#625 – Gremlins in the machine

Chris Gammell: This is The Amp Hour Podcast. Release March 26th, 2023. Episode 625. Gremlins in the Machine.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV Blog.

Chris Gammell: And I'm Chris Gammell of Contextual Electronics.

Dave Jones: That fire alarm was not me.

Chris Gammell: This time. I was watching a video of someone puncturing a spicy fire pillow this morning. A lithium cell going off. Oh, right.

Dave Jones: Yes, we've got that on the list. Yes, yes.

Chris Gammell: Oh, was that on the list? I didn't even see it on the list. Yeah, I think unmanaged. Someone had sent it to me.

Dave Jones: 615, put it on there.

Chris Gammell: Those things really, they still scare the crap out of me.

Dave Jones: For those of you who haven't seen it, we'll link it in. He cut through an 18650 cell. I presume it was fully charged, Tommy. I can't remember the exact details. With one of those pneumatic cordless pruner things, I guess you could call them. I guess they're like tree pruners or something.

Chris Gammell: Oh, like the one that this old Tony does too, huh? Yeah.

Dave Jones: What's this old Tony?

Chris Gammell: This old Tony is like another machining channel. Oh, okay. Right. Yeah. Right.

Dave Jones: Got it. It's great. Great job. When was the last time you saw an Arduino on his channel?

Chris Gammell: Yeah, a long time ago.

Dave Jones: For those who don't know, AVE stands for Arduino versus evil.

Chris Gammell: Yeah.

Speaker ?: Yeah.

Dave Jones: So, yeah.

Chris Gammell: It's gone far, far afield.

Dave Jones: He did one Arduino video like 10 years ago or something.

Chris Gammell: No, he did a couple.

Dave Jones: Oh, did he? I can't remember. It's so long again.

Chris Gammell: Yeah. Yeah.

Dave Jones: Oh, boy. I'm going to have to go back and watch them actually because I don't remember what they were. Interesting. So, but anyway. Yes. We had to delay the recording of this because there was a fire alarm. Fire alarms. Yeah. Apparently. Safety third. Safety third. Yes. It wasn't a real one. It was a test, even though they didn't come and tell us. They did give us forewarning, I don't know, a week ago or something.

Chris Gammell: Yeah.

Dave Jones: And yeah, apparently some people have complained that they couldn't hear the last alarm. So, yeah, they did like an audit of the building or something. So, they set it off and they left it going, you know, because normally like the siren actually turns off after a little, you know, after a couple of minutes or something. And then the lights just flash and the fire engine comes and, you know, et cetera. But no, they just kept it blaring.

Chris Gammell: Maybe different country to country too. I don't know. Oh, yeah. I'm sure. I'm sure we've got different regulations.

Dave Jones: Well, how does it work over there? Yankee land.

Chris Gammell: I don't really sit in commercial buildings anymore.

Dave Jones: Well, I was going to say, when was the last time you were in a multi-story commercial building?

Chris Gammell: Let's see. So, it would have been pre-pandemic. Yeah. Yeah. Yeah. I don't know the last time I was in around for a fire drill.

Dave Jones: Yeah.

Chris Gammell: They used to have them in my condo building and that was not fun. Oh, okay.

Dave Jones: Right.

Chris Gammell: Yeah. And like hotels, I feel like every time I go to a hotel. Oh, yeah. That's when they go off, you know. Yeah. Yeah.

Dave Jones: I've only had that happen once, I think. I was in Singapore and it was like 2 a.m. in the morning. The freaking fire alarm went off, you know.

Chris Gammell: Mm-hmm.

Dave Jones: So, it's like, yeah, that was startling.

Chris Gammell: And here's the thing, like, I've slept in a lot of hotel rooms, but I always remember the ones where the fire alarm goes off.

Dave Jones: Oh, right. Yeah, yeah. Well, I can vividly remember coming out at 2 a.m. in the morning. Yeah, exactly.

Chris Gammell: That is a jarring experience.

Dave Jones: Yeah. Yeah. And you don't come out in your pajamas. Like, you actually put- No, no, no. No, you go to the effort to put some actual clothes on and grab all your stuff just in case you can't come back, you know. So, it's like, yeah. No. Ah, well.

Chris Gammell: Well, speaking of being in hotels overseas, I just, I got back from Germany. I think I made it sans COVID this year. And, yeah. How long did the trip take? Trade show. On the way out there, it took me 28 hours.

Dave Jones: 28 hours, yeah.

Chris Gammell: Yeah, I missed a plane.

Dave Jones: That's basically the same from here. It'll take you 28 hours to get to Germany from here.

Chris Gammell: Yeah, yeah. It was a pretty rough day that day. All right. Not as bad on the way home. So, yeah. Right.

Dave Jones: Okay.

Chris Gammell: But, yeah. Still a bit of a haul. I think it's more like 40 if you get layovers and stuff. It's got to be a ton.

Dave Jones: Do you fly business or cattle?

Chris Gammell: Oh, no cattle class. Come on, man. Cattle. We're going to start up.

Dave Jones: Couldn't even swing it. Oh, why? All the money was in Silicon Valley Bank, was it? So, they didn't have enough liquidity to pay for business class.

Chris Gammell: That was not us. So, that's a good thing. That was a bit of a wild ride.

Speaker ?: Well, yeah.

Dave Jones: Like, should we talk about that? Because that's like apparently 50% of startups in Silicon Valley used Silicon Valley Bank. Like, for a while. We could talk about it.

Chris Gammell: I think probably people are sick of it by this point. Yeah, probably.

Dave Jones: That was a huge deal.

Chris Gammell: It does have actual impact. Yeah, sure, sure, sure. Yeah, yeah. I mean, like. Massive. A lot of people were scared.

Dave Jones: Yeah.

Chris Gammell: You know. It's a scary time when stuff like that happens, right? I mean.

Dave Jones: Yeah.

Chris Gammell: I don't know what else to say.

Dave Jones: Well, I see. People were complaining about, oh, what are you worried about? You got 250 grand, you know, federal deposit insurance. It's the same here, I think it's 200,000 or something here, but similar order, which the government actually protect your bank deposits. And you can still get it within, even if they didn't save the bank, which they did. Oh, well, they saved, it sold, and they secured all the deposits. They guaranteed all the deposits. But for three days there or something, it was all like, holy crap, we can't get our money out. Yeah, it was a weekend.

Chris Gammell: It was like a Friday or a Monday.

Dave Jones: And people don't realize that if you're even a small to medium business, 250K is nothing. Like you're just like, you cannot pay a week's wages. You are one week away from not being able to pay wages and people start walking out, right? Yeah. If you can't access that cash. So it's like, oh, why don't you spread it across 10 different banks? Hello?

Chris Gammell: Yeah.

Dave Jones: Have you tried the red tape involved in doing that with your accounting system? And oh, God, it's just shocking. It's a mess. Yeah. I just did my taxes the other day. That was fun.

Chris Gammell: Yeah. Okay.

Dave Jones: Yeah.

Chris Gammell: I'm glad it stuck around. I mean, I'm glad the depositors are not going to lose their dough.

Dave Jones: They're not going to lose their dough, but all their investors did. SVB is no more. Yeah.

Chris Gammell: Yeah. That's right. Yeah. So. Yeah. I mean, like when I think about people listening to this.

Dave Jones: They had a haircut is the industry term. They actually had to take a haircut.

Chris Gammell: A full haircut. Yeah.

Dave Jones: A full back and sides, you know. That's right. Yeah. Yeah. Is it called that in America? Back and sides?

Chris Gammell: What's that?

Dave Jones: A full back and sides is like, you know, short, short back and sides. Take it all the way down. Yeah. It's like a short, short back and sides.

Chris Gammell: I always hear like the military ones, like the high and tights and stuff like that. Oh, right. Okay. That's it.

Dave Jones: That's just slip in that number two and go all the way, you know.

Chris Gammell: Yeah. I think people listening to this, right. So if you're listening to this and you're like, well, why the hell should I care? I think one thing to think about is if you're ever going to join like a small company, you're probably not going to be able to ask them like, where do you have your money? I mean, maybe.

Dave Jones: Yeah. Right. Yeah. That's not an interview. It's not a recommended interview question, you know.

Chris Gammell: Yeah. Yeah.

Dave Jones: So just putting it out there.

Chris Gammell: I do think it's super legit. Like if you, you know, so I've been part of two, two and a half startups now, if you count supply frame as a startup.

Dave Jones: All right.

Chris Gammell: They were very, very big. Like, you know, it's super legit to ask about like, you know, what is, how much, how much runway do you have? That's the nice way to ask how much money do you have in the bank, right? Runway. Yes. That's the legit terminology. Yeah. That's the phrase you get. And, you know, past startups I've been at, they told me a number that was very satisfying during the interview and it still didn't really matter because condition changed and like, that's fine.

Dave Jones: Right. So they actively tell you, you didn't have to ask them. They actively.

Chris Gammell: No, I did ask them. Oh, okay. What I'm saying is they told me a number that was good and then there were still layoffs. So like that, that's because business change. And so there's like, you know, you can't control the stuff outside of the company and just how stuff like that changed. But, you know, it's really important. I think to, I would, you know, if I was getting hired, when I got hired by ABB, I didn't ask them about their cashflow. Right.

Dave Jones: No, exactly. They're so big.

Chris Gammell: Yeah. They're conglomerate. Right. And there was almost 0% chance you could get it.

Dave Jones: Well, you wouldn't ask that of a normal business. You wouldn't ask that of a normal in quote marks business. It'd only be a startup question.

Chris Gammell: Sure. I don't know. I don't know.

Dave Jones: Or is that just, I don't know.

Chris Gammell: Where do you draw the line to like normal business, I suppose, right? That's the real thing.

Dave Jones: Well, okay. If it's a small business, but any like business that like you're being interviewed by a HR person, for example, it's like, you know, like once you've reached that HR person level, the company's big enough to sort of like, I don't know. Yeah.

Chris Gammell: That's a good point. Yeah. So like I have a family member that runs a small business and I think, I don't know. I think that's a totally legit question. Like it's, it's not asking how much money do you have? It's, you know, if, if things went sideways tomorrow, how long could you survive? I think that's a sign of like fiscal responsibility and understanding your business. Like I think every business owner should at least understand that because there's like a, you know, there's a burn rate. Like you mentioned, right?

Dave Jones: If you can't make payroll, you're screwed.

Chris Gammell: I, I, I heard an interview question that I thought was, I had never heard it before, but I think it's actually a really great question is, which is like, have you never not been able to make payroll? I think that's actually a really great question to ask in an interview, right? I mean like, because it might've been a one-time thing, but a one-time thing is different than a one-time thing and like a three-time thing, even with good explanations.

Dave Jones: Well, it's debatable because you couldn't, can say a percentage of companies are going to take that the wrong way.

Chris Gammell: Yeah. I think it's a very well thought out.

Dave Jones: I think it's, it's, it shows that, you know, the business.

Chris Gammell: I think it's, it shows experience. It shows that you know the business. Okay. Right. And then you're looking out for your own interests and that you're, you know, like you're, how about this? Younger Chris would have been like, yeah, whatever, you know, I'll work anywhere. You know, like this is like, it just shows that like you have different priorities in life and that you, they need to be respected and like, you're not going to just stomp on my paycheck, you know, just because you're trying out some new crazy idea, like a new check mark system.

Dave Jones: Yeah. Well, if you ask that question, it means that you don't have a lot of screw you money. That to me, like if I was an employer, somebody asked that, I would take that as going, they don't have screw you money. Okay. They're going to stick around.

Chris Gammell: I don't think that's the case.

Dave Jones: Well, they're, they're more likely on average. They're more likely to stick around. No, you don't agree.

Chris Gammell: I don't think that question is the case. No, I think it, I mean, it does depend on your ask it though. Right. Like if you say like, I have to get every page, I have to get paid up to each other. Like that's just reeks of desperation. Right. But I think, I think what, you know, when I heard that question asked, it's, it's the sign of, that's the sign of experience, right? It's the sign of someone who's been through a bad startup in the past. Right. I think that's, that doesn't, that has nothing to do with like personal cashflow. I think.

Dave Jones: Sure. But some people might interpret that as, well, this person needs money. Sure. Yeah.

Chris Gammell: You, you, you can't, you can't stop how people interpret it. But anyways, I, I, I respect those questions in an interview coming from the person I'm interviewing. I think that's actually, like I said, the sign of experience and a good thing to ask. And, and if I'm not the person to answer it, then like, you know, then I should be able to point them at the person who can.

Dave Jones: Right. All right. Well, speaking of jobs.

Chris Gammell: Yeah. You're going to get one finally. I told you so.

Dave Jones: No, I've, I've got a perfect segue here. Okay. A company called Anthropic. They're advertising a prompt engineer.

Chris Gammell: Oh yeah. I saw this prompt engineer. Yeah. Yeah. That's great. Yeah. How did you find this? Was it just like on Twitter? Where was it?

Dave Jones: It showed up in my feed somewhere. Sorry. I don't know who posted it, but yeah. You know, things just show up in your Twitter feed at random, you know, because somebody else you follow retweeted it or something.

Chris Gammell: I don't know.

Dave Jones: Yeah. But it showed up and Anthropic with a slash instead of an I in PIC there. So yeah. Wank. You know, paid some consultant bloody, you know, a hundred thousand bucks to come up with that. Right. Right. Right. Yeah. And yeah. Anyway, so it's a, it's a prompt engineer. We've talked about this. It's basically using how to use chat GPT or other, some other AI. In this particular case, it's an AI company. They've got their own AI tech.

Chris Gammell: Yeah. It seemed like almost like an audit, audit of like their stuff. Right. So you're trying to find the most radical ways to kind of use and abuse. Use their own stuff.

Dave Jones: Use and abuse their own stuff. And, but yeah, that's. And, but the thing is it pays between 175,000 and 335,000 a year.

Chris Gammell: That is spicy. Yeah. That's a lot of dough. That is spicy.

Dave Jones: Right. That is a lot of dough.

Chris Gammell: That's in San Francisco though. So that's like, I don't know, like 50 grand and other places.

Dave Jones: Yeah. Yeah. You know. Exactly. Well, no, I would say that's, you know, that's a good, somebody, I think through the number of 140K at me here and I'm still going, well, 140K for a job like this. Literally, you could learn to do this in a day or two. I'm not kidding. There are no, basically.

Chris Gammell: Learn it in a day or two. Take a lifetime of chat GPT. Take a bluff.

Dave Jones: You, as long as you sell yourself well at the interview and, and you gave them examples of, you could cram for this in 24 to 48 hours, right? You can cram for this interview. Seriously. I am not kidding. And if you were well-spoken and, you know.

Chris Gammell: Do they have requirements on the education or no? What is the.

Dave Jones: Have you got the link?

Chris Gammell: I've got it here. Yeah.

Dave Jones: Yeah. Yeah. Here we go. You may be a good fit. If you have at least a high level familiarity with the architecture and operation of large language models, you don't even need detailed, like of how they work. You just need.

Chris Gammell: You know, here's the thing. These kinds of posts, they are super like open sounding, whatever. They're going to hire the person who worked at open, open AI in the past. That is the only way you're ever getting paid $335,000 a year is if you're like senior engineer at open AI working on the stuff that's really popular. That's the only way.

Dave Jones: But what if those jobs, what if you're that good and this is a nothing salary for that sort of position, which it might be. So they end up with all the work experience kiddies coming in and wanting, you know, 100% you know, like.

Chris Gammell: Yeah. Here's what I'll say when I look at this. I do not envy the person has to filter through the amount of crap they're going to get into their inbox for this job. Holy crap. Like this is just such an open.

Dave Jones: It's so open.

Chris Gammell: Anyone could apply for this, right? And with a salary.

Dave Jones: Here is what you're expected to do. Discover, test and document best practices, first wank word, for a wide range of tasks relevant to our customers and build up a library. So that just means muck around all day on the AI and see if you can come up with, right? And you've got to document it, right? You remember the difference between science and what's science and what's not is documented. Write it down. Write it down. It's writing it down. That's science, right? Build up a library of high quality prompts and prompt chains to accomplish a variety of tasks with an easy guide to help users search for one that meets the need. So you've just got to have a basic ability to write a procedure. Like within, as I said, 24, 48 hours, you could do this sort of stuff. Set up a fake website that you've done or a fake blog or something. And with like, seriously. And then build a set of tutorials and interactive tools. Sounds like you should apply, Dave.

Chris Gammell: That's the art of prompt engineering. I should. It sounds like you should apply. This is ridiculous. You have experience discussing prompt engineering, both on your widely lauded podcast and in your own artistic endeavors. Yes.

Dave Jones: Yeah. All you've got to do is sell yourself well and just show a few examples of what you've done.

Chris Gammell: That's true of any job, I think. Yes.

Dave Jones: Yes, it is. But they're not expecting much. They're not. It's not like you have to be a highlight. If you've written a few small Python programs, that's beneficial.

Chris Gammell: Given that your field of prompt engineering is arguably less than two years old, this position is a bit hard to hire for. This is PR. This is exactly what it is. It's PR. Yeah. It worked, though. I mean, like, hey, we're talking about it.

Dave Jones: Yeah, exactly.

Chris Gammell: Good on you, Anthropic.

Dave Jones: But anyway, yes.

Chris Gammell: Stinky AI companies. I'm so sick of all these. I'm so sick of AI, Dave. I'm so sick of it.

Dave Jones: But it did come true literally within months.

Chris Gammell: Sure.

Dave Jones: You know, and the term prompt engineering is now a career, right? And it's paying up to $335,000 a year.

Chris Gammell: Sure. Yeah, this is a new ceiling. Literally, we're gone for nothing. This is an actual job listing. Right, right. Sure, sure. Yeah. Yeah, but how do we know this wasn't written by an AI? Exactly.

Dave Jones: You don't.

Chris Gammell: You don't know. You don't know.

Dave Jones: Anyway, AI is going to be banned soon or parts of it. So that's my prediction. AI will be banned and limited everywhere.

Chris Gammell: Yep. In the other direction, right? So, like, I feel like all of this stuff is about, like, oh, we're not going to have to do anything. And, like, you know, you know, I've talked about, like, it's going to be co-working on stuff, working with prompts and, you know, doing that sort of stuff, all this GitHub CopilotX stuff and all of the AI code generation tools, whatever, whatever, whatever. But here's the thing. I still suck at a lot of things. Like, really bad at a lot of different things. And, like, what's going to happen in that, you know, let's prognosticate 20 years in the future, right? So now the AI stuff is so good that everyone's like, oh, I'll just, you know, put it in the machine. But somebody has to understand how the stuff works underneath, right? Somebody has to, like, know how to do bit masking, what it actually means at a basic level instead of just asking some computer how to do it, right? Is that a stupid thing to think?

Dave Jones: But it's turtles all the way down, right? You can go, I want to do this thing. And then the AI tells you, well, you need this shift in algorithm or something. And you have no idea what that is. Well, you chat GPT that. And it explains what that is. And then you find something else. Well, it's turtles all the way down, right? It'll just keep explaining until the cows come home. Yeah, but, like, actually understanding it doesn't... It doesn't matter.

Chris Gammell: Well, you don't have to understand it if it produces a working result. Yeah, but what about when it breaks? Like, here's the thing. It's not going to work the first time. This is what I keep coming back to. So there's two things here. One is, who's going to troubleshoot all this crap, right? How do we know if it's actually working as expected? Or is it just more, you know, writing new prompts to see if, oh, A didn't work. We should try B, blah, blah, blah.

Dave Jones: Some things are obvious. They either work or they don't. I mean, there's a ton of stuff that falls under their category.

Chris Gammell: It actually works, right? I mean, like, there's stuff where I've worked with experts and they don't know how it works after, you know, 10 years of working on this code, right? I mean, like, it's just like, there's sometimes there's just, there's gremlins in the machine, all right? Right. To use stupid language about it. Yes. Okay, so that's one thing. It's like, someone's got to troubleshoot and fix the broken crap and, like, and just, and there's some things where the large language models aren't going to have enough knowledge of the underlying system, right? Like, there's just some corner cases where it's just not going to work. And like, in, in that, that scenario. Okay. That's one thing. The second thing is now my daughter is going to grow up with this stuff widely available and it's going to probably teach her a lot of things, but it's also going to be a tool that she uses. And so now I think about like, all right, so now 20 years from now, what are the kids that grew up with this stuff? How do they understand the stuff all the way down, the turtles all the way down? Right. I mean, like, like, how do you, having, having grown up with this, right? How do you, how do you dig into some of the underlying knowledge? Now, I, you know, I've been very critical about the education system overall. And I don't, I don't think bottom-up engineering is the right method in a lot of ways, but also when these computers hand you answers.

Dave Jones: Yeah, but this is the ultimate top-down engineering, is it not?

Chris Gammell: It is totally. Yeah. It's totally, yes. It's the ultimate. Right. And so like, from a certain perspective, I should be like embracing this with open arms. And a lot of it, I am. You, you should, Mr. Contextual Electronics. I know. I know, Dave. I know. This is, you're feeling my, my stress here.

Dave Jones: Yeah.

Chris Gammell: But who's going to fix stuff?

Dave Jones: You can feel his world crumbling, folks. You can just. No, no, no.

Chris Gammell: Not crumbling. Just, I'm just introspective and, you know, tired. So, but like, okay. So like, then, so I worry about the troubleshooting now. What about troubleshooting in the future? Who's going to like, who's going to open the hood and dig underneath, right? And who's going to design the circuit board and understand how the circuit board works and be able to fix things? I don't know. Like that's, maybe this is the, the alternative to the prompt engineer is the, the mechanic, the digital mechanic of the future is the true power, the true power holder, because everyone can be a prompt engineer, but only certain people can be the digital mechanics. Right. I don't know. That, it doesn't, it doesn't stack up right now either. Right. I mean, like when I think about like some of the engineers I've worked with in the past who were like, they're really good at assembly language, right? They like grew up and learned assembly language. They're not getting paid that much more, you know, like in terms of like economic value, they're not getting paid that much more, or maybe even they're maybe being paid less because they've been, you know, pigeonholed into a career. Right.

Dave Jones: Yeah, exactly. They're in some niche that's really not that sought after. But so yeah, yeah. Somebody will want it. You'll, you'll always find a job, but it won't be like, you know, like a prompt, it won't pay like a prompt engineering job will.

Chris Gammell: Right. Right. Right. Right. I mean, you and I are still here writing C like a bunch of schlubs while, you know, some designer somewhere is using some high level language to make a fancy fancy website for 600 grand a year. You know, like, so economically, maybe it doesn't. So there's the economic piece of it. That's frustrating and continues to frustrate me. But then there's the actual, just practical, who's going to fix this stuff when it breaks? That's the part I don't understand. And I would love our audience to tell me.

Dave Jones: Well, there's always, you know, it's take any complex product that you've worked on. Right. And the number and the multidiscipline teams involved. Right. Chat GPT can only do a fraction of it sometimes. Sure. Sure. Right. So it's, it's not magic. Right. Yeah. It's going to completely replace and dominate certain fields. Like I just mentioned, I did my tax the other day. Right. You can't tell me that in 10 years time, AI is not going to completely take over the tax accounting side of things. Right. Tax accounting, lawyers. There's a strict set of rules published by the government. All it needs to do is read those and boom. And it knows how to, and you know, there's no point being a freaking accountant. You know, it's just no point. Well, there's still some point, but it's going to be like, some people will only trust humans and they want the human relationship and everything. But you can't say that, that industry is not going to be decimated by AI. Right.

Chris Gammell: Oh, I totally agree with that. Legal field.

Dave Jones: Just, just, you know, it's just doing legal, tedious paperwork. Not, not really the high level stuff, but the lower level stuff is just going to, it's just going to bam.

Chris Gammell: I think it might, it might bifurcate in the same way that we're talking about right here. Right. Yeah, it could. The prompt engineer of the accounting field will be the ones who are able to, you know, twiddle the bits or whatever the equivalent is for these, you know, AI engines that are doing accounting stuff. But then the fixers, there's also the fixers.

Dave Jones: Is, is that a race to the bottom financially? Like job wise, if you're a prompt engineer, it might be highly sought after now, 330 K a year, but in 10 years time, it's 30 K a year because that's a basic expectation of every high school graduate. Right. Is that you can run chat GPT. So they can do prompt engineering.

Chris Gammell: Yeah.

Dave Jones: For X task.

Chris Gammell: You know, these are, these are really good questions. Yeah. Probably.

Dave Jones: I mean, back when I was a boy, right. Just the ability to use Microsoft office tools was a highlight on your resume. Right. Being able to use spreadsheets was a highlight. That's wild, man. I know. It's really wild. Seriously. Right. But, but that was a highly sought after thing. Right. I was, I think we mentioned this a couple of months ago. I was considered a God at my, at one company I worked at because I could insert pictures in Word documents. Right. It was like, you know, people were bowing down. This is just like.

Chris Gammell: And now look at you. You insert photos on the internet. Yeah. Yeah. Right. Sequential order, 24 frames per second.

Dave Jones: It's nuts. Oh God. Yeah. Yeah.

Chris Gammell: All right. So what I was getting at, what I was starting to lean towards with this is I not actually tied to AI because I'm trying not to think about AI as much as possible, but I have been thinking about learning plans and educating myself. Foo. Educate yourself, foo. Yeah. So I've actually been thinking through this, you know, I got a lot to learn, Dave, and I'm trying to figure out like, so I, you know, I think I've told you I've been, I've been doing language classes and in the past I've done like piano classes. Yep. But there are no electronic classes. I mean, or like even coding classes, I think they're always kind of structured, not like lessons are, not like a piano lesson is or a language lesson is.

Dave Jones: What are you getting at?

Chris Gammell: Well, that's like a structured kind of thing where there's like drills. So part of me is like, Oh, you should go and revise contextual electronics to do this sort of thing. And I'm like, back up, Chris, you got to learn this stuff first. So what I'm really thinking is like, I got to like, you know, lay out a learning plan for myself to do this sort of thing, right? It's almost like I need like a coach and then I need like a, not a curriculum, but like a-

Dave Jones: You're going to have to give me a specific example because I'm not really following you here.

Chris Gammell: Here's some things that I said I would like to be doing. I wrote a post about this on my consulting forum. So this is, this is the, uh, the structure, the structure of this. One thing I said that I'd like to do is build a embedded Linux system like Jay Carlson wrote about. So you remember that post, a fantastic post when Jay came on the show and was talking about it. Yeah. That was amazing.

Dave Jones: Right. Yeah. Yeah.

Chris Gammell: Yeah. So Jay has like three very seminal pieces on his blog. One is like the $1 microcontroller. Another is the one that built an embedded system. And then one's about his lab. All of them are just fantastic. Among other things. I think there's some others on there. Another one on the firmware side of things. Right. So basically bottom up, pull together an RTOS design without any assistance. So just kind of just do it myself. So that's kind of like a higher level IOT style product. Yeah. But do it without all of the resources that I kind of have available day to day. So take like a free RTOS design, kind of plug it all together and kind of build it from the bottom up.

Dave Jones: Well, once again, it's all going to depend on which RTOS you use, you know, like.

Chris Gammell: Yeah. Yeah. In that case, that would be free RTOS. Specifically free RTOS. And this was a response from a forum member saying, you know, as well saying like, yeah, you should like glue stuff together, like the network stack and the, you know, putting in different drivers and yeah, having a good project to go towards that. So that's kind of like a high level system.

Dave Jones: You want software stack engineering skills. Do you? Yeah.

Chris Gammell: Is that, I think the, the real thing I want there is the one thing I want to get like the drill on and get better at is being able to kind of glue together pieces. Right. So like some of the really good firmware engineers I've seen, they're able to like take examples from various disparate parts of the internet. Yes.

Dave Jones: And then just glue, like within minutes, they're all tied together. I know. I'm so jealous. Yeah. Yeah. Yeah. Exactly. Yeah.

Chris Gammell: So they'll like pull in like an LVGL driver and they're like, oh yeah, I know. I've seen this kind of display before and on the, I get a display driver going with like some example code.

Dave Jones: Yeah. I know. I hate it. Yeah, exactly.

Chris Gammell: And it's like, it's, it's fantastic. And, but like that comes, like I, I realized that comes from hard work and then a lot of experience and I need to just kind of find that experience and like cycle on that. Right. The other one was like low level type stuff. So like that, that, uh, risk 32 part, right? Like the, the one, the cheaper one that I have to buy a bunch of, that's right. The CS 32 V zero zero three, like, so then that's like a low level one. And so I want to work more on like building low level kind of bit masking to, you know, like basically direct register talking in C that sort of thing. Another thing was building out build pipelines, like CICD build pipelines for firmware. Cause that's something that's just like deployable everywhere. Right. If you get into that, you build containers. That's something I want to do more scripting. I feel like there's a lot of scripting stuff that I need to get better at. So like Python really just learning Python.

Dave Jones: And then I said, I thought you were, were already a Python guy.

Chris Gammell: I have used Python, but I'm not comfortable in Python.

Dave Jones: Okay. Oh, okay. That's news to me.

Chris Gammell: Yeah. Yeah. No, no. And then like, I feel like I need to build up. So I've been using a lot of free CAD lately. I've been feeling better about it, but like, still it's like a, that's like a muscle you got to work out. You know what I mean? So like, it's almost like Sophie Wong was on the contextual electronics podcast early on. And she talked about doing the 30 days of like modeling, like of like objects. I feel like that's like the kind of like repetition type of stuff. Yeah.

Dave Jones: That sounds about right. You'd want to be doing models for a month before you get kind of fluid. Right. Right.

Chris Gammell: And then there's like, but then there's like, so like thinking about, so that's a good example, right? So that's like, that's like working out is like a, that's like a, every kind of day you're trying new things. Maybe you're pulling in new ways to, you know, so maybe I'm working on like revolving shapes in 3d modeling to like get more comfortable with that sort of thing. But like just thinking about, I need to build that muscle, right? You need to like build that muscle and then maintain that muscle. And so that's what I'm talking about here. Like that's like a learning, like I think about a learning plan is like, or like, if I was taking a language course, he'd be like, oh, well, the teacher would be like, well, you're, you're good at modeling 3d boxes in, in 3d software, but you're not good at doing cones. Let's work on cones today.

Dave Jones: And that, I see where you're coming from. Okay. Right.

Chris Gammell: So like, imagine that in each of these fields that I was talking about, I also said, yep. Content. So I said, basically hardware firmware, software engineering, 3d modeling and content. Yeah. And that's, that's kind of like, when I think about that sort of thing and you know, it's a lifelong goal, but like, I need like a coach, I guess. Like a.

Dave Jones: Well, you can, there's this guy who has this website called contextual electronics who will guide you through these sorts of things.

Chris Gammell: That's not what he does. I mean, yeah, he does that. He does that for some people, but like, I need that for me. I need like a coach. Just need coach coaching too, you know? Right.

Dave Jones: Well, I don't know. Is there, has anyone seen anyone out there that offers that sort of electronics tutoring?

Chris Gammell: Coaching. Yeah. I guess coaching. I think about it like sports coaching though, almost like it's like sports coaching. Like even like Tiger Woods has a coach. It's a high performance.

Dave Jones: Would that be a high performance? No, that's not a high performance coach. It's kind of like a. Yeah. Yeah.

Chris Gammell: I don't know. I think one thing is like, you need to like have someone who also sees your weakness and then like can help break it down. Right. That's like a really good coach does that.

Dave Jones: But unfortunately that's not just like, you know, you use the example of music or, you know, a friend, like a language or something like that. That's, it's not as easy to spot in engineering because it's, it's there because the whole process is so long and convoluted that you can't just go, okay, let's just sit down for an hour and see what your weaknesses are. It's like, you know, like you've got to do, do a project over months and then like, oh,

Chris Gammell: it's just, is it even possible? One thing with coaching, like a really, I think coaches are good at like, like when you are avoiding something, that's like a really good sign for a coach to like hone in on it. Yes. Right. Like, so if you're like, so like you're like weightlifting and stuff like that. Like if you're like, like, like, like favoring your, your right shoulder, they might be like, what, why are you doing that? Right. You're like, you're pulling your right shoulder. Why are you doing that? And it's like, let's go work on building up your right shoulder more or something like that, you know? Yeah. But that's really tough. It is tough to do. And it's, I mean, honestly, it's expensive to time and money.

Dave Jones: Yeah. No, I'm not even sure that's entirely possible in the electronics design realm.

Chris Gammell: I mean, I just told you the things that I suck at though. As I said, the timeframes are just like.

Dave Jones: That part helps, right? I know what I need to work on. We can find individual people, I guess, to help you on specific things, but knowing what the specific things are is pretty much on you, I think.

Chris Gammell: You're right. Right. It's almost like you need like a, you know, you know what you really need is like a, you know, like in like epic magical quests, there's always like a, you've, you know, you kind of come upon the new master and then they teach you that thing. And then you go along your journey, then you find a new master. That's the kind of thing that I need, you know, like fantasy story.

Dave Jones: Well, in some industries, you've got absolutely no hope at all. I just sent you a link to Lewis Rossman's. I really liked this video rant that he did five days ago. It's titled the lower rungs on the economic ladder are gone. And it's basically him. It's a really good explanation of how he used to do things, how he built up his channel, built up his business. Well, he's built up his business basically. And how you can't really do that today. The same rungs on the ladder are not, are physically not there for someone to follow in his footsteps after all the hard work he's been done doing for a decade, trying to build all these tutorials to help people out, do the exact same business, start the same business he's on. He explains quite methodically how those lower rungs just aren't there anymore. You almost literally can't get into the business he's in now the same way that he did. It doesn't exist.

Chris Gammell: And he's talking about YouTube, not, not like repair. No, no, no.

Dave Jones: I know he's talking about the repair business, not the YouTube business. So he's talking about the repair business and a lot of people have followed in his foot. Yeah. And it's like, and the fact that, you know, you can't get, not only can't you get the parts anymore, but how they're locked in and all that sort of, but all the, all the equipment and everything else, it's something kind of like, yeah, there's, there's all these rungs that he stepped up, may, you know, you know, earn a bit of money, bought this tool and earned a bit more money, bought this, you know, these screens, bought them in bulk, stepped up and how that, those rungs just aren't there anymore. They've just, it's, it's complete rug pull for the entire repair industry. Yeah.

Chris Gammell: So I haven't watched this yet, but I mean, it is kind of that, that happens in a lot of industries, right. I've like the, the, the proven path, right. Yep. That's why I think to speak about education systems, present company included, you know, anytime someone says like, I will show you the path, really what they're saying is I will show you the path of the way I did it. Right.

Dave Jones: Which worked 10 years ago. That's right. Exactly.

Chris Gammell: They always need to take that with a grain of salt. Right. And like, you know, that's, I think that's even in traditional education, right. Like they like, oh, well you need to learn ABC and then D it's like, well, nobody cares about AB or C anymore. And I just need to go straight to D. Right. Like that. Yeah. So unfortunately in educational types scenarios, it's always a part of a problem is just like, you can only really say the path that has been in the past.

Dave Jones: And yeah. Will the AI, will the prompt engineering become that D and AB and C are gone. So many people hyper-focus on the next decade on D, on learning D that they give up AB and C. And then they realize in a decade's time, they've completely come a gutsy because they don't have skills in AB and C.

Chris Gammell: Yeah. I think the, it's important to always be flexible at the end of your journey. Right. Because I mean, I think about like the people that, I guess the people, some of the people that sign up for my course, right. A lot of them were like, yeah, you know, I, I went to school for electrical engineering. It was like, the money was supposed to be so good in that. And then like, you know, I graduated in 2000 and like the bottom fell out, you know, there was nothing there. And so then I went into IT eventually. Right. That was like a very common story that I heard.

Dave Jones: Oh, totally.

Chris Gammell: And then they're, they loved electronics though, like, because it's awesome. Of course. Right. I mean, like, and then they wanted to get back to it. Like that is a very common thing. So, but like, there was just, there was no there there.

Dave Jones: I, I, I did that once I realized this, uh, oh, this, all this networking stuff. When, when the networking boom took off, I thought I should probably get skills in this. So I actually went and did a course on, on. Oh, really? Yeah. Yeah. That was about God there. That was maybe late nineties, you know? So I, yeah, I, I have a diploma in networking or something it is, you know? Yeah.

Chris Gammell: So what made you decide not to pursue it further?

Dave Jones: Oh, that was just, I just thought it was a handy skill to have. I just kept doing engineering work. I didn't end up changing jobs, but I thought, you know, that's where the industry was headed. Oh yeah. You know?

Chris Gammell: Okay. Interesting.

Dave Jones: But nah, as, as it turns out, didn't really need those skills a lot. And that industry tanked, you know, all the networking jobs just started to tank fairly, fairly fast after that. So yeah.

Chris Gammell: I did hear someone talking about how fiber like installations are up again. Like there's been like a leap forward in fiber. I don't know if it's like trans-oceanic fiber, but like, oh, who was telling me about this? I think it was one of my former coworkers telling me that like the fiber industry has been roaring back in certain areas because of some big leap forward in terms of like networking speed.

Dave Jones: Okay. Well, it just seems old to me. Yeah.

Chris Gammell: I was surprised about it. Yeah. I know. Exactly. Exactly. Exactly. But also like, it does seem like almost incremental. Even if you go from like 10 gig to a hundred gig, like, yeah, that's a, you know, two decade improvement, but like, I don't need that. You know, do you?

Dave Jones: But the good thing about fiber is once it was once the fiber is installed, then you just, you don't change, you typically don't change the fiber. You just change the tech on both ends and magically you've up high order. Right. Yeah.

Chris Gammell: The physical layer is the same pretty much. Right. But yeah. Yeah.

Dave Jones: There you go. We, I, how did we get dragged back into bloody AI again? I don't know.

Chris Gammell: Well, you, you said it. You said it. Yeah. Did I? That one. Did I start it? Yeah, you did. Yeah, you did. But you carried it on. I did. Yeah. It's tough not to talk about these days. It's everywhere too. And it's, eh. It's.

Dave Jones: Yeah, I know. Yeah.

Chris Gammell: Yeah. What kind of electronics do you want to talk about? We'll see. Maybe we could start a timer.

Dave Jones: Oh, electronics.

Chris Gammell: I know, right? We could start a timer and see how long it takes us to get dragged back into AI. What do you think?

Dave Jones: Right. Okay. Yep.

Chris Gammell: All right. Did you see the hardware that has now been released that I made that I was telling you about?

Dave Jones: Oh, that video you sent me? Yeah. The video. The one on the trade show stand? Yeah. Yep. Or is there another update to that?

Chris Gammell: There's another one. Here is a link for you. This is. Maybe I didn't. I talked to. When I had my coworker Mike on the show, we were talking about a little bit. But it's just the reverse mount LED thing.

Dave Jones: This is the one we talked about the other week. This is the exact same link. We did.

Chris Gammell: Yeah, you're right. Yeah.

Dave Jones: Exact same link, dude. I guess I'm not making that much new hardware, Dave. Yeah. No.

Chris Gammell: See, this is why I need a learning plan to make more hardware. Right.

Dave Jones: Now, have we got any electronic stuff on our list here? How about those triple five timers, eh? What? Eh?

Chris Gammell: That's an old throwback right there. What's up with the triple five timers, Dave?

Dave Jones: Oh, boy.

Chris Gammell: Mm-hmm.

Dave Jones: No, somebody was talking on the forum the other day about an excellent link about they're talking about the fake chips on eBay. They bought a fake 07.4 op-amp.

Chris Gammell: Oh.

Dave Jones: Quad op-amp. It turned out to be an LM324. But it's rebadged as a, you know, 07.4. And they're quite different. Quite, quite different chips, you know?

Chris Gammell: Yeah, yeah. Like the TL 07.4 or something else?

Dave Jones: Yeah. Yeah. Just a normal TL. It had the TI branding or something on it, you know? Uh-huh. And yeah, it turned out to be. No, when they actually ran it through the basic tests, then, you know, they'll get in crossover distortion. They'll get in all, you know, other things that are-

Speaker ?: Yeah, get close to the rail.

Chris Gammell: The rails were like three volts on either end or whatever.

Speaker ?: Yeah, yeah, yeah.

Dave Jones: Exactly, yeah. So you can tell, you know, this is not a TL 07.4, you know?

Chris Gammell: It's crazy. It's crazy that like there's so much money in the gray market system that it's like it can pay for that.

Dave Jones: Yeah.

Chris Gammell: Yeah.

Dave Jones: It's crazy. It's crazy. And, you know, like the solution's obvious. Don't ever buy parts on eBay. Only buy them from the legitimate distributors. But then other people saying, oh, the legitimate distributors are too expensive.

Chris Gammell: Mm-hmm.

Dave Jones: You know, it's like, are they? Really? Well, we can buy onesies for, you know?

Chris Gammell: Right, right.

Dave Jones: Practically nothing.

Chris Gammell: Do you ever watch- I know you like a lot of like bad movies. Mm-hmm. Do you ever watch Super Troopers? You know, that comedy?

Dave Jones: Super Troopers.

Chris Gammell: No? Oh, add that one to your list. You'll like that one.

Dave Jones: You're talking about Starship Troopers?

Chris Gammell: Nope. Super Troopers. It's about highway cops in Vermont.

Dave Jones: Super Trooper is a ABBA album back in the 70s.

Dave Jones: Okay. Yeah.

Chris Gammell: One of the characters in that, he has a line that's like, desperation is a stinky cologne. And that's what I think about buying people, people buying chips off eBay. That's a stinky cologne. Right. A sign of desperation. Okay.

Dave Jones: Yeah, yeah. Yeah, but I, I don't know.

Chris Gammell: I mean, what else are you going to do, you know?

Dave Jones: It's just, I was thinking about doing, well, I, here's a video idea that I just had yesterday, which I might do. Hit me.

Chris Gammell: Yeah.

Dave Jones: I thought about what is, what is the best project I can make for 10 cents or a dollar.

Chris Gammell: For 10?

Dave Jones: 10 cents? Yeah. Yeah, because see, because I've, I've, you might be surprised, but I've done a video on the one cent regulator, right? Okay. Well, voltage regulators, one cent. I bet you I can get an op amp for a couple of cents.

Chris Gammell: Paduk or CH32? Yeah.

Dave Jones: I can get a three cent Paduk microcontroller, right? You know.

Chris Gammell: You count the PCB. Well, we kind of talked about this on the show. Remember we said the $100 prototype as well. Remember that? Right. That was more like shipping and all the other crap you'd have to do that too. Yeah, yeah, yeah.

Dave Jones: No, but I thought maybe if you, you know, a dollar bomb cost or something, you know? Yeah, sure, sure. The $1 bomb challenge, that could be like a contest, you know?

Chris Gammell: That could be great. Yeah.

Dave Jones: Yeah.

Chris Gammell: Okay.

Dave Jones: That'd be fantastic. Anyway, yeah, you know, you could, basically that would be, you would have to do that in reverse. You would have to, and I'm saying, I'm not saying that I haven't done this before, come up with actual ideas for projects based on cost and availability. You know, you start with a case. You see a gorgeous case and you go, wow, I wonder what I can do with that case. Oh, that's sexy. Sure. And then you come up with the idea. Same thing here, right? You go to, you know, you go to, you know, you go to, you know, you go to, you know, LCSC, for example, right? That'd be your go-to destination, right? Which is the Chinese equivalent to DigiKey and Mausa. You go there and then you sort by price. You say, okay, I need a one cent voltage regular. I need a one cent op amp. I need a three cent microcontroller, right? What other stuff can I get and what can I build, right? Okay. I've got 20 cents worth of parts. That leaves me 80 cents for my PCB. You know, what can I make for my dollar?

Chris Gammell: How many do you have to make too? I guess that's another thing, right?

Dave Jones: Well, you just have to maybe, you know, specify like a hundred volume or something so that you could get like, you know, and, and assuming that you handmade them all, you know, say there's no assembly cost and you just, you know, something like that. Yeah. Yeah. And maybe you take out postage, right? Just because that comes out in the wash, right? It's like.

Chris Gammell: Well, not if you're paying a dollar per. Yeah.

Dave Jones: But still, you're saying like, uh, if you were going to be extending this to a hundred million

Chris Gammell: units.

Dave Jones: Even a thousand units, right? Once you extend it to a thousand units, postage becomes absolutely trivial. Right.

Chris Gammell: I don't know, man. I've been shipping some stuff and it's, it gets up there, you know, like if you have like multiple, so like, how about this? I finally closed out, uh, the round of parts that I bought in March of 2022. They finally were all delivered, but they were delivered in about eight different batches. No, 12 different batches. Right. Yeah. It was like pretty insane. It was.

Dave Jones: So you paid like 30 bucks for each one or did the distributor absorb that cost?

Chris Gammell: They know they charge for each one. Yeah.

Dave Jones: Ooh. Ouch.

Chris Gammell: Yeah. Yeah. And so, I mean, it does add, it's not like, you know, in the long run, yeah, it's not a big deal. And I mean, at this point, honestly, I'm just happy to have parts. Yeah.

Dave Jones: But if you get them all from their one source and they're all in stock and you buy, you know, like a thousand of each, then postage isn't going to be a huge deal.

Chris Gammell: Right. Right. I'm just saying it can.

Dave Jones: Yeah. Of course it can. Impact things. Yeah. Anyway, I thought that'd be cool. So there you go. That's a great idea. I like it. That's great.

Chris Gammell: Yeah.

Dave Jones: Yep. The $1 project. Yeah.

Chris Gammell: Bring it. Bring it on. I found something on our list that I thought would be kind of interesting. Arm seeks to raise prices ahead of a hotly anticipated IPO.

Dave Jones: Hell no. This is not electronics. This is just.

Chris Gammell: This is, this is, this is electronics because, so I saw this actually, Ray Ozzy, former guest of the show, shared this on LinkedIn and had some good insight on it. And so it's an FT article. We'll link it in. But basically, Arm is, you know, the IP maker in many of the chips that we all know and love.

Dave Jones: Yes. Arm, Arm don't make chips for those who don't know. That's right.

Chris Gammell: That's right.

Dave Jones: They make the core. They make the IP core. Yeah. And other people license. They're purely a licensed company. So.

Chris Gammell: And they usually do it on a per chip basis, right? So you sell. Per part. A hundred million chips. It's, you know, some fractional cent per, you know, the IP. There's like upfront cost. Then usually it's an IP thing. And now they're saying, no, we're actually going to change how we charge for this sort of thing. And so Ray posted this and had.

Dave Jones: Oh, they've made a pivot in the way they, in their fundamental.

Chris Gammell: I think because they're trying to juice. I mean, they say they're trying to juice the numbers so that they look better when they IPO. If people don't remember, Arm was owned by Masasan and the, what the hell is that Japanese fund called?

Dave Jones: Yep.

Chris Gammell: Masasan is the, whatever it is. It's the, it's the monstrous fund that bought Arm and also owned WeWork. So really great. Oh, right.

Dave Jones: Yeah.

Chris Gammell: Yeah.

Dave Jones: Yeah. WeWork.

Chris Gammell: And then they were supposed to get bought, I think by NVIDIA a while back.

Dave Jones: Oh yeah, that's right. And that was stopped. That was, yeah.

Chris Gammell: And now, so then they said, okay, well, we got blocked there. Now we're going to go IPO sometime this year. And now, so they're trying to make that look really good so that they get lots of dollars and everybody's happy, happy, whatever.

Dave Jones: Right.

Chris Gammell: But this fundamentally changes things, right? So now it's basically saying it's not about how many chips you sell and we'll charge you per, you know, per chip. It's now big chunk of money and you got to, you're on your own. Right.

Dave Jones: Well, maybe this is a response to the HISC-V thing. Everyone's jumping on the HISC-V bandwagon, right?

Chris Gammell: Well, maybe, but they still own 95% of the market. So that's...

Dave Jones: Yeah. But maybe they can see the curve, that hockey stick curve starting to...

Chris Gammell: I think they're nervous. But this isn't actually a move that you would make if you were nervous about it. You would cut prices. You wouldn't juice prices. So I do think that this is going to have some impacts generally, but I think RISC-V will get accelerated as a result of this, not the other way around. And I couldn't be happier.

Dave Jones: Well, imagine if you were an existing ARM customer and you were on a royalty basis and you paid per how many chips you use and all of a sudden they come along to you, oh, sorry, we're changing the agreement and now we want 50 million bucks cash up front, but we'll give you the chips for free. It's like, where are you going to get that coin from?

Chris Gammell: Exactly. Right? Yeah. Yeah. It changes everything. It changes...

Dave Jones: That's just like, yeah.

Chris Gammell: Yeah. It's pulling the rug out from under. So, and I think, you know, a lot of the times, you know, you and I are used to like tiny, teeny tiny microcontroller type of things. That's not where they're making their money. They're making their money on smartphones, you know, the A98, whatever the hell these models, monomers are these days that are in like the Snapdragon, all those other crazy ones that are out there. So like, that's really what this is going towards. And I'm sure that's the big one. But like, I guarantee that, you know, Qualcomm, as much as, you know, they're not talking about it yet. There's just, there's RISC-5 that is just getting infiltrated to everything here because like this is a cost center for all of them. So it's just like, I think this is just going to be a huge accelerator of that sort of thing.

Dave Jones: Right.

Chris Gammell: Especially in China. Right. I mean, if you think about like the stuff being made there where there's, so like they said, the average price for a smartphone chip is $40 for Qualcomm, 17 for MediaTek and $6 for Unisoc. And so, you know, if you don't think Unisoc is trying to save some coin of that $6, like they, I mean, they are, right? Like as you go down in margin, you're trying to save more money. So why wouldn't you start developing your own IP or even buying just cheaper IP? You know, like remember RISC-5 is a instruction set architecture. It's not an actual processor. However, many of the tools then can be developed or to work with your, your stuff, the core that you develop. So you could still buy a core that might be cheaper, you know, that still uses the same instruction sets, but it might be cheap, way cheaper than an ARM core, but then you could still have some tooling around it because of the stuff that's out there. So we'll see how it goes.

Dave Jones: There you go. Wow.

Chris Gammell: Yeah. Thanks to Ray for posting that. That's an interesting piece.

Dave Jones: There's going to be a lot of butthurt people soon, I think.

Chris Gammell: That Ray Ozzy who helped invent Azure and a bunch of other things. He's going places. He knows his stuff.

Dave Jones: All right. We have no other electronics related stuff on our list.

Chris Gammell: No, there's other stuff on here. There must be, there must be Dave. Come on.

Dave Jones: Will, will a bridge rectifier work after it's been laser decap, decapulated? Is that even a word? No, I'm just inventing words. That's, that's kind of cool. I need to know the answer. Will it actually work? Oh, this is a great time-lapse video of a chip being lasered. Oh.

Chris Gammell: Is it like the one where they kind of raster the top of it off basically and you get down to the, just to the silicon?

Dave Jones: Yeah. This is what they're doing here on this, on this video. Wow. Oh, you can see the individual connections and then, oh man, I need to do this. I need a laser. Maybe I should take up the sponsor, the video sponsorships of all those companies offering me a laser.

Chris Gammell: Oh, really?

Dave Jones: Cutter. Oh yeah. Yeah.

Chris Gammell: Laser cutter is a little different just so you know. Why? Well. What do I need? A laser engraver? Laser cutter? I think you need it to be more dialed in. I don't know, actually. This is like laser ablation, I think, when they're taking the top off of a chip, like the polymide off. I think it's like ablation.

Dave Jones: Sure, but it's just, it's just dialing, dialing in the right power, right? I mean. Yeah.

Chris Gammell: You know what? I wouldn't actually know where to start with that sort of stuff, so I'll just, I'll stop talking about it. Yeah. I don't, I don't actually know.

Dave Jones: Yeah. Anyway, that's interesting. I've never seen inside a bridge rectifier, like as in a large four pin flat pack bridge rectifier before. That's pretty cool.

Chris Gammell: Yeah. I wonder if that was in, I have my open circuits thing here. I wonder if they, they put that in there.

Dave Jones: There you go.

Chris Gammell: Yeah.

Dave Jones: I like it. I need to start playing with lasers.

Chris Gammell: Yeah. How's your, how is your, your tooling set up around the lab? Are you?

Dave Jones: No, it's, I know I need to. Pick and place was a no. No. Yeah. No, unfortunately it's, um, no, the pick and place it, I, I just had a lead on a good one, but it wouldn't physically fit through the door. So yeah. That's rough. I had to. Yeah. Yep. No. Yeah.

Chris Gammell: So on the laser side of things, you know, so I, I did a bunch of laser cutting for the conference. Like I made like backing plates, basically like things that I mounted demos to. I think I talked about it last time.

Dave Jones: Right.

Chris Gammell: And like, if you need anything like that, it's not, I personally think it's not worth it to get a laser cutter unless you like know you're going to be doing that all day, every day. Yeah. I think it's just like there's services out there where you can just pay for it much cheaper. But. Sure. Is there decapping services?

Dave Jones: Well, the same thing with a 3d printer these days, you could argue that, you know, if you're in the U S is, I don't think it's that. I don't even think, is there a company in Australia that'll 3d print locally and send it to you for Nix?

Chris Gammell: Oh, a hundred percent. Yes. Yeah. I guarantee there is. Yeah. Okay.

Dave Jones: Yeah. Search that.

Chris Gammell: Yeah. I, I, I think it's a, I think there's a little bit more in guess and check. Well, at least how I do 3d modeling, there's a little bit more guess and check. So like up to the point where I like, I know the fit's going to be perfect.

Dave Jones: Yep.

Chris Gammell: Whereas like if I'm doing laser cutting stuff is often like artistic. I mean, there's still like form and fit type of stuff.

Dave Jones: Right.

Chris Gammell: But.

Dave Jones: Oh yeah. There you go. Digital manufacturing accessible from five Aussie bucks. There you go. From, from five Aussie bucks. I'll. Zelda 3d will. From one to a thousand parts. Yeah.

Chris Gammell: Sounds like you should be ordering some this afternoon. That's what I'm thinking.

Dave Jones: One to three days. Yeah. Yeah. You should. Sweet.

Chris Gammell: Next step is free CAD. Thomas Sirlander did a review of key CAD or not key CAD free CAD. I don't know if you know his channel, but. I don't. It was not a glowing recommendation, but it was a, it was a review from a major YouTuber in the 3d space. So. Okay. That's something. That's something that doesn't happen that often. It's still incredibly rough. The. Using it. So.

Dave Jones: Is it just, have we, have we reached the point where 3d printer and laser cutter, you're almost, you know, it's almost pointless having your own because the, the services out there are so cheap and you can't beat, you know, this hundred thousand dollar printer printing your part versus your little dinky $300 eBay jobby printing it.

Chris Gammell: Like there's no contest. Like I said, I think the only thing is fit, like initial fit and finish. And like, and like holding a part in your hand. Like there's still no.

Dave Jones: Physically. Yeah. Yeah.

Chris Gammell: There's still no like, you know, like we talked to Jerry Ellsworth about like the promise of like holding a digital prototype and like an AR style prototype in your hand like that.

Dave Jones: Right.

Chris Gammell: That's like the promise of AR that never really landed, but there's still no like good way to get a sense of scale. Like, you know, like when you've done a circuit board and then you get it back from the manufacturing, you're like, oh my God, this is so much smaller than I thought it was going to be. Yeah. Yeah.

Dave Jones: Right.

Chris Gammell: Yeah. Like, so like that same effect, but like 3D print, like 3D printing, like to be able to even throw a rough model into my printer and then be like, oh, that was way too small. Very useful. Yeah. Yeah.

Dave Jones: Yeah. It's useful. But, you know, but are you like, is it like good enough to use, you know, cause if you want to like use it.

Chris Gammell: I think, I think then what you do is you use the rough model, right? You get your fit and finish, whatever you get. And then you order it.

Dave Jones: So you can have that in an hour or two. That's right. Your rough model. And then in a day or two, you'll have your.

Chris Gammell: That's right.

Dave Jones: Real part. You know?

Chris Gammell: Yep. Yeah. And you can iterate on it too. Right. So I can get probably 10 iterations in a day versus, you know, any kind of shipping back and forth, that sort of thing.

Dave Jones: Still quite useful.

Chris Gammell: I think, I think there still is some value, but not, there's no value in me having a hundred thousand dollar printer in my house. Right. Because to get the cash back from that, you need to be able to like doing it.

Dave Jones: Quite a step up in level. It's like the pick and place machine, right? You've got to step up to the $20,000 machine before it's like, it's really quite good.

Chris Gammell: Have you seen those assist machines for pick and place? So it's like still hand pick and place, but it like steadies your hand and it's got vacuum pickup and those sorts of things.

Dave Jones: They're a manual automated. They call it like a semi manual pick and place machine. And they've got like arms and they've got trays and they've got like a suction thing and you, you know, and yeah, yeah, yeah.

Chris Gammell: Yeah, exactly. Yeah.

Dave Jones: They can get really complicated actually. They can get really complex. You can pay, you know, a fair bit of coin for those.

Chris Gammell: Sure. And then yeah, like foot switches on the pressure thing. Like that's what I think about is like a in between kind of thing that actually, you could still be, you know, like it's not worth it. Like where is that crossover point for a pick and place? You know, it's like 10, a hundred units, 50 units.

Dave Jones: Yeah, exactly. It's still that very narrow window. There's still that narrow operational viability window, you know?

Chris Gammell: Yeah. But I think it's also the complexity of the board, you know?

Dave Jones: And it's like, and it's the same with the 3d printer. Like, yeah, I've got one, but if I want to use it, it's not just set up, ready to go. Like I'd have to, oh God, I've got to get it. And then new software doesn't work with the old firmware and oh God, I got to like, you know, it takes me all day just to get the thing bloody operational again to do my one print.

Chris Gammell: I have taken care of that by buying a brand that is pretty bulletproof, but I know the feeling. I felt, I felt that exact feeling with my old printer and now I'm past that.

Dave Jones: So I think I'll just take one of the companies that have offered me a 3d printer. I'll just do a review on my second channel. Then I'll have a new 3d printer.

Chris Gammell: Buy it, man. It's nothing better than buying one that's bulletproof. Yeah. I haven't, I haven't.

Dave Jones: Well, they've offered them to me for free.

Chris Gammell: So, you know, I know, but those aren't going to be bulletproof. That's what I'm saying. You know, like community supported bulletproof so that you don't have to think about it. What is your brand then? I have a Prusa Mini. Prusa Mini. That's Prusa Mini has been flawless. It's been no, no thought process involved. It's just been, it's an actual tool. It is a joy, a joy to use. And it's an open source hardware company. So extra bonus.

Dave Jones: Right.

Chris Gammell: Yeah.

Dave Jones: I'm just checking if they've contacted me. Prusa haven't. No, but you know, I've got cruelty just, is that how you pronounce it?

Chris Gammell: That's right. Yeah. Yeah.

Dave Jones: They contacted me, they contacted me the other day. You know, it's like, you know, what is wrong with that one? Right. Nothing. Right.

Chris Gammell: There's just a lot of fiddling. You know, like that's what I looked, I looked at those too. I don't know. I don't know why.

Dave Jones: Why is the Prusa just magic?

Chris Gammell: I don't know.

Dave Jones: I bet you I can find people that hate the Prusa and you go, oh no, he gave me all sorts of problems. Betcha.

Chris Gammell: Okay. Okay. Fine. Yeah. That's fine. Yeah. I mean, you're right. You're right. Fine. Go with the other one. I just think it's, you know, buying hands off tools is a joy, right? I mean, that's, I think you can at least agree with that statement.

Dave Jones: Right. Well, there you go. The New York Times seems to agree with you. The Prusa Mini Plus offers the best overall 3D printing experience thanks to its combination of print quality, reliability, and desktop appropriate use. That's for 2023. Now, nah, I am looking at, right? I am looking at this Prusa Mini, right? I am looking at this damn thing and I'm going, no way, that is reliable. I know. It doesn't look good, right? That just looks like a thrown together. The base of it is 3D printed.

Chris Gammell: I know. I know. It's insane. It's insane.

Dave Jones: Bullshit. You can't tell me that is reliable.

Chris Gammell: It is printed hundreds of prints for me and I've calibrated it the first time.

Dave Jones: That looks like the dodgiest 3D printer I've seen in the last five years. Seriously.

Chris Gammell: Reach out to Prusa. Get one. You know, be like, I'm a skeptic and I want to have one of your printers. It's so easy. 100% endorsement from here. I hope they send me one. Oh, man. If only I had any sway with anyone, Dave.

Dave Jones: That just looks like the dinkiest 3D printer.

Chris Gammell: Two things in life. I need a coach that can help guide me through learning more. I need another Prusa printer. I'm going to Prague in June, so maybe I can go visit them.

Dave Jones: No, look. The New York Times article says flaws, but they're not deal breakers. No, sorry. What are the flaws? I'm not.

Chris Gammell: It's fine, Dave. You don't have to get one. You don't print that often. Here's the thing. I tell people this all the time. The thing that you need to do if you are interested in 3D printing in the space, you don't have to get good at the printing part. Assume, you know, there are other good printers out there too, right? Right. Just let's leave that whole conversation behind. There are other good printers. Once it's dialed in, you should never have to touch it again, hopefully.

Dave Jones: Yeah, right.

Chris Gammell: Then what? You're just printing other people's designs? No, you need to like, if you're using a 3D printer in any like, you need to get good at modeling. And that part is frigging hard. Yes. It is a lot harder in FreeCAD than other ones. Like, you know, I love Fusion 360. I loved it. I don't want to use it anymore, but it's very good software. And the other ones are too. But you need to get good at that. And that is the skill you have to build. Not the dialing of 3D printers. That's not how it used to be. You know, that's... Totally agree. The MakerBot days are over, thankfully. Yeah. Yeah.

Dave Jones: Oh, yeah. I can remember when I got a rigid support arm for my MakerBot. Oh, God. It made a world of difference. Big days. Big days.

Dave Jones: Yeah. Oh, it's heaven. You know. Oh, God. Yep. Anyway.

Chris Gammell: Yeah. And then you go and print another Benchy, right? Right.

Dave Jones: Print another Yoda head, you know? Yeah, exactly. Yep. Yep. Oh, God. Those were the days. Anyway. All right. There you go. Our amp hours up.

Chris Gammell: Mostly taken up with AI. Sorry. Dave, we made it. We made it.

Dave Jones: AI and 3D printing, for God's sake.

Chris Gammell: I know. What are we? We should... Pathetic. You know what we should do? Change our name. Next time, we should be on ChatGPT Live. We can just be like, what should we talk about next? Yeah, yeah.

Dave Jones: And we can...

Chris Gammell: And be like...

Dave Jones: We can get ChatGPT to suggest things.

Chris Gammell: That's right. Please suggest it.

Dave Jones: We are an electronics-focused podcast. Please suggest us things to talk about.

Dave Jones: Like, everything... Isn't everything just going to become the same?

Chris Gammell: Oh, man. We are talking about it again. Isn't everything just going to be the same?

Dave Jones: Isn't...

Chris Gammell: Oh, Dave. Oh, boy. We're headed into a weird time. We're headed into a weird time.

Dave Jones: Oh, boy. All right.

Chris Gammell: All right.

Dave Jones: That's it.

Chris Gammell: Talk to you soon.

Dave Jones: Catch you next time.

Speaker ?: Bye. Bye. Thank you.