00:53:05.700 --> 00:53:06.120
<v Chris Gammell>Good job.

00:53:07.140 --> 00:53:08.060
<v Chris Gammell>That was not me.

00:53:08.940 --> 00:53:09.860
<v Chris Gammell>So that was all you.

00:53:09.940 --> 00:53:10.140
<v Chris Gammell>Yeah.

00:53:10.640 --> 00:53:10.840
<v Chris Gammell>Yeah.

00:53:11.500 --> 00:53:12.440
<v Chris Gammell>Um, yeah.

00:53:12.820 --> 00:53:13.060
<v Chris Gammell>Yeah.

00:53:13.480 --> 00:53:13.740
<v Chris Gammell>Okay.

00:53:14.040 --> 00:53:14.280
<v Chris Gammell>Okay.

00:53:14.620 --> 00:53:18.360
<v Chris Gammell>Uh, one thing that I did want to bring up on, so, you know, you had mentioned some of the

00:53:18.360 --> 00:53:19.080
<v Chris Gammell>parts you're working on.

00:53:19.340 --> 00:53:23.540
<v Chris Gammell>I, I found your old post about Apple air tag teardowns.

00:53:23.540 --> 00:53:26.380
<v Chris Gammell>I was using this as a reference point for the little board I've been building.

00:53:26.860 --> 00:53:31.020
<v Chris Gammell>And, uh, I was wondering if you'd been doing anything with this in the meantime, the air

00:53:31.020 --> 00:53:32.380
<v Chris Gammell>tag and stuff like that.

00:53:32.380 --> 00:53:33.700
<v Colin O'Flynn>No, I haven't looked at the air tag.

00:53:33.780 --> 00:53:38.500
<v Colin O'Flynn>So, I mean, a quasi unrelated, not a security site, but just for my own, you know, interest,

00:53:38.800 --> 00:53:42.800
<v Colin O'Flynn>product development had been looking a little more back at the Bluetooth tracker system and

00:53:42.800 --> 00:53:43.440
<v Colin O'Flynn>how that works.

00:53:43.440 --> 00:53:50.420
<v Colin O'Flynn>But, um, you know, I haven't looked into all of the, yeah, with the more recent trackers

00:53:50.420 --> 00:53:51.560
<v Colin O'Flynn>or any of them are using.

00:53:52.000 --> 00:53:53.760
<v Chris Gammell>I mean, I think this is the main one in the market, right?

00:53:53.780 --> 00:54:00.060
<v Chris Gammell>It's not, there's like, Oh, the new Samsung one just dropped and I'm, I'm standing in line

00:54:00.060 --> 00:54:00.500
<v Chris Gammell>to get it.

00:54:00.540 --> 00:54:05.780
<v Chris Gammell>It's like, no, no, I got a solved problem, but it is a, it is, this is a very commercially

00:54:05.780 --> 00:54:07.220
<v Chris Gammell>available and accepted solution.

00:54:07.580 --> 00:54:08.840
<v Colin O'Flynn>I think that is good.

00:54:08.900 --> 00:54:10.080
<v Colin O'Flynn>I mean, so I'm, I use Android.

00:54:10.080 --> 00:54:12.620
<v Colin O'Flynn>I don't have an iPhone and I wanted a tracker for some stuff.

00:54:12.620 --> 00:54:14.360
<v Colin O'Flynn>And like, I bought tile.

00:54:14.500 --> 00:54:15.260
<v Colin O'Flynn>It's fine.

00:54:15.400 --> 00:54:18.440
<v Colin O'Flynn>But like, yeah, there's so few people on the network.

00:54:18.540 --> 00:54:20.880
<v Colin O'Flynn>I was like testing where it's going to be discovered.

00:54:21.800 --> 00:54:25.880
<v Colin O'Flynn>I'm basically going to have to drive around if I lose it and like hope I lose enough.

00:54:25.880 --> 00:54:30.020
<v Chris Gammell>See, this is like, now you can get into drones, get a new thing, you know, have a scanner yourself,

00:54:30.160 --> 00:54:30.400
<v Chris Gammell>whatever.

00:54:30.660 --> 00:54:30.800
<v Chris Gammell>Yeah.

00:54:30.900 --> 00:54:31.040
<v Chris Gammell>Yeah.

00:54:31.040 --> 00:54:32.860
<v Chris Gammell>Network effects are like really powerful.

00:54:33.020 --> 00:54:34.480
<v Chris Gammell>It's like an important thing to have.

00:54:35.120 --> 00:54:35.600
<v Chris Gammell>So, um.

00:54:35.760 --> 00:54:36.280
<v Colin O'Flynn>No, totally.

00:54:36.440 --> 00:54:38.080
<v Colin O'Flynn>So that's, that's a big part of it.

00:54:38.400 --> 00:54:40.180
<v Chris Gammell>It's definitely some cool little hardware too.

00:54:40.460 --> 00:54:40.860
<v Colin O'Flynn>Yeah.

00:54:40.860 --> 00:54:41.420
<v Colin O'Flynn>It is amazing.

00:54:41.420 --> 00:54:42.820
<v Colin O'Flynn>They packed in.

00:54:43.200 --> 00:54:43.360
<v Chris Gammell>Yeah.

00:54:43.640 --> 00:54:43.880
<v Chris Gammell>Cool.

00:54:44.000 --> 00:54:46.140
<v Chris Gammell>So what, you had mentioned more expressive stuff as well.

00:54:46.200 --> 00:54:47.440
<v Chris Gammell>So what are you doing with expressive these days?

00:54:47.480 --> 00:54:48.640
<v Chris Gammell>I guess we were, oh, that's the other thing.

00:54:48.660 --> 00:54:51.780
<v Chris Gammell>We were talking about the, the, um, security wise, like the HCI.

00:54:52.160 --> 00:54:52.660
<v Colin O'Flynn>Oh yeah.

00:54:52.700 --> 00:54:52.880
<v Colin O'Flynn>Right.

00:54:52.960 --> 00:54:53.160
<v Colin O'Flynn>Okay.

00:54:53.160 --> 00:54:54.480
<v Chris Gammell>Quote hack that was happening.

00:54:54.540 --> 00:54:57.220
<v Chris Gammell>It wasn't really a hack that was like undocumented HCI.

00:54:57.380 --> 00:54:57.640
<v Chris Gammell>Right.

00:54:57.640 --> 00:54:58.760
<v Colin O'Flynn>Yeah, exactly.

00:54:58.760 --> 00:55:02.180
<v Colin O'Flynn>So, so, I mean, and that was like, you know, and this was, I, I guess, uh, I forget if I

00:55:02.180 --> 00:55:03.800
<v Colin O'Flynn>meant to bring it up with the cherry chariot.

00:55:03.920 --> 00:55:04.740
<v Colin O'Flynn>The interesting thing, right.

00:55:04.740 --> 00:55:08.740
<v Colin O'Flynn>About this is, um, this is a cool example for compartmentalization, right.

00:55:08.780 --> 00:55:12.840
<v Colin O'Flynn>Even if there wasn't a hack, but it's like, you know, expressive NRF, you have these sort

00:55:12.840 --> 00:55:16.640
<v Colin O'Flynn>of binary blobs that you're linking in.

00:55:16.640 --> 00:55:21.120
<v Colin O'Flynn>Um, and so if there's security threats in them and they're going to be a big target,

00:55:21.220 --> 00:55:23.120
<v Colin O'Flynn>everyone's going to want to find a attack.

00:55:23.140 --> 00:55:23.340
<v Colin O'Flynn>Yeah.

00:55:23.380 --> 00:55:23.600
<v Colin O'Flynn>Right.

00:55:23.620 --> 00:55:26.920
<v Colin O'Flynn>For one of those, those stacks, because then it applies to everything.

00:55:27.620 --> 00:55:27.980
<v Colin O'Flynn>So.

00:55:28.560 --> 00:55:28.820
<v Chris Gammell>Yeah.

00:55:28.820 --> 00:55:32.060
<v Chris Gammell>And when this came out, I remember like all the things were like, this chip isn't, you

00:55:32.060 --> 00:55:33.800
<v Chris Gammell>know, it was for the originally SP 32.

00:55:33.800 --> 00:55:36.920
<v Chris Gammell>And it was like, this chip is in millions of consumer devices.

00:55:36.920 --> 00:55:39.820
<v Chris Gammell>And now they're vulnerable to Bluetooth, you know, over the air attacks.

00:55:39.820 --> 00:55:41.740
<v Chris Gammell>And it's just like, well, maybe.

00:55:42.480 --> 00:55:42.840
<v Colin O'Flynn>Yeah.

00:55:42.920 --> 00:55:43.080
<v Chris Gammell>Yeah.

00:55:43.080 --> 00:55:43.340
<v Chris Gammell>Exactly.

00:55:43.480 --> 00:55:43.720
<v Chris Gammell>Right.

00:55:43.720 --> 00:55:46.420
<v Chris Gammell>I'm going to get out of like my, my daughter's Tony box.

00:55:46.640 --> 00:55:48.560
<v Chris Gammell>Like, oh no.

00:55:49.400 --> 00:55:51.380
<v Chris Gammell>Which I think that is not the point for security.

00:55:51.440 --> 00:55:51.760
<v Colin O'Flynn>I get it.

00:55:51.800 --> 00:55:53.480
<v Colin O'Flynn>So like, yeah, yeah.

00:55:53.520 --> 00:55:54.300
<v Colin O'Flynn>Well, and that was the thing, right.

00:55:54.340 --> 00:55:56.620
<v Colin O'Flynn>So like you compartmentalize it and that would be really good.

00:55:56.620 --> 00:55:58.500
<v Colin O'Flynn>But I mean, an interesting thing there, right.

00:55:58.500 --> 00:56:01.880
<v Colin O'Flynn>Is like one of the things when I've talked to people before and I don't have a good solution

00:56:01.880 --> 00:56:09.920
<v Colin O'Flynn>for this is for security, how many, you know, people want to have very dramatized headlines.

00:56:10.460 --> 00:56:12.320
<v Colin O'Flynn>And so it's a bit tricky.

00:56:12.320 --> 00:56:17.140
<v Colin O'Flynn>Like, you know, I was looking at some access control systems a while ago, a few years ago.

00:56:18.560 --> 00:56:21.900
<v Colin O'Flynn>And some of the stuff I probably will publish it eventually.

00:56:22.020 --> 00:56:26.340
<v Colin O'Flynn>I never did because I felt bad because it was like, I was looking at one system and it was

00:56:26.340 --> 00:56:28.140
<v Colin O'Flynn>no better or worse than any other system.

00:56:28.540 --> 00:56:28.660
<v Colin O'Flynn>Right.

00:56:28.660 --> 00:56:34.060
<v Colin O'Flynn>But if I publish it, I was guaranteed just knowing the industry that, you know, the other

00:56:34.060 --> 00:56:38.020
<v Colin O'Flynn>vendors would use that to be like, oh, look, we're so good because these other people

00:56:38.020 --> 00:56:39.080
<v Colin O'Flynn>got, you know, had this.

00:56:39.240 --> 00:56:40.840
<v Colin O'Flynn>But it's like, no, you know, that wasn't the point.

00:56:40.840 --> 00:56:42.900
<v Colin O'Flynn>The point was, was that so.

00:56:42.900 --> 00:56:48.560
<v Chris Gammell>See what you got to do is start a TikTok channel and then publish your work and then get sponsored

00:56:48.560 --> 00:56:53.000
<v Chris Gammell>by the other ones and be like, Colin approves door locks.

00:56:53.260 --> 00:56:53.860
<v Chris Gammell>44.

00:56:54.340 --> 00:56:54.620
<v Chris Gammell>Click.

00:56:54.980 --> 00:56:55.360
<v Chris Gammell>I don't know.

00:56:55.460 --> 00:56:55.620
<v Chris Gammell>Yeah.

00:56:55.840 --> 00:56:56.880
<v Chris Gammell>This is the one that the best.

00:56:57.460 --> 00:56:57.780
<v Chris Gammell>Yeah.

00:56:57.780 --> 00:57:03.220
<v Chris Gammell>When you say security system, like that's like, like door locks and things like it or, or,

00:57:03.400 --> 00:57:04.320
<v Chris Gammell>or something more.

00:57:04.520 --> 00:57:04.800
<v Colin O'Flynn>Yeah.

00:57:04.840 --> 00:57:08.180
<v Colin O'Flynn>More like access control, you know, like the card we're installing.

00:57:08.300 --> 00:57:10.220
<v Colin O'Flynn>I was installing one of their building actually.

00:57:10.320 --> 00:57:10.460
<v Colin O'Flynn>Yeah.

00:57:10.500 --> 00:57:11.280
<v Colin O'Flynn>Card reader stuff.

00:57:11.360 --> 00:57:12.460
<v Colin O'Flynn>So, you know, they use lots of it.

00:57:12.500 --> 00:57:17.720
<v Colin O'Flynn>And, you know, generally I think most people know the, well, the older ones had no security,

00:57:17.720 --> 00:57:20.520
<v Colin O'Flynn>but a lot of them would have keys in the readers and stuff like that.

00:57:20.680 --> 00:57:21.120
<v Colin O'Flynn>That are really.

00:57:21.340 --> 00:57:21.520
<v Colin O'Flynn>Wow.

00:57:21.520 --> 00:57:21.960
<v Colin O'Flynn>Yeah.

00:57:22.680 --> 00:57:23.080
<v Colin O'Flynn>Yeah.

00:57:23.220 --> 00:57:23.400
<v Colin O'Flynn>Right.

00:57:23.440 --> 00:57:25.100
<v Colin O'Flynn>And it's like kind of for the threat model.

00:57:25.220 --> 00:57:26.400
<v Colin O'Flynn>It's not amazing.

00:57:26.620 --> 00:57:28.640
<v Colin O'Flynn>It's better than the old ones.

00:57:29.580 --> 00:57:31.700
<v Colin O'Flynn>You know, I think newer ones now have secure elements.

00:57:31.900 --> 00:57:36.380
<v Colin O'Flynn>So also at the time the company was coming out with a newer reader that had a secure element

00:57:36.380 --> 00:57:37.860
<v Colin O'Flynn>that basically would solve the problem.

00:57:38.040 --> 00:57:40.440
<v Colin O'Flynn>But it got delayed because this was over COVID time.

00:57:40.760 --> 00:57:45.180
<v Colin O'Flynn>So again, I kind of felt bad because it was like, you know, they, they were trying to

00:57:45.180 --> 00:57:47.400
<v Colin O'Flynn>solve it, but I think it came out finally.

00:57:47.400 --> 00:57:50.060
<v Colin O'Flynn>So I, I kind of, you know, lost interest.

00:57:50.060 --> 00:57:54.720
<v Chris Gammell>So I, what I'm hearing here, Colin, is you've lost your killer instinct is the black hacker

00:57:54.720 --> 00:57:58.760
<v Chris Gammell>that we all think you are, you know, now you're just like a teddy bear.

00:57:59.860 --> 00:58:00.220
<v Chris Gammell>Basically.

00:58:00.640 --> 00:58:01.000
<v Chris Gammell>Yeah.

00:58:01.320 --> 00:58:03.720
<v Chris Gammell>Age, you know, like soup is the thing.

00:58:03.820 --> 00:58:07.000
<v Colin O'Flynn>I like soup, you know, hang out.

00:58:07.400 --> 00:58:07.760
<v Chris Gammell>Yeah.

00:58:07.880 --> 00:58:08.120
<v Colin O'Flynn>Yeah.

00:58:08.180 --> 00:58:08.680
<v Colin O'Flynn>I mean, I am.

00:58:08.780 --> 00:58:12.240
<v Colin O'Flynn>The one thing I've worked on, so my other COVID project too, was I started writing down

00:58:12.240 --> 00:58:16.180
<v Colin O'Flynn>a lot of this, like, you know, production, basically trying to write a book on production,

00:58:16.480 --> 00:58:17.300
<v Colin O'Flynn>small scale production.

00:58:17.560 --> 00:58:17.840
<v Chris Gammell>Yeah.

00:58:18.160 --> 00:58:19.280
<v Chris Gammell>We mentioned that before the show.

00:58:19.280 --> 00:58:19.360
<v Chris Gammell>Yeah.

00:58:19.640 --> 00:58:22.060
<v Chris Gammell>We were texting about that and I totally forgot about it.

00:58:22.080 --> 00:58:22.840
<v Chris Gammell>So good.

00:58:22.940 --> 00:58:23.900
<v Chris Gammell>I'm glad you brought that up.

00:58:24.060 --> 00:58:26.800
<v Chris Gammell>Well, it's all promotion, but yeah, I mean, there's nothing to promote because I don't

00:58:26.800 --> 00:58:28.100
<v Chris Gammell>have a, it's not out yet.

00:58:28.640 --> 00:58:32.280
<v Chris Gammell>Well, we're going to have hopefully a landing page in the show notes that people can go and

00:58:32.280 --> 00:58:34.620
<v Chris Gammell>sign up to learn more directly from Colin.

00:58:34.920 --> 00:58:35.760
<v Chris Gammell>So what are you going to write about though?

00:58:35.820 --> 00:58:38.140
<v Chris Gammell>So you're how to do, you said production?

00:58:38.540 --> 00:58:38.760
<v Colin O'Flynn>Yeah.

00:58:38.800 --> 00:58:39.080
<v Colin O'Flynn>Small scale.

00:58:39.180 --> 00:58:42.220
<v Colin O'Flynn>So, I mean, really, and, and I was a bit careful with how I scoped into game because,

00:58:42.220 --> 00:58:43.900
<v Colin O'Flynn>you know, there's a lot, a lot of people have done this.

00:58:43.960 --> 00:58:46.600
<v Colin O'Flynn>Once you start, you think you're really smart and you've done a good job, right?

00:58:46.600 --> 00:58:51.120
<v Colin O'Flynn>And then you start talking to other people and realize that like, oh man, other people

00:58:51.120 --> 00:58:53.940
<v Colin O'Flynn>are way smarter and did a better job, but it's too late.

00:58:54.320 --> 00:58:55.420
<v Colin O'Flynn>Um, I made my choices.

00:58:56.060 --> 00:58:57.180
<v Colin O'Flynn>Pen has been put to paper.

00:58:57.380 --> 00:58:57.700
<v Colin O'Flynn>Exactly.

00:58:57.700 --> 00:59:01.920
<v Colin O'Flynn>So, um, so it's really more, it's a combination of my own experiences, right?

00:59:01.940 --> 00:59:04.240
<v Colin O'Flynn>It's like a memoir on small scale production.

00:59:04.660 --> 00:59:06.920
<v Colin O'Flynn>Um, but the idea was to cover everything, right?

00:59:06.980 --> 00:59:11.400
<v Colin O'Flynn>Because, you know, we've gone to embedded world and book booths and like, okay, how do

00:59:11.400 --> 00:59:12.740
<v Colin O'Flynn>you, how do you book a booth?

00:59:12.840 --> 00:59:17.180
<v Colin O'Flynn>Ah, you've got to rent all this super expensive furniture at stupid prices.

00:59:17.260 --> 00:59:17.500
<v Colin O'Flynn>Okay.

00:59:17.500 --> 00:59:18.100
<v Colin O'Flynn>How do you do that?

00:59:18.760 --> 00:59:22.480
<v Colin O'Flynn>Um, so it was kind of like to take people from, I've got a cool idea.

00:59:22.480 --> 00:59:27.120
<v Colin O'Flynn>And then even if you don't want, even if you're not going to, you know, make a product, what

00:59:27.120 --> 00:59:29.700
<v Colin O'Flynn>does that process actually look like?

00:59:30.320 --> 00:59:30.720
<v Colin O'Flynn>Right?

00:59:30.780 --> 00:59:35.880
<v Colin O'Flynn>Like how much work is in idea to, I can ship a box to a person.

00:59:36.640 --> 00:59:37.040
<v Chris Gammell>Yeah.

00:59:37.160 --> 00:59:41.140
<v Chris Gammell>That's actually really, that should be like recorded reading before any Kickstarter or

00:59:41.140 --> 00:59:45.080
<v Chris Gammell>crowd supply came campaign kicks off and be like, well, you have to go through Colin's

00:59:45.080 --> 00:59:46.820
<v Chris Gammell>checklist and add up all your time.

00:59:47.000 --> 00:59:49.560
<v Colin O'Flynn>And it was like in doing it.

00:59:49.580 --> 00:59:50.220
<v Colin O'Flynn>Cause I didn't do that.

00:59:50.220 --> 00:59:53.360
<v Colin O'Flynn>I would just like started being like, Oh, I can make some PCBs.

00:59:53.540 --> 00:59:54.120
<v Colin O'Flynn>Let's do it.

00:59:54.160 --> 00:59:54.780
<v Colin O'Flynn>Let's do a Kickstarter.

00:59:54.940 --> 00:59:57.880
<v Colin O'Flynn>And this is back when Kickstarter didn't really enforce anything.

00:59:58.080 --> 00:59:59.120
<v Colin O'Flynn>So you could just do it.

01:00:00.000 --> 01:00:03.700
<v Colin O'Flynn>Um, yeah, it was, it's a bit of like, you know, lessons learned over that.

01:00:03.980 --> 01:00:04.260
<v Chris Gammell>Yeah.

01:00:04.640 --> 01:00:08.080
<v Chris Gammell>Do you think a lot of it will change appreciably in the next 10 years as well?

01:00:08.100 --> 01:00:11.140
<v Chris Gammell>I mean, that's the other thing is like some stuff has changed from 10 years ago, but

01:00:11.140 --> 01:00:11.660
<v Chris Gammell>you know.

01:00:11.980 --> 01:00:12.200
<v Colin O'Flynn>Yeah.

01:00:12.280 --> 01:00:14.360
<v Colin O'Flynn>Well, so the problem is I started this over COVID, right?

01:00:14.360 --> 01:00:15.580
<v Colin O'Flynn>So it's now five years old.

01:00:15.620 --> 01:00:16.760
<v Colin O'Flynn>So I've had to go through an update.

01:00:17.560 --> 01:00:17.920
<v Chris Gammell>Oh yeah.

01:00:17.920 --> 01:00:20.420
<v Colin O'Flynn>You know, it's about 70% plus written.

01:00:20.620 --> 01:00:23.300
<v Colin O'Flynn>So it's actually pretty, pretty far.

01:00:23.660 --> 01:00:28.080
<v Chris Gammell>What I hearing here is that you have a future upgrade path for your, your book as well.

01:00:28.080 --> 01:00:28.280
<v Chris Gammell>Right.

01:00:28.280 --> 01:00:29.940
<v Chris Gammell>You can have like a 26th edition.

01:00:31.380 --> 01:00:33.640
<v Chris Gammell>The needed in 14 just came out and that's a better.

01:00:34.200 --> 01:00:34.560
<v Chris Gammell>Exactly.

01:00:34.820 --> 01:00:34.960
<v Chris Gammell>Right.

01:00:35.060 --> 01:00:35.500
<v Colin O'Flynn>Well, yeah.

01:00:35.560 --> 01:00:37.660
<v Colin O'Flynn>So that's, that'll be one of the things to add, I guess.

01:00:37.660 --> 01:00:43.660
<v Colin O'Flynn>But I mean, I, I'm trying to, a lot of it's super stuff that hasn't changed in 30 years.

01:00:43.660 --> 01:00:43.940
<v Colin O'Flynn>Right.

01:00:44.000 --> 01:00:44.720
<v Colin O'Flynn>This kind of go right.

01:00:44.880 --> 01:00:48.200
<v Colin O'Flynn>Even simple, what I would call simple, but people don't think about it.

01:00:48.200 --> 01:00:48.360
<v Colin O'Flynn>Right.

01:00:48.360 --> 01:00:51.120
<v Colin O'Flynn>Like how do you deal with inventory from an accounting perspective?

01:00:51.760 --> 01:00:52.400
<v Colin O'Flynn>Oh, right.

01:00:52.820 --> 01:00:53.040
<v Chris Gammell>Right.

01:00:53.080 --> 01:00:58.640
<v Colin O'Flynn>There's a lot of stuff that you can get some really basic examples, but I've, I just struggled

01:00:58.640 --> 01:01:01.760
<v Colin O'Flynn>to find good examples specific to electronics hardware.

01:01:01.760 --> 01:01:05.900
<v Chris Gammell>And you don't think that would be regionally specific either Canada versus us in terms

01:01:05.900 --> 01:01:10.060
<v Chris Gammell>of like, I guess taxation is you have to consider it at some point.

01:01:10.160 --> 01:01:12.660
<v Chris Gammell>It's all just like, just think about this and go talk to a professional.

01:01:12.980 --> 01:01:14.060
<v Chris Gammell>That's, that's good advice.

01:01:14.160 --> 01:01:16.340
<v Chris Gammell>And that's basically it.

01:01:16.380 --> 01:01:20.340
<v Colin O'Flynn>It's like a lot of it is, you know, because a lot of it will change and, you know, I've

01:01:20.340 --> 01:01:23.660
<v Colin O'Flynn>tried to include details where I can, but a lot of it is just like, here's something

01:01:23.660 --> 01:01:25.860
<v Colin O'Flynn>again, things people don't, I don't know.

01:01:25.960 --> 01:01:29.820
<v Colin O'Flynn>I don't think they fully, some people think about sometimes you just don't have time to

01:01:29.820 --> 01:01:30.400
<v Colin O'Flynn>think about it.

01:01:30.400 --> 01:01:35.460
<v Colin O'Flynn>Um, so at least if you knew up front, you could say, okay, I need to make sure I talk

01:01:35.460 --> 01:01:40.220
<v Colin O'Flynn>to, you know, someone about how I'm going to deal with all this inventory that's coming

01:01:40.220 --> 01:01:40.540
<v Colin O'Flynn>in.

01:01:40.680 --> 01:01:44.340
<v Chris Gammell>And, you know, I remember even when to go talk to professional, like, so you're not showing

01:01:44.340 --> 01:01:45.680
<v Chris Gammell>up at like an accountant's office.

01:01:45.680 --> 01:01:47.940
<v Chris Gammell>They're like, there's not, there's nothing here for me to do, right?

01:01:47.940 --> 01:01:52.340
<v Chris Gammell>Like being like having like that kind of like path and guide is very useful actually.

01:01:52.920 --> 01:01:53.040
<v Colin O'Flynn>Yeah.

01:01:53.200 --> 01:01:53.700
<v Colin O'Flynn>Well, no, exactly.

01:01:53.780 --> 01:01:53.880
<v Colin O'Flynn>Right.

01:01:53.900 --> 01:01:57.400
<v Colin O'Flynn>And that's like, you know, I, I don't know if there's still stories of this, but I remember

01:01:57.400 --> 01:02:01.460
<v Colin O'Flynn>from early Kickstarter, there'd be a number of times when people would get the Kickstarter

01:02:01.460 --> 01:02:02.020
<v Colin O'Flynn>money.

01:02:02.460 --> 01:02:02.660
<v Colin O'Flynn>Right.

01:02:02.660 --> 01:02:05.360
<v Colin O'Flynn>But they wouldn't have the expenses yet because they didn't realize that, oh, if you just

01:02:05.360 --> 01:02:08.280
<v Colin O'Flynn>buy the parts, it doesn't actually count until you ship it.

01:02:08.280 --> 01:02:08.520
<v Colin O'Flynn>Right.

01:02:08.520 --> 01:02:10.360
<v Colin O'Flynn>Like you can't realize the cost.

01:02:11.100 --> 01:02:13.180
<v Colin O'Flynn>Um, and so they screwed themselves because.

01:02:13.980 --> 01:02:17.020
<v Colin O'Flynn>You know, they, they needed to defer it or stuff like that.

01:02:17.020 --> 01:02:18.320
<v Colin O'Flynn>Like they didn't do it in time.

01:02:18.640 --> 01:02:22.620
<v Colin O'Flynn>And they basically on paper had a super profitable business because they had all the input money

01:02:22.620 --> 01:02:24.240
<v Colin O'Flynn>and they didn't have expenses yet.

01:02:24.240 --> 01:02:25.080
<v Colin O'Flynn>Yeah.

01:02:25.080 --> 01:02:25.380
<v Colin O'Flynn>Right.

01:02:25.640 --> 01:02:25.960
<v Colin O'Flynn>Right.

01:02:26.020 --> 01:02:30.840
<v Colin O'Flynn>Like they, they hadn't set it up well enough that they could amend their return or stuff

01:02:30.840 --> 01:02:31.260
<v Colin O'Flynn>like that.

01:02:31.500 --> 01:02:31.860
<v Chris Gammell>Yeah.

01:02:31.860 --> 01:02:37.820
<v Chris Gammell>So like Kickstarter money gets delivered December 31st and the new, yeah, all the parts are purchased

01:02:37.820 --> 01:02:39.220
<v Chris Gammell>on January 1st or something like that.

01:02:39.260 --> 01:02:39.560
<v Colin O'Flynn>Yeah.

01:02:39.560 --> 01:02:40.240
<v Colin O'Flynn>Stuff like that.

01:02:40.240 --> 01:02:40.420
<v Colin O'Flynn>Right.

01:02:40.660 --> 01:02:44.620
<v Colin O'Flynn>And even in that case, like in, you know, depending on what you're using, you may not even be able

01:02:44.620 --> 01:02:48.560
<v Colin O'Flynn>to call it an expense until you ship stuff out because it's, oh, it's just an asset.

01:02:48.700 --> 01:02:48.840
<v Colin O'Flynn>Right.

01:02:48.860 --> 01:02:49.860
<v Colin O'Flynn>Like now you have an asset.

01:02:50.100 --> 01:02:50.400
<v Chris Gammell>Yeah.

01:02:50.400 --> 01:02:50.740
<v Chris Gammell>Right.

01:02:50.820 --> 01:02:51.020
<v Chris Gammell>Right.

01:02:51.020 --> 01:02:51.220
<v Chris Gammell>Right.

01:02:51.220 --> 01:02:52.760
<v Chris Gammell>Like we have a hundred.

01:02:52.980 --> 01:02:53.380
<v Chris Gammell>It's inventory.

01:02:53.380 --> 01:02:55.420
<v Chris Gammell>You have to get taxed on that, at least in the U S I mean.

01:02:55.740 --> 01:02:55.860
<v Chris Gammell>Yeah.

01:02:55.980 --> 01:02:56.480
<v Chris Gammell>Well, yeah, exactly.

01:02:56.520 --> 01:02:56.840
<v Colin O'Flynn>That's it.

01:02:56.860 --> 01:02:57.000
<v Colin O'Flynn>Right.

01:02:57.020 --> 01:03:01.320
<v Colin O'Flynn>So like you had a hundred K of profit and then you spent a hundred K so you have zero cash,

01:03:01.320 --> 01:03:03.640
<v Colin O'Flynn>but on the accounting perspective, right.

01:03:03.660 --> 01:03:07.740
<v Colin O'Flynn>You had all profit and asset like, right.

01:03:07.800 --> 01:03:08.440
<v Colin O'Flynn>You had no expense.

01:03:08.500 --> 01:03:08.880
<v Colin O'Flynn>You have zero.

01:03:09.420 --> 01:03:09.540
<v Colin O'Flynn>Yeah.

01:03:10.380 --> 01:03:11.860
<v Colin O'Flynn>So that, yeah, that's stuff like that.

01:03:11.880 --> 01:03:12.020
<v Colin O'Flynn>Right.

01:03:12.060 --> 01:03:17.980
<v Colin O'Flynn>That like, you know, I've, when I've been talking to people starting this stuff, it's not crazy,

01:03:18.560 --> 01:03:23.300
<v Colin O'Flynn>but I also think it's, uh, it's easy to miss if you're, you know, just coming out

01:03:23.300 --> 01:03:25.720
<v Colin O'Flynn>of school, like was sort of a thought process.

01:03:26.440 --> 01:03:26.800
<v Colin O'Flynn>Yeah.

01:03:26.840 --> 01:03:27.520
<v Colin O'Flynn>I've got a good idea.

01:03:27.700 --> 01:03:31.340
<v Chris Gammell>That is one thing I wanted to ask you about too, since you are a teacher and you're kind

01:03:31.340 --> 01:03:33.540
<v Chris Gammell>of, you're dealing with the youths.

01:03:33.540 --> 01:03:39.780
<v Chris Gammell>Um, and we were talking about like their use of AI and kind of, uh, generally just students,

01:03:39.780 --> 01:03:44.820
<v Chris Gammell>you know, kids these days, uh, as a general, what, what are, what trends are you seeing

01:03:44.820 --> 01:03:45.060
<v Chris Gammell>there?

01:03:45.500 --> 01:03:45.640
<v Colin O'Flynn>Yeah.

01:03:45.720 --> 01:03:45.920
<v Colin O'Flynn>Yeah.

01:03:45.920 --> 01:03:46.420
<v Colin O'Flynn>I mean, it's interesting.

01:03:46.520 --> 01:03:49.840
<v Colin O'Flynn>There's tons of chat GBD, like, um, and actually what was terrible.

01:03:49.840 --> 01:03:54.400
<v Colin O'Flynn>I don't know how many, you know, how much readers are plugged into it, but they, they

01:03:54.400 --> 01:03:59.700
<v Colin O'Flynn>released, um, a few months ago, a deep, what was called deep research mode research.

01:03:59.780 --> 01:03:59.960
<v Colin O'Flynn>Yeah.

01:04:00.080 --> 01:04:00.340
<v Colin O'Flynn>Yeah.

01:04:00.340 --> 01:04:03.860
<v Colin O'Flynn>Where it would like give you citations and everything, which is terrible timing.

01:04:03.860 --> 01:04:07.980
<v Colin O'Flynn>Cause in the class, I was just teaching a class, um, in the winter, it's like January

01:04:07.980 --> 01:04:09.780
<v Colin O'Flynn>to April term.

01:04:10.200 --> 01:04:12.820
<v Colin O'Flynn>And I had a big research paper is one of the assignments.

01:04:12.820 --> 01:04:15.940
<v Colin O'Flynn>Cause I was like, oh, that's something that chat GBD doesn't do a great job on.

01:04:15.940 --> 01:04:20.900
<v Colin O'Flynn>You know, it's like, boom, they released the exact thing that, um, but yeah.

01:04:20.960 --> 01:04:26.500
<v Colin O'Flynn>So the, the thing is, I mean, it's kind of an existential question of, are we asking students

01:04:26.500 --> 01:04:30.620
<v Colin O'Flynn>to do things that, you know, will be automated?

01:04:30.620 --> 01:04:32.520
<v Colin O'Flynn>So is it a stupid thing to ask them to do?

01:04:32.980 --> 01:04:33.380
<v Colin O'Flynn>Right.

01:04:33.480 --> 01:04:33.800
<v Colin O'Flynn>Got it.

01:04:34.140 --> 01:04:38.940
<v Colin O'Flynn>Am I asking them to do manual multiplication when everyone's going to use a calculator?

01:04:39.480 --> 01:04:39.880
<v Colin O'Flynn>Right.

01:04:40.100 --> 01:04:40.480
<v Colin O'Flynn>Right.

01:04:41.040 --> 01:04:44.900
<v Colin O'Flynn>You know, so, and, and it was kind of how I'm trying to approach it and in ways of making

01:04:44.900 --> 01:04:45.740
<v Colin O'Flynn>new questions.

01:04:45.820 --> 01:04:48.500
<v Colin O'Flynn>I mean, one of the assignments actually have them do so that in this course they were,

01:04:49.000 --> 01:04:50.220
<v Colin O'Flynn>um, I was covering for someone.

01:04:50.300 --> 01:04:54.060
<v Colin O'Flynn>So it's like a, they're using Ross to, to make a robot do various tasks.

01:04:54.260 --> 01:04:55.480
<v Chris Gammell>Oh, yeah.

01:04:55.920 --> 01:04:56.140
<v Colin O'Flynn>Yeah.

01:04:56.140 --> 01:04:57.400
<v Colin O'Flynn>So, I mean, I learned a lot doing it.

01:04:57.760 --> 01:04:58.840
<v Colin O'Flynn>Had to pretend I knew.

01:04:59.840 --> 01:05:05.360
<v Colin O'Flynn>Um, but yeah, so I have them ask chat TBD to make Ross to code, to drive the robot in a

01:05:05.360 --> 01:05:08.160
<v Colin O'Flynn>square, which it does an awesome job at, right?

01:05:08.180 --> 01:05:10.220
<v Colin O'Flynn>Because there's a million tutorials doing that.

01:05:10.220 --> 01:05:15.540
<v Colin O'Flynn>Um, I then say, okay, you know, use the same prompt and ask it to drive in like a maple

01:05:15.540 --> 01:05:15.900
<v Colin O'Flynn>leaf.

01:05:16.120 --> 01:05:18.180
<v Colin O'Flynn>And it just, so Canadian.

01:05:18.380 --> 01:05:18.740
<v Colin O'Flynn>Come on.

01:05:18.860 --> 01:05:19.960
<v Colin O'Flynn>I had to tie that in.

01:05:20.040 --> 01:05:20.160
<v Colin O'Flynn>Sorry.

01:05:20.240 --> 01:05:23.660
<v Colin O'Flynn>It had to be something, something that the witty tutorials, but it exploit, like it does

01:05:23.660 --> 01:05:24.440
<v Colin O'Flynn>not work.

01:05:24.480 --> 01:05:25.880
<v Colin O'Flynn>It doesn't even come close to working.

01:05:26.160 --> 01:05:27.000
<v Colin O'Flynn>Oh, yeah.

01:05:27.320 --> 01:05:27.560
<v Colin O'Flynn>Right.

01:05:27.640 --> 01:05:30.180
<v Colin O'Flynn>And the point is, it's like, it is a good tool.

01:05:30.700 --> 01:05:33.960
<v Colin O'Flynn>You could ask chat TBD, you know, get me the vector of a maple leaf.

01:05:33.960 --> 01:05:35.000
<v Colin O'Flynn>Like you could get it there.

01:05:35.100 --> 01:05:35.200
<v Colin O'Flynn>Sure.

01:05:35.200 --> 01:05:35.360
<v Colin O'Flynn>Right.

01:05:35.360 --> 01:05:39.360
<v Colin O'Flynn>You could say, make a function that drives the robot to an X, Y location and it'll make

01:05:39.360 --> 01:05:40.120
<v Colin O'Flynn>that and that'll work.

01:05:40.140 --> 01:05:42.020
<v Colin O'Flynn>Cause there's lots of examples, right?

01:05:42.020 --> 01:05:42.940
<v Colin O'Flynn>Like you'd tie it together.

01:05:42.940 --> 01:05:46.840
<v Colin O'Flynn>So it's kind of to prove, cause like what I was seeing as students would just like put

01:05:46.840 --> 01:05:49.080
<v Colin O'Flynn>everything in and it either worked or didn't work.

01:05:49.180 --> 01:05:49.280
<v Colin O'Flynn>Right.

01:05:49.380 --> 01:05:51.400
<v Chris Gammell>And then they try it again, try it again, try it again, try it again.

01:05:51.400 --> 01:05:51.620
<v Chris Gammell>Yeah.

01:05:51.620 --> 01:05:55.520
<v Colin O'Flynn>Like there is no, it was teaching them how to be able to make small testable things.

01:05:56.020 --> 01:05:56.380
<v Colin O'Flynn>Right.

01:05:56.380 --> 01:05:59.880
<v Colin O'Flynn>Like it is a tool that's useful, but make sure it's testable.

01:05:59.880 --> 01:06:01.760
<v Colin O'Flynn>You're responsible for what it's doing.

01:06:02.440 --> 01:06:03.440
<v Colin O'Flynn>Um, yeah.

01:06:03.440 --> 01:06:03.480
<v Colin O'Flynn>Yeah.

01:06:03.960 --> 01:06:07.860
<v Chris Gammell>So, yeah, I saw an article that was about like, basically the, the, the skills of the

01:06:07.860 --> 01:06:11.820
<v Chris Gammell>future are one of the skills of the past, which is just troubleshooting, like being able

01:06:11.820 --> 01:06:15.580
<v Chris Gammell>to troubleshoot, you know, there is going to be there, you know, you have to assume there's

01:06:15.580 --> 01:06:20.060
<v Chris Gammell>some AI slop in the mix somewhere, fine, whatever, but you still need to be able to figure out

01:06:20.060 --> 01:06:24.300
<v Chris Gammell>what's going right, what's going wrong, how to test your assumptions, like just getting

01:06:24.300 --> 01:06:24.880
<v Chris Gammell>better at that.

01:06:24.880 --> 01:06:29.760
<v Chris Gammell>If you, if you were only teaching that like troubleshooting, I feel like that's, that's all

01:06:29.760 --> 01:06:32.480
<v Chris Gammell>every course would have been like every course would have been useful for that sort of

01:06:32.480 --> 01:06:33.480
<v Chris Gammell>thing, you know?

01:06:33.480 --> 01:06:33.480
<v Speaker ?>

01:06:33.480 --> 01:06:34.480
<v Colin O'Flynn>Yeah, absolutely.

01:06:34.480 --> 01:06:36.900
<v Colin O'Flynn>And that's, I mean, that was an interesting chat GPT usage.

01:06:36.900 --> 01:06:42.240
<v Colin O'Flynn>I saw actually, you know, we had some sample code that we're giving them that, you know,

01:06:42.240 --> 01:06:43.680
<v Colin O'Flynn>another prof wrote.

01:06:43.680 --> 01:06:45.860
<v Colin O'Flynn>So like, it's not, it doesn't exist per se, right?

01:06:45.860 --> 01:06:51.200
<v Colin O'Flynn>Like he made it really specific for whatever interface and there'd be errors and like the

01:06:51.200 --> 01:06:55.660
<v Colin O'Flynn>errors would come up and they were putting those errors into chat GPT, which of course

01:06:55.660 --> 01:07:00.320
<v Colin O'Flynn>was not going to work because like they were just, they were not to know their messages.

01:07:00.320 --> 01:07:03.320
<v Colin O'Flynn>Like it was just, this is an error message for, right.

01:07:03.360 --> 01:07:05.420
<v Chris Gammell>But that was still the first thing was like, oh, do that.

01:07:05.840 --> 01:07:05.940
<v Chris Gammell>Yeah.

01:07:06.020 --> 01:07:09.700
<v Chris Gammell>Does that mean there's more like when there's like a educate or sorry, like a testing kind

01:07:09.700 --> 01:07:13.580
<v Chris Gammell>of paradigm then as well, you know, education always has like some kind of certification,

01:07:13.580 --> 01:07:17.460
<v Chris Gammell>testing, whatever, does it mean that everything has to be kind of in person?

01:07:17.580 --> 01:07:22.080
<v Chris Gammell>Even if there is like an allowed AI agent that's maybe with them in it, like, but it's just

01:07:22.080 --> 01:07:25.700
<v Chris Gammell>like onsite you're, you're watching the, the assessment happen.

01:07:26.340 --> 01:07:26.440
<v Colin O'Flynn>Yeah.

01:07:26.580 --> 01:07:26.780
<v Colin O'Flynn>Yeah.

01:07:26.780 --> 01:07:27.300
<v Colin O'Flynn>I don't know.

01:07:27.380 --> 01:07:28.460
<v Colin O'Flynn>I mean, that's, it's a good question.

01:07:28.760 --> 01:07:29.980
<v Colin O'Flynn>That's stuff like that.

01:07:30.060 --> 01:07:34.220
<v Colin O'Flynn>I don't know if at least in our university, I haven't seen good answers to it yet.

01:07:34.620 --> 01:07:35.600
<v Colin O'Flynn>How are they going to deal with that?

01:07:35.660 --> 01:07:36.980
<v Colin O'Flynn>I mean, I was using a lot.

01:07:37.060 --> 01:07:40.340
<v Colin O'Flynn>So for this class, they just had a final project.

01:07:40.340 --> 01:07:44.560
<v Colin O'Flynn>So like, you know, they had to make a functioning robot that did various things.

01:07:45.140 --> 01:07:48.620
<v Colin O'Flynn>If they used AI to generate code, I didn't really check.

01:07:48.700 --> 01:07:52.960
<v Colin O'Flynn>I kind of, to some degree allowed it because I said, if you can get it to work, right.

01:07:53.080 --> 01:07:56.260
<v Colin O'Flynn>Like that's, that's, that's up to you, how you generate the code.

01:07:56.440 --> 01:07:56.960
<v Colin O'Flynn>But yeah.

01:07:57.080 --> 01:08:01.380
<v Colin O'Flynn>So for pure, I mean, for pure tests, like the thing I think we're getting at rate is like

01:08:01.380 --> 01:08:09.080
<v Colin O'Flynn>asking students to rate a research paper on some basic topic is like a non, basically

01:08:09.080 --> 01:08:10.540
<v Colin O'Flynn>that there's no way you can do that.

01:08:10.640 --> 01:08:12.280
<v Colin O'Flynn>You're saying, yeah, they just don't do that anymore.

01:08:12.560 --> 01:08:12.680
<v Colin O'Flynn>Yeah.

01:08:12.760 --> 01:08:16.620
<v Colin O'Flynn>Like, well, people still do, but I just think you're, you're totally like some students are

01:08:16.620 --> 01:08:23.060
<v Colin O'Flynn>honest and not using chat GPT that like, they're probably at a disadvantage because like, and

01:08:23.060 --> 01:08:27.420
<v Colin O'Flynn>it's so hard, you know, they're trying to use tools to detect AI, but I don't, I don't

01:08:27.420 --> 01:08:28.780
<v Colin O'Flynn>really trust them that much.

01:08:28.860 --> 01:08:30.260
<v Colin O'Flynn>My reliability is low.

01:08:30.260 --> 01:08:32.560
<v Colin O'Flynn>So like you're accusing students that haven't right.

01:08:32.620 --> 01:08:35.140
<v Colin O'Flynn>I'm using it and letting others through.

01:08:35.380 --> 01:08:40.300
<v Chris Gammell>So I just feel like the, the problem is always the schools are there to test and certify whatever.

01:08:40.300 --> 01:08:40.640
<v Chris Gammell>Right.

01:08:40.660 --> 01:08:45.320
<v Chris Gammell>And like, that's a huge part of the educational process to like say that you're okay to move

01:08:45.320 --> 01:08:45.460
<v Chris Gammell>on.

01:08:45.460 --> 01:08:49.680
<v Chris Gammell>But like the actual focus on learning, like that is, that's always the tough thing of like,

01:08:49.740 --> 01:08:50.760
<v Chris Gammell>they're often decoupled.

01:08:50.840 --> 01:08:52.280
<v Chris Gammell>They've been decoupled for years and years anyways.

01:08:52.280 --> 01:08:52.660
<v Chris Gammell>Right.

01:08:52.860 --> 01:08:56.480
<v Chris Gammell>Even say for all of education of just like, just because I test on something doesn't mean

01:08:56.480 --> 01:09:01.720
<v Chris Gammell>of actually learned it, if anything, like if it just blew up the testing process and it

01:09:01.720 --> 01:09:05.440
<v Chris Gammell>was just focusing on the learning process, that would ultimately be a good thing.

01:09:05.520 --> 01:09:08.500
<v Chris Gammell>But I don't know if it ever would go that direction, you know?

01:09:09.020 --> 01:09:09.280
<v Colin O'Flynn>Yeah.

01:09:09.460 --> 01:09:10.280
<v Colin O'Flynn>No, you're totally right.

01:09:10.340 --> 01:09:12.940
<v Colin O'Flynn>Like that's, I mean, there's been some discussions.

01:09:13.120 --> 01:09:17.580
<v Colin O'Flynn>I know other universities have looked at this about like, you know, what is the value of marks

01:09:17.580 --> 01:09:18.080
<v Colin O'Flynn>per se?

01:09:18.280 --> 01:09:19.100
<v Colin O'Flynn>Because yeah.

01:09:19.420 --> 01:09:19.600
<v Colin O'Flynn>Yeah.

01:09:19.680 --> 01:09:25.380
<v Colin O'Flynn>You know, as soon as you have marks and testing and then becoming like, did you get 87% or 85%?

01:09:26.000 --> 01:09:26.120
<v Colin O'Flynn>Right.

01:09:26.200 --> 01:09:26.460
<v Colin O'Flynn>Right.

01:09:26.480 --> 01:09:27.640
<v Colin O'Flynn>Is a thing.

01:09:27.640 --> 01:09:29.540
<v Colin O'Flynn>And then it's, yeah.

01:09:29.580 --> 01:09:34.300
<v Colin O'Flynn>Which then really is like, okay, what's the, what's the rubric for the assignment?

01:09:34.300 --> 01:09:36.000
<v Colin O'Flynn>Cause I have to exactly hit that.

01:09:36.140 --> 01:09:36.280
<v Colin O'Flynn>Right.

01:09:36.360 --> 01:09:36.540
<v Colin O'Flynn>Right.

01:09:36.540 --> 01:09:40.280
<v Colin O'Flynn>Because I need to maximize my marks because that's going to in the future impact a scholarship

01:09:40.280 --> 01:09:41.060
<v Colin O'Flynn>or something.

01:09:41.860 --> 01:09:43.040
<v Chris Gammell>And yeah, yeah, yeah.

01:09:43.140 --> 01:09:47.040
<v Chris Gammell>Ties back to money and you know, like how people are there and what they're, like it's

01:09:47.040 --> 01:09:47.780
<v Chris Gammell>a service industry.

01:09:47.780 --> 01:09:52.380
<v Chris Gammell>You're here to like administer a service and, you know, certify based on this so that, you

01:09:52.380 --> 01:09:53.360
<v Chris Gammell>know, all that stuff.

01:09:53.400 --> 01:09:53.500
<v Chris Gammell>Yeah.

01:09:53.500 --> 01:09:56.400
<v Chris Gammell>So it really, it does ripple far out, but someone should.

01:09:56.400 --> 01:09:58.420
<v Chris Gammell>I would still be saying, are the kids actually learning anything?

01:09:58.420 --> 01:09:59.560
<v Chris Gammell>Like, am I learning anything?

01:09:59.640 --> 01:10:02.500
<v Chris Gammell>Like, do I learn anything when I, you know, use an AI engine?

01:10:02.600 --> 01:10:06.920
<v Chris Gammell>Like, yeah, maybe, maybe glancing, but like, do I need depth of knowledge for situations

01:10:06.920 --> 01:10:09.480
<v Chris Gammell>where I'm not going to have access to that stuff?

01:10:09.480 --> 01:10:10.300
<v Chris Gammell>Like, I don't know.

01:10:10.300 --> 01:10:15.600
<v Chris Gammell>It just, the changing nature of like the, the second brain kind of the offloaded to an

01:10:15.600 --> 01:10:16.440
<v Chris Gammell>AI sort of thing.

01:10:16.440 --> 01:10:16.980
<v Chris Gammell>Like, I don't know.

01:10:16.980 --> 01:10:17.120
<v Chris Gammell>Yeah.

01:10:17.120 --> 01:10:17.820
<v Colin O'Flynn>Not very much.

01:10:17.900 --> 01:10:19.320
<v Colin O'Flynn>I mean, my, my, you know, excellent.

01:10:19.400 --> 01:10:20.760
<v Colin O'Flynn>This is getting very existential now.

01:10:20.800 --> 01:10:21.120
<v Colin O'Flynn>I'm sorry.

01:10:21.260 --> 01:10:21.420
<v Colin O'Flynn>Sure.

01:10:21.600 --> 01:10:21.760
<v Colin O'Flynn>Yeah.

01:10:21.840 --> 01:10:22.900
<v Colin O'Flynn>But it's like, right.

01:10:22.980 --> 01:10:27.760
<v Colin O'Flynn>Are we getting to a stage where you'll have like a, you know, knowledge will become,

01:10:27.760 --> 01:10:32.520
<v Colin O'Flynn>we'll kind of level off in a way because, you know, if everyone's just asking chat

01:10:32.520 --> 01:10:37.700
<v Colin O'Flynn>GPT, like who's posting blog posts and who's posting, you know, stack overflow questions

01:10:37.700 --> 01:10:38.500
<v Colin O'Flynn>and stuff like that.

01:10:38.500 --> 01:10:38.900
<v Colin O'Flynn>Right.

01:10:38.940 --> 01:10:42.180
<v Colin O'Flynn>So it's like, if AI doesn't know the answer, then.

01:10:42.420 --> 01:10:42.540
<v Chris Gammell>Yeah.

01:10:42.840 --> 01:10:45.720
<v Chris Gammell>There's no novel, novel discovery and like publishing about it.

01:10:45.740 --> 01:10:46.820
<v Chris Gammell>Like, what does it ingest?

01:10:47.280 --> 01:10:47.840
<v Colin O'Flynn>Yeah, exactly.

01:10:47.940 --> 01:10:48.080
<v Colin O'Flynn>Right.

01:10:48.100 --> 01:10:51.620
<v Colin O'Flynn>Like where, where does it, if ever, if no one's, if people aren't publishing as many

01:10:51.620 --> 01:10:55.300
<v Colin O'Flynn>things, cause they're not, you know, just writing stack overflow questions and stuff

01:10:55.300 --> 01:10:55.660
<v Colin O'Flynn>like that.

01:10:55.660 --> 01:10:55.980
<v Chris Gammell>Yeah.

01:10:56.060 --> 01:10:56.300
<v Chris Gammell>Right.

01:10:56.380 --> 01:10:59.680
<v Colin O'Flynn>Like you're just asking chat GPT and that doesn't become public.

01:10:59.680 --> 01:11:03.280
<v Chris Gammell>Like, um, maybe it trains on it.

01:11:03.680 --> 01:11:07.240
<v Chris Gammell>That would be the most old man thing for us to do is like soup.

01:11:07.500 --> 01:11:09.800
<v Chris Gammell>And then like, Hey, do you remember the good times?

01:11:10.640 --> 01:11:11.640
<v Colin O'Flynn>Stack overflow?

01:11:12.040 --> 01:11:12.200
<v Colin O'Flynn>Yeah.

01:11:12.640 --> 01:11:13.460
<v Colin O'Flynn>Back in the day.

01:11:13.800 --> 01:11:14.600
<v Colin O'Flynn>Oh, good times.

01:11:14.600 --> 01:11:16.740
<v Colin O'Flynn>Slash people yelling at you.

01:11:17.380 --> 01:11:21.940
<v Colin O'Flynn>Back in my day, people were grumpy about things and AI's learned from it.

01:11:22.400 --> 01:11:22.700
<v Colin O'Flynn>Yeah.

01:11:23.740 --> 01:11:24.100
<v Colin O'Flynn>Yeah.

01:11:24.240 --> 01:11:24.540
<v Colin O'Flynn>Yeah.

01:11:24.540 --> 01:11:25.060
<v Colin O'Flynn>So I don't know.

01:11:25.120 --> 01:11:26.620
<v Colin O'Flynn>I mean, it is a bit for research, right?

01:11:26.620 --> 01:11:29.460
<v Colin O'Flynn>That's where like AI, you know, and again, that's what I'm trying to show.

01:11:29.460 --> 01:11:34.300
<v Colin O'Flynn>Like, Hey, it's good for basic stuff, but, and it could be good for more complex things,

01:11:34.420 --> 01:11:40.100
<v Colin O'Flynn>but you're also, you know, if you want to push the boundary, then you can't ask it to

01:11:40.100 --> 01:11:40.700
<v Colin O'Flynn>push the boundary.

01:11:40.700 --> 01:11:44.980
<v Colin O'Flynn>Cause it doesn't know, like, you know, it's, you know, and I think just removing the magic

01:11:44.980 --> 01:11:50.020
<v Colin O'Flynn>is almost what they need because for some students, I feel like they do write really

01:11:50.020 --> 01:11:50.240
<v Colin O'Flynn>big.

01:11:50.320 --> 01:11:51.240
<v Colin O'Flynn>I can ask this anything.

01:11:51.320 --> 01:11:52.520
<v Colin O'Flynn>And it gives a good answer.

01:11:53.240 --> 01:11:53.540
<v Chris Gammell>Oh yeah.

01:11:53.760 --> 01:11:53.960
<v Chris Gammell>Yeah.

01:11:54.400 --> 01:11:57.140
<v Chris Gammell>Everybody learns that the hard way, like a couple of times before they're like, Oh,

01:11:57.140 --> 01:11:59.500
<v Chris Gammell>maybe I should think about this and be strategic.

01:11:59.920 --> 01:12:00.600
<v Colin O'Flynn>Yeah, exactly.

01:12:00.720 --> 01:12:00.820
<v Colin O'Flynn>Right.

01:12:00.820 --> 01:12:02.300
<v Colin O'Flynn>And the problem I see is right.

01:12:02.340 --> 01:12:07.180
<v Colin O'Flynn>That if you're teaching a class, like, you know, how many classes you've, you've seen

01:12:07.180 --> 01:12:10.680
<v Colin O'Flynn>where it's like, this is the same material from the last 30 years.

01:12:10.680 --> 01:12:11.000
<v Colin O'Flynn>Right.

01:12:11.000 --> 01:12:14.620
<v Colin O'Flynn>And like, so of course it's giving a great answer for every class question.

01:12:14.780 --> 01:12:15.180
<v Colin O'Flynn>Yes.

01:12:15.420 --> 01:12:15.660
<v Colin O'Flynn>Right.

01:12:16.460 --> 01:12:17.220
<v Colin O'Flynn>It's not novel.

01:12:17.460 --> 01:12:22.300
<v Chris Gammell>I guess some of it is the onus kind of goes back on the professor to put more input in

01:12:22.300 --> 01:12:25.620
<v Chris Gammell>for novel, novel questions for novel solutions, that sort of thing.

01:12:25.620 --> 01:12:25.980
<v Chris Gammell>But yeah.

01:12:26.140 --> 01:12:26.360
<v Colin O'Flynn>Yeah.

01:12:26.420 --> 01:12:26.960
<v Colin O'Flynn>Yeah, exactly.

01:12:26.960 --> 01:12:29.660
<v Colin O'Flynn>Which is, and I think that's kind of what we're getting at.

01:12:29.740 --> 01:12:29.900
<v Colin O'Flynn>Right.

01:12:29.960 --> 01:12:31.540
<v Colin O'Flynn>Is, which I don't think is a bad thing.

01:12:31.540 --> 01:12:36.100
<v Colin O'Flynn>Like, you know, when I went through university, there was, there wasn't chat GPT, but there was

01:12:36.100 --> 01:12:40.540
<v Colin O'Flynn>people that had, you know, burned CDRs of a whole bunch of scanned assignments from like

01:12:40.540 --> 01:12:42.500
<v Colin O'Flynn>10 years.

01:12:42.500 --> 01:12:43.700
<v Chris Gammell>And they would be like, here's all these.

01:12:43.700 --> 01:12:44.780
<v Chris Gammell>There is an old sentence.

01:12:45.700 --> 01:12:47.520
<v Chris Gammell>Burn CDRs of old assignments.

01:12:47.960 --> 01:12:48.260
<v Colin O'Flynn>Yeah.

01:12:48.380 --> 01:12:48.580
<v Colin O'Flynn>Right.

01:12:48.660 --> 01:12:52.020
<v Colin O'Flynn>But I mean, it's to some degree the same, like cheating's always there.

01:12:52.220 --> 01:12:54.460
<v Chris Gammell>Like, yes, it's more effort.

01:12:54.460 --> 01:12:57.980
<v Chris Gammell>And at the end of the day, those people did also did not learn those things.

01:12:57.980 --> 01:12:58.180
<v Chris Gammell>Right.

01:12:58.200 --> 01:13:00.420
<v Chris Gammell>I mean, that's same, same, same, right.

01:13:00.420 --> 01:13:05.260
<v Chris Gammell>It's like, not like the person who, you know, just formulaically, you know, use the old

01:13:05.260 --> 01:13:08.240
<v Chris Gammell>assignments to, you know, fill in the blanks and get the output.

01:13:08.300 --> 01:13:10.300
<v Chris Gammell>It's like, they got the grade, they didn't get the knowledge.

01:13:10.640 --> 01:13:13.940
<v Chris Gammell>And then like, okay, but we need the knowledge.

01:13:14.020 --> 01:13:15.900
<v Chris Gammell>We still need people who know things.

01:13:16.140 --> 01:13:16.260
<v Chris Gammell>Right.

01:13:16.740 --> 01:13:17.100
<v Chris Gammell>Exactly.

01:13:17.260 --> 01:13:20.480
<v Colin O'Flynn>And that's kind of what I get at when people, you know, cause some props will really be

01:13:20.480 --> 01:13:22.640
<v Colin O'Flynn>like, ah, chat GPTs, you know, ruined everything.

01:13:23.400 --> 01:13:26.620
<v Colin O'Flynn>Um, you know, I think it's always been ruined in a similar way.

01:13:26.620 --> 01:13:31.200
<v Colin O'Flynn>Like it's, you know, if you, if you're realistic, maybe it's more obvious now.

01:13:31.200 --> 01:13:35.940
<v Chris Gammell>Like these, these kids are using abacuses and that, that makes the math easier.

01:13:36.460 --> 01:13:36.820
<v Chris Gammell>Yeah.

01:13:38.220 --> 01:13:38.580
<v Chris Gammell>Yeah.

01:13:39.000 --> 01:13:43.340
<v Chris Gammell>Well, I mean, are you, uh, you know, you interact with students and stuff like that too.

01:13:43.420 --> 01:13:46.620
<v Chris Gammell>Does it make you more hopeful though, that they are, have access to more knowledge?

01:13:46.660 --> 01:13:48.920
<v Chris Gammell>Is it more novel solutions for things as well?

01:13:49.360 --> 01:13:49.520
<v Colin O'Flynn>Yeah.

01:13:49.520 --> 01:13:53.400
<v Colin O'Flynn>I mean, I definitely see, um, you know, especially, I mean, this course was interesting because,

01:13:53.540 --> 01:13:56.420
<v Colin O'Flynn>because it was sort of a new topic for me because it was this sort of cover.

01:13:56.420 --> 01:14:00.260
<v Colin O'Flynn>Um, so, you know, there was like a few groups like, oh, you know, like I made this whole

01:14:00.260 --> 01:14:03.360
<v Colin O'Flynn>simulator for my robot and it's like, oh, wow, that's, I didn't even know.

01:14:03.460 --> 01:14:07.160
<v Colin O'Flynn>That's why I don't know how to do that because this isn't stuff I do a lot, you know?

01:14:07.160 --> 01:14:10.500
<v Colin O'Flynn>So there are still, when I, you know, when I went through, I remember being like, you

01:14:10.500 --> 01:14:16.640
<v Colin O'Flynn>know, there's a few really super engaged students who want to do all the things.

01:14:16.880 --> 01:14:18.360
<v Colin O'Flynn>Um, so it still seemed that way.

01:14:18.360 --> 01:14:22.860
<v Colin O'Flynn>Like it definitely, there was some very engaged students and I think, you know, they were doing

01:14:22.860 --> 01:14:27.980
<v Colin O'Flynn>way, you know, like there were 3d printing tons of stuff and a lot of people were doing

01:14:27.980 --> 01:14:32.340
<v Colin O'Flynn>that, but, you know, looking at PCB designs, they were doing advanced coding.

01:14:32.800 --> 01:14:33.980
<v Colin O'Flynn>Um, that's awesome.

01:14:34.220 --> 01:14:34.400
<v Chris Gammell>Yeah.

01:14:34.520 --> 01:14:35.120
<v Chris Gammell>That's the thing.

01:14:35.120 --> 01:14:39.740
<v Chris Gammell>Like it's, it is like, it opens up new spaces and things, you know, barriers have come down.

01:14:39.820 --> 01:14:40.380
<v Chris Gammell>That's really great.

01:14:40.380 --> 01:14:40.920
<v Chris Gammell>I feel like.

01:14:40.920 --> 01:14:46.220
<v Chris Gammell>And I guess my, my experience with that is just like, uh, being in the hardware industry,

01:14:46.460 --> 01:14:49.740
<v Chris Gammell>interact with, you know, younger people are coming in that are like more software savvy

01:14:49.740 --> 01:14:50.020
<v Chris Gammell>too.

01:14:50.020 --> 01:14:54.140
<v Chris Gammell>And just like software methodology brought into the hardware space, like that has been

01:14:54.140 --> 01:15:01.340
<v Chris Gammell>very positive as a general development, um, kind of across the board, you know, it's chip

01:15:01.340 --> 01:15:04.460
<v Chris Gammell>vendors can't get away with the same kind of stuff that they might've put out.

01:15:04.460 --> 01:15:08.240
<v Chris Gammell>And then, but then they also hire smart people to build software than that.

01:15:08.240 --> 01:15:11.340
<v Chris Gammell>It's just kind of like rising tide sort of thing lifts all boats, which is good.

01:15:11.780 --> 01:15:12.560
<v Colin O'Flynn>Not very much so.

01:15:13.200 --> 01:15:13.340
<v Chris Gammell>Yeah.

01:15:13.680 --> 01:15:16.060
<v Chris Gammell>Like if I started out in Ross too, right now, I wouldn't know where to start.

01:15:16.060 --> 01:15:19.820
<v Chris Gammell>So like if there's a bunch of beginner content that like everybody can start with and that

01:15:19.820 --> 01:15:21.880
<v Chris Gammell>helps me too, when I go into a new topic area.

01:15:22.180 --> 01:15:22.220
<v Chris Gammell>So.

01:15:22.420 --> 01:15:23.280
<v Colin O'Flynn>Yeah, no, exactly.

01:15:23.340 --> 01:15:23.460
<v Colin O'Flynn>Right.

01:15:23.480 --> 01:15:23.900
<v Colin O'Flynn>That's the thing.

01:15:23.960 --> 01:15:28.080
<v Colin O'Flynn>There was, there was tons of material out there and like, yeah, like this is, and the,

01:15:28.120 --> 01:15:31.800
<v Colin O'Flynn>you know, the discussion in this course and, you know, we keep having this discussion in

01:15:31.800 --> 01:15:34.840
<v Colin O'Flynn>various courses is, you know, where do you start?

01:15:34.900 --> 01:15:36.180
<v Colin O'Flynn>Because I think there's only so much time.

01:15:36.180 --> 01:15:37.460
<v Colin O'Flynn>There's only so much you're going to cover.

01:15:37.680 --> 01:15:37.720
<v Colin O'Flynn>Yeah.

01:15:37.720 --> 01:15:38.040
<v Colin O'Flynn>Right.

01:15:38.160 --> 01:15:38.380
<v Colin O'Flynn>Right.

01:15:38.380 --> 01:15:42.380
<v Colin O'Flynn>So it's like, do you start at like, you know, we're still doing resistor networks and stuff,

01:15:42.460 --> 01:15:45.600
<v Colin O'Flynn>which is important, but also you then don't have time for.

01:15:46.120 --> 01:15:46.260
<v Chris Gammell>Yeah.

01:15:46.440 --> 01:15:46.660
<v Chris Gammell>Yeah.

01:15:46.740 --> 01:15:49.400
<v Chris Gammell>I haven't run into many resistor networks myself in.

01:15:49.540 --> 01:15:49.780
<v Chris Gammell>Yeah.

01:15:50.160 --> 01:15:50.760
<v Chris Gammell>Well, exactly.

01:15:51.280 --> 01:15:55.520
<v Colin O'Flynn>So like it's, but the argument is, it's like, well, this is used and, you know, heat flow

01:15:55.520 --> 01:15:57.120
<v Colin O'Flynn>and there's a bunch of stuff where it's used.

01:15:57.180 --> 01:15:57.460
<v Colin O'Flynn>Yeah.

01:15:57.520 --> 01:15:57.760
<v Chris Gammell>Right.

01:15:58.000 --> 01:15:58.200
<v Chris Gammell>Right.

01:15:58.460 --> 01:16:00.780
<v Colin O'Flynn>Which isn't wrong, but also.

01:16:01.380 --> 01:16:01.740
<v Colin O'Flynn>Right.

01:16:02.100 --> 01:16:06.380
<v Chris Gammell>I, I, I, sometimes I feel like it comes down to like practical versus academic, like end goal

01:16:06.380 --> 01:16:07.140
<v Chris Gammell>as well.

01:16:07.140 --> 01:16:11.980
<v Chris Gammell>So like, if someone's going to be like going to get a PhD in like some thermodynamic field,

01:16:11.980 --> 01:16:13.880
<v Chris Gammell>then they probably need to know that sort of thing.

01:16:13.880 --> 01:16:19.160
<v Chris Gammell>But like me as like, you know, a dummy, dumb, dumb engineer, who's just going to make things,

01:16:19.640 --> 01:16:20.100
<v Chris Gammell>build things.

01:16:20.100 --> 01:16:21.520
<v Chris Gammell>And like, hopefully they work.

01:16:21.900 --> 01:16:26.220
<v Chris Gammell>Uh, you know, it's just a different outcome that I'm hoping for.

01:16:26.220 --> 01:16:29.500
<v Chris Gammell>Like, I never wanted to know all the base physics about things.

01:16:29.540 --> 01:16:34.180
<v Chris Gammell>Cause it's like, I just wanted to know enough about them to, to move on and like make a thing,

01:16:34.300 --> 01:16:34.680
<v Chris Gammell>you know?

01:16:34.920 --> 01:16:36.560
<v Chris Gammell>So education.

01:16:36.560 --> 01:16:39.520
<v Chris Gammell>Well, you got yourself into quite a space there, Dr.

01:16:39.580 --> 01:16:39.980
<v Chris Gammell>O'Flynn.

01:16:40.320 --> 01:16:40.660
<v Chris Gammell>Yeah.

01:16:40.860 --> 01:16:47.140
<v Colin O'Flynn>That's something, but I mean, it is, uh, you know, I think it's interesting to see what

01:16:47.140 --> 01:16:48.000
<v Colin O'Flynn>people are coming up with.

01:16:48.000 --> 01:16:49.300
<v Colin O'Flynn>Like it is, as I said, it is amazing.

01:16:49.780 --> 01:16:55.540
<v Colin O'Flynn>You know, what the capabilities are and lots of people using, you know, little Linux things

01:16:55.540 --> 01:16:56.460
<v Colin O'Flynn>and stuff like that.

01:16:56.460 --> 01:16:56.700
<v Colin O'Flynn>Yeah.

01:16:56.940 --> 01:16:57.200
<v Chris Gammell>Yes.

01:16:57.460 --> 01:16:57.620
<v Chris Gammell>Yeah.

01:16:57.820 --> 01:16:58.820
<v Chris Gammell>It's, uh, there's, yeah.

01:16:58.820 --> 01:17:00.900
<v Chris Gammell>Starting from cool, cool spots and doing cool things.

01:17:00.900 --> 01:17:01.360
<v Chris Gammell>That's great.

01:17:01.760 --> 01:17:01.980
<v Colin O'Flynn>Yeah.

01:17:02.340 --> 01:17:02.820
<v Chris Gammell>All right.

01:17:02.820 --> 01:17:06.500
<v Chris Gammell>If people want to see the cool things that you're working on, Colin, where can, uh, where

01:17:06.500 --> 01:17:07.280
<v Chris Gammell>can they find your work?

01:17:07.280 --> 01:17:11.820
<v Colin O'Flynn>So right now I have a half updated blog and call them a Flynn.com.

01:17:11.920 --> 01:17:13.260
<v Colin O'Flynn>That's probably the best spot.

01:17:13.860 --> 01:17:15.740
<v Colin O'Flynn>Um, yeah, I'm trying to post more.

01:17:15.820 --> 01:17:18.000
<v Colin O'Flynn>So I've, I, I recently recorded a video.

01:17:18.000 --> 01:17:23.640
<v Colin O'Flynn>So for this class, actually, uh, kind of, you know, I called an intro to PCB video series,

01:17:23.640 --> 01:17:31.960
<v Colin O'Flynn>but again, tons of caveats because, um, the, it was designed to be shoved within like one

01:17:31.960 --> 01:17:32.900
<v Colin O'Flynn>or two classes.

01:17:32.900 --> 01:17:37.660
<v Colin O'Flynn>So, you know, there's only so much you could do in there, but yeah, I, my, my goal is to

01:17:37.660 --> 01:17:39.920
<v Colin O'Flynn>get more content posted.

01:17:40.440 --> 01:17:43.060
<v Colin O'Flynn>Um, a few more teardowns too.

01:17:43.060 --> 01:17:48.160
<v Colin O'Flynn>So I, one of the things I was looking at with some RF stuff, so like a Marine X-band radar,

01:17:48.800 --> 01:17:51.500
<v Colin O'Flynn>um, looking at how that works and yeah.

01:17:51.640 --> 01:17:51.860
<v Colin O'Flynn>Yeah.

01:17:51.940 --> 01:17:53.080
<v Colin O'Flynn>A few new topics, right?

01:17:53.100 --> 01:17:56.700
<v Colin O'Flynn>I mean, one of the fun things about academia is you have so much crossover, so I'm trying

01:17:56.700 --> 01:18:00.700
<v Colin O'Flynn>to embrace that a little more than just pure security, stuff like that.

01:18:01.040 --> 01:18:02.600
<v Chris Gammell>More fun things coming from Colin then as well.

01:18:02.720 --> 01:18:04.900
<v Chris Gammell>And we'll, uh, you know, you'll be back at five years.

01:18:05.000 --> 01:18:08.520
<v Chris Gammell>We'll talk about it then and see how it is soup and, you know, see how it goes.

01:18:09.160 --> 01:18:11.380
<v Colin O'Flynn>See how life is chat GPT.

01:18:11.380 --> 01:18:13.400
<v Colin O'Flynn>Maybe it'll just do the interview for us actually.

01:18:13.500 --> 01:18:13.980
<v Chris Gammell>Yeah, exactly.

01:18:14.080 --> 01:18:14.760
<v Chris Gammell>Yeah, exactly.

01:18:14.880 --> 01:18:15.520
<v Chris Gammell>No content.

01:18:16.500 --> 01:18:17.200
<v Colin O'Flynn>Yeah, exactly.

01:18:17.440 --> 01:18:18.080
<v Chris Gammell>Schedule it now.

01:18:19.520 --> 01:18:19.880
<v Chris Gammell>All right.

01:18:19.920 --> 01:18:20.660
<v Chris Gammell>Well, thanks for being here Colin.

01:18:20.860 --> 01:18:21.160
<v Chris Gammell>Good chat.

01:18:21.180 --> 01:18:22.180
<v Chris Gammell>Thanks so much for having me Chris.

01:18:22.180 --> 01:18:52.160
<v Chris Gammell>Outro Music
