AI Can Build It. Taste Is What Makes It Valuable.

[00:00] Dave Pengelley: This is our this [00:00] Richard Webbe: Okay. [00:00] Dave Pengelley: is our lead in banter that um as as we go live, Richard is about to tell Matt about a Christmas ham story regarding phraseologies and sayings that we don't know the origins of. [00:12] Richard Webbe: We are. So for the audience, we were all talking about our colloquialisms and where they come from. And Dave, I was saying, has perfect recall for all the phrases that we hear in popular culture that come from TV shows and can always tell me which TV show it comes from. And Matt was just saying, good morning. I I know that's Saying, but I don't know where it comes from. And that led me to the explanation of the Christmas ham story. And this is a really good one, right? So what happens is, you know, the girl's with their mum, and mum's got the ham. She's going to put it in the oven. But before she puts it in the oven, she cuts the end off. The daughter quietly observes this and says, Okay, flashwater generation. There's the daughter with her family sitting there, and she's cutting the end of the ham off and then putting it in the oven. And her daughter goes, Mum. Why do you cut the end of the ham off? And she goes, I don't know. It's something we've just always done. So she rings her mum and says, Mum, why do we always cut the end of the ham off? She goes, Because it never fitted in my oven. [01:15] Richard Webbe: So while everyone thinks it's a cooking issue, it's actually the old ovens were quite small because today we have these huge ovens or whatever, you can freaking chuck a cow in there if you want and you're fine. So um, and that's what I find with the colloquialisms, they they become Popular culture, they become, as the poet laureates would say, into our modern language, and we use it all the time, and often we don't know why. And examples of that are: I'm stuck on the threshold. Well, where did the word threshold come from? Well, in the old days, the floor of a house had no carpet, and they used the threshings from the wheat when they were farming to collect all the rubbish, and then they'd throw the threshings out. That was their version of vacuuming. And the threshold was that bit of wood they put at the bottom of the door. So the wind didn't blow all the threshold out or into the house. [02:02] Dave Pengelley: Well, it [02:03] Richard Webbe: But [02:03] Dave Pengelley: holds [02:03] Richard Webbe: of [02:03] Dave Pengelley: the [02:03] Richard Webbe: course, [02:03] Dave Pengelley: threshold in the house. [02:04] Richard Webbe: exactly [02:04] Dave Pengelley: It's the threshold. [02:05] Richard Webbe: the thresh threshold. And so hence the threshold you know on the threshold, you're on the border of going into something or out of something. I am on the threshold. I could go like this for hours. So [02:15] Dave Pengelley: I [02:15] Richard Webbe: God [02:15] Dave Pengelley: I was [02:16] Richard Webbe: could. [02:16] Dave Pengelley: watching this random documentary about building things because I was, you know, home alone between things. I don't know, I just was like lost, and I was just skimming through Amazon or whatever and found there's like the house stuff's built documentary series. I was like, that looks interesting. I think I think I got me because the first episode had um Gray Nichols cricket bats. I was like, ooh, I'm in. I'm in. Tell me about how they make uh English willow. But one of the ones, um Doc Martin's boots, and he was going through the boot factory and referencing all these terms that we use, like as modern, like in the in the parlance of our times, that come from boot making. Even things like sciving off. So skiving off, people are familiar with means you know, they're slacking off, they're taking a rest, they're not getting the job done. That is a term in from boot making where when the boots on the sole of And then I think they trim the edge off. That's the the sky and they're skive taking the scive off. And that is the only job in the whole bootmaking factory you sit down for. So where's that person? Are they sciving off? [03:10] Matt: Yeah. [03:10] Dave Pengelley: And that means they weren't on their feet. They were doing this one particular role in boot making. And that became, you know, that's the person having a rest because they're they're sitting down. And it was like crazy as he went through. I can't even remember them all, but they're like five, like a half a dozen different terms that he referenced as part of the boot making process that are these sort of common day terms that we talk about. [03:30] Richard Webbe: I've got one more good one for you. I watched a new movie, which is a classic kids' action movie, which was a bit silly at times, but enjoyable to watch with my young son, and it's Matchbox. [03:41] Dave Pengelley: I we watched that the other day with John [03:44] Richard Webbe: And what [03:44] Dave Pengelley: Cena. [03:44] Richard Webbe: was the story we learned off that day? The guy's talking, and he said, the man who first made matchbox toys, as in the little tiny cars, his kids were only allowed to take a toy to school that could fit in a matchbox. So he used to carve little cars. That's where matchbox toy cars come from. [04:06] Dave Pengelley: Phenomenal. [04:07] Richard Webbe: Matt, [04:07] Matt: I [04:08] Richard Webbe: you [04:08] Matt: have [04:08] Richard Webbe: look [04:08] Matt: one. [04:08] Richard Webbe: quietly amused or impressed. I'm hoping impressed. [04:11] Matt: I can I can contribute one, just uh like a random one. Um my my partner's mum got a new car and uh my partner said the other day in a message, how's the new whip? And you know, like have you guys heard the term whip to do with cars? Like like [04:28] Dave Pengelley: Yeah, [04:28] Matt: no [04:28] Dave Pengelley: that's [04:29] Matt: no [04:29] Dave Pengelley: where I [04:29] Matt: no like [04:29] Dave Pengelley: was. [04:29] Matt: like Whip in a WHIP. [04:32] Richard Webbe: Yep. [04:32] Matt: Um that that is a is back to the horse-drawn carriage days because you get a new a new carriage comes with a new whip, has your new whip. [04:41] Richard Webbe: Ah. [04:41] Matt: Like and yeah, it it it I I thought it was a modern term. I thought it was to do with like Japanese drift culture or something, like whipping around the corners. [04:50] Richard Webbe: Well, years from now, people will be asking what does AI stand for? And and they'll be asking what does SI stand for? Because we've learned during the week that's super intelligence, thanks to our friends. At Elon Musk and Mr. Trump and others. I love it. [05:03] Matt: Yeah, yeah, they've actually fully rebranded now. Um all their team is now talking about SpaceX SI. Um yeah, it's crazy. [05:11] Dave Pengelley: But what what do you mean? What is what is SI? Who's who's jumping [05:14] Richard Webbe: So [05:15] Dave Pengelley: on that [05:15] Richard Webbe: forth, [05:15] Dave Pengelley: bandwagon, [05:15] Richard Webbe: that's right. [05:16] Dave Pengelley: right? [05:16] Richard Webbe: I [05:16] Dave Pengelley: All [05:16] Richard Webbe: did [05:17] Dave Pengelley: right, [05:17] Richard Webbe: see [05:17] Dave Pengelley: let's [05:17] Richard Webbe: Benioff did that during [05:18] Dave Pengelley: uh [05:18] Richard Webbe: the [05:19] Dave Pengelley: let's [05:19] Richard Webbe: week. [05:19] Dave Pengelley: let's do it let's do a thing. It's fun. [05:33] Dave Pengelley: Funny funny scene Benioff just jump on the latest bandwagon. Obviously, he is my former employer. Um, but like he's known for being in with the a different crowd than sort of the Trump crowd, right? Like he's Mr. Davos World Economic Forum guy, who Trump's not a big fan of. And so to see Benioff falling in line with the SI thing is like, ah, the winds are shifting. The power is moving away from Davos back to Washington. And Mark is following along like a like a good CEO. Very interesting. [06:06] Matt: Just comes to mind like the the for D chess thing again. Like, you know, I feel like these guys they all know what they're doing. And [06:13] Dave Pengelley: Oh. [06:14] Matt: everything is [06:14] Dave Pengelley: Mark is is a premier salesman. He knows exactly what he's doing. He is no one's fool. Um, but the I mean it the I put the out the joke that Salesforce, they rebrand every year anyway. They just had their annual conference where they announced AI Force or whatever [06:29] Matt: a good idea. very [06:29] Dave Pengelley: it was. like a month ago [06:32] Matt: Yeah. [06:32] Dave Pengelley: they when they had Dreamforce and sort of a month later it's SI Force but yeah I um alright I'm [06:38] Richard Webbe: I [06:38] Dave Pengelley: shocked it's [06:38] Richard Webbe: did [06:38] Dave Pengelley: not the [06:38] Richard Webbe: see [06:38] Dave Pengelley: lightning [06:39] Richard Webbe: that [06:39] Dave Pengelley: Einstein [06:39] Richard Webbe: and laugh. [06:39] Dave Pengelley: 360 super agent intelligence force one cloud like it's just the little monikers they just put on their product for anyone who's ever been in the Salesforce ecosystem it's [06:49] Richard Webbe: Another [06:49] Dave Pengelley: pretty [06:49] Richard Webbe: example [06:49] Dave Pengelley: wild [06:49] Richard Webbe: of what I call hijack marketing. Attach your product to the latest words in the marketplace and pinch that marketing dollar to use for yourself. [07:00] Matt: Like when he made his school group, he he just called it Claude Code Architects. And you know, everyone that went to school searching for Claude Code, they were like, Oh, this group looks good. [07:11] Richard Webbe: Wow, that's powerful. That is very powerful. So I'm looking towards that fearless anchor man here. What are we up to today, Dave? [07:21] Dave Pengelley: Oh, I'm uh I was just saying a man, I'm really full. It's bad to eat too much before a podcast because then you can't think it. I've got a bit of food cramping going on. Um that's what everyone wants to hear about. Dave's food camp. No, we're not talking about food and diet and tracking apps this week. There is so much going on, um, news, wins, etc. Um, just general state of the industry. I mean, before we jump into news stuff, like let's talk about the SI force thing and the world SI thing. What do you think is behind this whole rebranding of AI to SI? Is this an anti-China thing? Is this a like meta-game? Gives their their mascot a little cute little fluffy butt, and so people are meant to think it's less i in like threatening because you know the rumors around AI and Skynet have been distributed for decades. Let's call it something different, and then people won't be scared of it. Like what what do you guys think? [08:15] Matt: I think that's it. I genuinely think that AI has like bad connotations in the market. Even the recent post from the Xbox CEO, she I don't have the link prepared, sorry, but she was talking about how everyone's been mashing together all these AAA games and you know, completely just blending source code for stuff that shouldn't exist. [08:35] Dave Pengelley: I saw Call of Duty Tony Hawk or something the other day where [08:38] Matt: Yeah. [08:38] Dave Pengelley: someone's running around in fatigues riding a skateboard. [08:41] Matt: Yeah. And then that's that same thing. Imagine both of those in Minecraft. Like I saw that one this morning. Um and and she basically said that, yeah, look, let's let's embrace it. We've made Xbox labs and blah, blah, blah. Let's do some cool things. And nowhere in that whole post did she mention like AI. And there were so many comments p s saying like, just say the thing. Like we know that you're admitting that the thing exists. [09:07] Richard Webbe: It's [09:07] Matt: Just [09:07] Richard Webbe: a dirty [09:07] Matt: name [09:08] Richard Webbe: word. [09:08] Matt: it. [09:08] Richard Webbe: It's a dirty [09:09] Matt: Yeah. [09:09] Richard Webbe: word, man. You might have seen my article I wrote last week on this point about everyone complaining about data centers and AI, and and and I I put I intentionally put in the article. So while you're complaining about new data centers and AI, and I still don't understand why people complain about data centers compared to other things, um, you know, it could be. Right now, an algorithm in there is sorting out the cancer that you're going to get in four years' time and have the cure ready for you. People are so down on the technological advancements at the moment. And I I have to say it, you know, love him or hate him. I thought the funniest thing was, well, our governments, governments all over the world, are talking down AI as a point of shame or ushame, or oh my God, this is dangerous. Donald Trump stands up and goes, We're going accelerating, we're going full AI. We're going SI and we're gonna keep going. Now, [10:03] Dave Pengelley: No [10:04] Richard Webbe: if you [10:04] Dave Pengelley: never [10:04] Richard Webbe: think [10:04] Dave Pengelley: go full retard. [10:05] Richard Webbe: Yeah, we're gonna go full red down. I was avoiding using that word. But the funny thing is, people complain about cars, they complain about medicine, they complain about vaccines, they complain about safety belts, steering wheels. Yeah, people um uh I think it was Isaac Asimov and Alvin Toffler. If you want to look up those two, uh I think Alvin Toffler wrote the book Future Shock, and it's all about How humans need to catch up with the technology around us for us to accept it. And I think that's a marketing mistake that our technology industry has gotten wrong. We run ahead too fast to make money without bringing everyone with us and go, hey, this is going to be a good thing. This is going to be a great thing. And so one of the headlines that I spotted during the week, and a lot, and I was just talking to an AI consultant about 10 minutes ago, not you two, but uh someone else, and it's the amount of money that people are not making in the AI space. The AI tag on the end of their badge or their job title, but no one's making any money. No one's actually profitable, and no one's actually realizing what they want to realize or what they want to get. It's almost like the telcos, and I wrote an article about that this week. The telcos enable all of this with their pipes connecting everything, but they're not making any money out of this. And they're trying to work out how to they miss the point with the online world. They monetize being able to sell mobile phones and that. But in terms of the data and the actual big value. value that exchanged across a network with a high connecting mobile phone, the telcos are still out to out to lunch, they're still missing out on the billions or trillions they should be getting. It was all taken by Mr. Benioff and you know our mates at Facebook and Meta and a few other places. So there's an amazing cultural story happening as you you just said that relates to advertising, acceptance and where we go from here and all [11:53] Dave Pengelley: When when I was at Telstra, they started trying to get on this bandwagon by selling what they were calling T Suite, which was their sort of managed cloud. Applications. This was 15, 16 years ago before people knew what the cloud was. Software as a service was still a relatively new thing. I don't think Netflix was even out yet, not in its current streaming form. Um and they were getting the Office 365 stuff from Microsoft, and they were like reselling it and packaging it out under their T Suite offering. But I was a I was a sales rep then, and it was really hard because people like they all they all have they all had their small business server 2003 little HP box in the corner [12:30] Richard Webbe: Yes. [12:30] Dave Pengelley: or in Iraq that was running their exchange. They were like, it's wild now to think that people running their own mouse servers on their ADSL connections in their offices and there were there were thousands of them all across Australia. People running their own exchange servers. [12:44] Richard Webbe: Very true. [12:45] Dave Pengelley: Just [12:45] Richard Webbe: And [12:45] Dave Pengelley: wild. [12:45] Richard Webbe: the shadow IT and all of it. So my old boss, Ben Horowitz, of course, uh now um heading up with Andreessen um Andreessen Horowitz, the venture capital private equity firm. I think it was Ben if he ever sees this and says, Richard, I didn't do that. It's a great Guy, by the way, uh and a best selling author. Um, he he he wrote a white paper called The Zero, yeah, I was gonna hold my book up then, but I'm not a best-selling author yet. Don't laugh, you two. Sorry, I have to do it now. Anyway, he wrote a great little uh he grew uh wrote a great little um uh white paper that I loved called the Zero Latency Data Center. And I reckon someone needs to do something like this for AI, and I I might even try and write it myself. So what happens is when I was at BHP, IT and to all my BHP IT execs we caught up for our mid year Christmas dinner a few months ago. We would have someone in a mine in Northwestern Australia and say, We need a server, we need HR, we need spreadsheet with the Microsoft thing that Dave just spoke about, we're processing it. We need all of that here now. And I, you know, you say to everyone, how long do you think it took the IT department to get something out there? Six to nine months? Because you've got to do security and standard operating environment and all of that sort of stuff. So what are the Do they're losing a million bucks a day when they're not digging in the ground? So they just go down to the Harvey Norman down the road and do their shadow IT, which was a massive danger from security and reliability, safety, all that sort of stuff. So avoiding the term necessity is the mother of invention, right? [14:20] Richard Webbe: Ben wrote this zero latency data center, and he basically said it is much more cost-effective, culturally acceptable, to just have a bunch of servers sitting in the corner of. Of the room ready to go. So one of your divisions rings up and says, I need HR payroll safety, you know, onboarding, and you just grab that server and put it on a truck and ship it off to them. And of course, that led to the depth of isn't cloud clever. The amount of server that we can have running, not even being used, but turn it on or off straight away, makes the world a lot more efficient. So, how are we going to get there with AI? How are we going to give AI this ubiquitous access that's not threatening and Culturally accepted, and people understand the value of means. [15:05] Matt: Kinda depends what you mean as far as like what you want it to do, like what is the application. Like [15:12] Richard Webbe: I [15:12] Matt: like [15:13] Richard Webbe: think [15:13] Matt: Dave [15:13] Richard Webbe: the killer [15:13] Matt: mentioned, [15:13] Richard Webbe: application, the killer application is automation between lists of data where you can get that data together and dispose of it, or discern it, or divest it into the systems that need to know or the people that need to know. [15:29] Matt: I think. So like ingestion, translation. Because it's quite safe to do just read-only tooling. You know, like you could have all of your data accessible across all of your different services just through read-only tooling that you create. And then it just depends on where that where that information stream goes. So if you're sending that through your normal API channels or if you have your server box in the corner running your small LLM of choice, you know, that that kind of thing would be pretty accessible now. And but if that's useful to anyone, like that sort of thing is still yet to be sort of seen. [16:13] Richard Webbe: And this, but this is the point, the core that you wrote that you introduced me to that I've been using, and I can tell you now it's it's useful and helpful, and I get the full UI understanding of it. Um, people don't have the connection. All they hear is AI and they panic and run around and say I'm gonna lose my job. [16:29] Matt: Yeah. [16:29] Richard Webbe: No one is Really promoting that we will save lives, save fires, save people, send people. We're going to do everything with this stuff. And as the economic confusion and the cultural confusion sort of levels out, it gets in the way. I mean, our government's talking about holding slowing down on AI. Oh my God, did you see what's our mate who owns NVIDIA? He launched AI Europe [16:54] Matt: Jensen. [16:54] Richard Webbe: yesterday or this morning. Did you see that? [16:57] Matt: No, I actually didn't [16:58] Richard Webbe: It was [16:58] Matt: catch [16:58] Richard Webbe: on the [16:59] Matt: that. [16:59] Richard Webbe: socials, he said. We are AI Europe, and what they're going to do on the very point you're making, they're going to set up what they call AI factories. So if we think of AI as a generic automation tool that connects all of the data sources that we need to make use, he's going to set them up all across Europe. [17:21] Richard Webbe: Thank you. And he had a glass. If you look up the video, he had a glass of scotch in his hand when he was talking about it. It was very funny. And this is you know, here's the Australian government talking about slowing down AI and we need more controls and more. Oh my God, by the time we please our own government, Europe and the US will so far be so far ahead of us, we will voluntarily become a third world country. [17:44] Matt: Hmm, it's already pretty bad to uh operate a business here. [17:50] Richard Webbe: Exactly. I mean, exactly. Sorry, Dave, I know you have something to say, but you went blurry and Max headroom me there for a second. [18:00] Richard Webbe: Do [18:00] Dave Pengelley: Yeah, [18:00] Richard Webbe: you think [18:00] Dave Pengelley: and and I'm still my computer's strong. I can [18:04] Richard Webbe: we'll keep [18:05] Dave Pengelley: Hello? [18:05] Richard Webbe: talking [18:08] Matt: We can [18:08] Dave Pengelley: What? [18:08] Matt: eat fine, [18:09] Dave Pengelley: I'm [18:09] Matt: yeah, just [18:09] Dave Pengelley: here. [18:09] Matt: visually it's a bit weird. [18:11] Richard Webbe: about it? [18:13] Dave Pengelley: Yeah. Um [18:15] Richard Webbe: Sorry. [18:15] Dave Pengelley: the web that website I brought up with Jensen's thing was hacking my computer and slowing everything down. As soon as I close that tab, it was better. Weird. Weird new side. [18:22] Matt: You [18:22] Dave Pengelley: Too many [18:22] Matt: never [18:23] Dave Pengelley: ads, [18:23] Matt: know. [18:23] Dave Pengelley: I think, on that site. Way too many ads on that site. Um yeah, I mean uh one of the big problems is people don't know what Problems the AIs can fix. I mean, it's a lack of imagination in some ways. And it's not entirely their fault because you don't know what you don't know. There's it's beyond the Dunning Kruger effect. It's just you don't know. And and you know, like cutting the end off the ham that you talked about, you know, let's let's bring our banter section back into back into the AI chat. That's just how we do things. And I don't know that there's a better way or a different way, or that there's a there's reasons why that doesn't apply anymore. We don't challenge as as as the um the design thinking people would like to say, we don't challenge the awesome. Orthodoxies. It's just that's the way things are done and it works. And so we continue to do it that way. Um, and when you really push these, and the AIs, you know, to a to a point, don't actually help you imagine. Like if you just say, Hey, can I do this? 50 50, you'll get, no, not really, or they'll give you the not the best way of doing it. Like constantly, when I'm asking to build things or do things, it gives me suggestions. I'm like, hang on, why wouldn't we use this tech? Or what about if we did that? And it goes, Oh, you're right, I should have included that. Yeah, but I'm also concerned about this. Oh, well, if you're concerned about that, then we should definitely do this. It's like, uh, like it still requires so much of that human critical thinking and subject matter expertise. I mean, this is that argument between, you know, vibe coding, sort of slop apps, and agentic engineering. Is do you actually have any concept of the things you're asking it to do and steering it at all? Or are you just sort of punching the ball across and saying run wild? So [19:57] Matt: And there's a huge difference too, um, that we're seeing now. With people who do know what they're doing, but they don't know how to scale it. You know, like even those people who are technical experts that, you know, are super skeptical about AI and you know, they might be dabbling and they're like, oh yeah, look, it can do some things. But then there's the people that are like, let's let's say less intelligent than them, less experienced than them, and you're getting a fax. [20:23] Richard Webbe: Sorry, I was oh I I muted. I'm trying to mute, sorry, guys. [20:32] Matt: Oh man, that's just it's [20:33] Dave Pengelley: It's [20:33] Matt: been a long [20:33] Dave Pengelley: live [20:33] Matt: time [20:33] Dave Pengelley: live [20:33] Matt: since [20:34] Dave Pengelley: television. This is awesome. I love it. [20:35] Matt: Um [20:36] Richard Webbe: That's that one. [20:37] Matt: Yeah, so the person that like knows less can achieve more value because they understand how to actually scale and play the system more. Um I used to experience this when I play games. Like, you know, I used to learn a game so heavily and go deep into guides, but I never really like mastered the game because I never played it. I was just spent too long learning it. Like, so it's that same concept, I guess, in at times with these with these things. Like no no one really can connect unless you actually are in there using it every single day and like you the critical thinking and imagination, like you said. [21:10] Dave Pengelley: It's funny, my son, like he he's 15 and we play with some of these sort of strategy board games. And the first time he plays a game, he'll he'll go deep in thinking he's got this amazing strategy, and he'll he'll all sort of lean into like hard into one sort of core play. And almost inevitably it doesn't pay off for him because he's gone too hard. Like he's just he's just focused. Focused on one particular thing where he thinks if he can maximize that, that'll do the the lot. But you know, by the second or third game, he's worked it out and he does significantly better because of that played experience. Like I can, oh, I've devised all these strategies. But what actually works when I put the put the pedal to the metal and that experiential, I mean, we talked about this uh when I was talking about my game stuff for the other week, that iter iterative building, I think, is generally a better path forward than this whole dark factory concept. Now, there might be people that have got these really sophisticated Ralph Looping Dark Factory gas cities going on that are doing amazing things. Although when you look at Steve Yeegie's comics that he makes, his agents have lost the plot half the time as well. So I'm I'm I'm less sure of that. But I think, you know, do it, review it, do it, review it, do it, review it is a really strong loop with humans in it to get the best outcomes. Because if you just send the thing away for two days to build stuff. Who knows where it's gonna go? [22:31] Matt: Yeah, I've tried to adopted a really front-heavy mindset now. So that, you know, kind of like going into a board game and you read the instructions, like what are the rules? So imagine giving your your agents this intense encyclopedia. You know, they're not going to really follow it all. They're going to get lost. They're not going to do anything. But instead, imagine it happens essentially automatically. So you try and do something and kind of like a type system in a language like TypeScript or Rust. You know, you get this immediate. feedback when you do something that's wrong that's actually typed. You know, what you said about like do it and then review it, the extra step there is retrospective, you know, adjustment so that the next doing doesn't repeat the same mistake. So you basically over time get this buildup of a code base that self-corrects and it teaches the agents how to do things properly without you having to do it. And I'm feeling massive success with that now where like I don't have as much panic and stress in what I'm building, I can basically just send it off and all I need to approve now is mock-ups. [23:36] Dave Pengelley: Yeah. [23:36] Matt: Like, you know, w what [23:38] Dave Pengelley: Yeah. [23:38] Matt: does it actually look like and how does it feel? [23:40] Dave Pengelley: After after multiple iterations of my game that I'm building where I've got Chat GPT doing a lot of the planning and then giving the work through to Hermes, which is also running an an open AI uh model most of the time. Um, we're well down the track and we're just it's the next level bits of the build are getting heavier, but I've been running on a very fairly vanilla Hermes with no soul specification, nothing, just you know As the coding harness. And we've I've had some issues lately where I've just like all these repeated testing and these micro testing loops where it's just like always asking for approvals. So when I finally got it to break through that and finish that block of work, I was like, all right, we're going to do a full efficiency review around how we're actually working between the Chat GPT project instructions and the Hermes. And Hermes can go look at all the work it's done and self-assess itself. And I threw like Astra Max at it and went, just go and do a full Self-assessment. Here's the the project prompt from Chat GPT. You can look at yourself. Give me a plan of recommendation. It gave me a bunch of recommendations. And finally it said, I don't think you should customize your soul MD. I'm like, what are you talking about? Why wouldn't you? It's like a basic one. Oh yes, I I over this is like Max Astra. And it's like, you're right, I overcompensated, and we shouldn't do this, and we should actually put some sort of guidelines for how we do things in there. I was like, uh, but I took all that, went back to chat, chat reviewed it all the way. Yep, I would also. Just think about these and that didn't come, and this one conflicted with that. So take that back to Hermie. So a little bit back and forth. And then finally they all agree they loved each other and bang. New working instructions. So we can move forward with tests where we need them, checkpoints where we need them, better transparency between the two tools. Um, and it's, you know, like you were saying, Matt, it's that that learning and that regression, not so much regression, but that um retrospective back into the operating layer in order to make sure you move forward more efficiently. [25:27] Matt: Retrospective [25:27] Richard Webbe: It's interesting. [25:28] Matt: is such a fun word. Like I I I've never really used it much and like a lot of this AI vocabulary that we just seem to all be adopting, you know, like I [25:37] Dave Pengelley: Well [25:37] Matt: um [25:37] Dave Pengelley: re retra is from like sprints and agile type stuff. [25:40] Matt: Yeah, okay. Well that yeah [25:42] Dave Pengelley: Yeah, yeah. [25:42] Matt: so Matt Pocock has put together a really good retro skill as part of his pack. Um I would suggest having a look at it for that, is that sort of workflow? Like you if you're keen on being in the loop and and doing that deep retro, like yeah, his thing's really good for that. Um [26:00] Dave Pengelley: Got annoyed where it would just stop building and say, Can you approve me to running this test? I'm like, why am I approving you running a freaking test? Just run the run the tests. Tell me when when there's actual things I need to log in and play to actually [26:11] Richard Webbe: I'm [26:11] Dave Pengelley: do [26:11] Richard Webbe: I'm [26:11] Dave Pengelley: human [26:12] Richard Webbe: sorry, [26:12] Dave Pengelley: actual. [26:12] Richard Webbe: Dave. I love it when the checks and balances come back to annoy you. [26:17] Dave Pengelley: Ah, yeah, it's one of those things where and then I was setting up another agent to help me with um my church treasurer stuff the other day. And the agent set up the other agent, like with all these super onerous read-only things. I'm like, why? Why have you locked this thing down so much? Like, what does that achieve? Like, what is it to you if I want this agent to be able to write me a file or do something like, Oh yes, I'm sorry, I was trying to be protective and went too far. And it's like, oh, come on, just that's that's [26:42] Richard Webbe: So [26:42] Dave Pengelley: weird. Like [26:43] Richard Webbe: let's [26:43] Dave Pengelley: we talk about these things being unsafe and going rogue. I find completely the opposite, especially with the open AI models, um, in the Hermes harnesses [26:51] Richard Webbe: go. [26:52] Dave Pengelley: I run them in, they they are really, really to the letter of the law. They will not like cross a T that doesn't need crossing. It's it's Wild. [27:00] Richard Webbe: So let's have a have a step back because I like where this conversation's heading, which is something Matt often says. So if I was to get AI, you know, 50 years ago, and we'd never really seen a car, and I got AI to design a car, a personal mobility device that we could all use, assuming weirdly in a with in this world alternate universe, we didn't have any then. And then I got the design of that car, and then I got some agents, go tell me. What impact this car would have on the world, on people. And if it came out and said, well, a lot of people are going to get pregnant in them. Lots of people are going to die in them. Lots of people are going to commit crime in them. Lots of people are going to do bad things in them. Lots of people will move around when they probably shouldn't be. The car would probably never be built, really, if you were, if you had that lens on it. And this is where I find the stuff I know that you two do so well. You get a problem and you're very creative and you solve the problem. I know I come from a slightly different angle from a commercial and a cultural point of view, although I think we're all pretty good on the cultural side. So I ride my bike up and down Beach Road here in Melbourne Bayside every day. And I was riding last night instead of yesterday morning, and I had an epiphany. Because as I ride along, I've got the sun setting on my left, beautiful view of. The water, and I have all these new houses on my right, and every house is lit up with a five foot by six foot LCD screen, as bright and as colourful as you can imagine. And it gave me this weird feeling of where the new technology and the natural world were kind of, do you know what I mean? They were like they were poles apart, yet they're only about 20 feet across the road apart. So being a little bit of an artist like my dad, I came home and I did a sketch. Of what that actually looked like. And it just reminds me of what we're talking about. The fight between culture, reality, protection, productivity. It is a weird mishmash that humans aren't catching up with. And just for fun, I printed it off. But that was my sketch I did last night from my bike ride. And I can tell you now, it actually looks like [29:23] Dave Pengelley: That's [29:23] Richard Webbe: that. [29:24] Dave Pengelley: great. [29:24] Richard Webbe: So [29:24] Dave Pengelley: That's [29:24] Richard Webbe: it actually looks like that. See the blue, they're the they're the sorry on this side, they're those plasma screws. And as I ride along on my bike, every day I see the beautiful water and the boats and the fishermen, and and I see all these people sitting in their houses with these massive TV screens on their walls. [29:42] Matt: Did [29:42] Richard Webbe: God [29:42] Matt: you draw [29:42] Richard Webbe: knows. [29:43] Matt: that? [29:43] Richard Webbe: Sorry? [29:43] Matt: Did you actually draw [29:45] Richard Webbe: Yes, [29:45] Matt: that? [29:45] Richard Webbe: yes, I did art for many years. Don't worry about it. [29:47] Matt: Wow. [29:48] Richard Webbe: I'm [29:48] Matt: Maybe [29:48] Richard Webbe: actually going [29:49] Matt: don't [29:49] Richard Webbe: to use [29:49] Matt: worry [29:49] Richard Webbe: that. [29:49] Matt: about it. [29:49] Richard Webbe: The [29:49] Matt: That's [29:50] Richard Webbe: reason [29:50] Matt: that's [29:50] Richard Webbe: that's [29:50] Matt: really [29:50] Richard Webbe: that's [29:50] Matt: good. [29:51] Richard Webbe: the reason that's it's only crayon, only takes a few minutes. I used to do caracatures too. I'll I would I would try one of us, but but Dave's got AI. Doing it much better than I could ever do it. But [30:03] Matt: That's [30:03] Richard Webbe: the [30:03] Matt: still [30:03] Richard Webbe: funny [30:04] Matt: like [30:04] Richard Webbe: thing is. [30:04] Matt: you know, this this is literally your story and your point here in action, like juxtaposed within itself. [30:12] Richard Webbe: That's [30:12] Matt: The fact that you've [30:13] Richard Webbe: it. [30:13] Matt: done a drawing about the juxtaposition within a podcast that talks about this juxtaposition is just [30:23] Dave Pengelley: Yeah, you're [30:25] Dave Pengelley: I I love it. I mean as you guys know, I love the demo and a demo and a demo and the podcast. Podcast. It's very, very cool. Um, your your artist skills are way better than mine, uh, Richard. This was my effort the other week. [30:38] Richard Webbe: I love [30:38] Dave Pengelley: Um, [30:38] Richard Webbe: it. [30:39] Matt: That's the [30:39] Richard Webbe: I [30:39] Matt: best [30:40] Richard Webbe: love [30:40] Matt: person I [30:40] Richard Webbe: it. [30:40] Matt: can do as well. [30:42] Richard Webbe: Yeah, I had I hadn't drawn for a while. I was trying to design the front the front um fence for my new house. It hadn't been built yet when I was building it a few years ago. So I hadn't got pen to paper for a while. I've been practicing. So I've grabbed some crayons and I I won't show it, but I drew this, looks like a two-year-old's picture. of what the front fence and the house and the trees would look like. But I'll st I'll bring it up next week. Do you know I took a photo of that and they were not very far apart at all. They were very, very close. [31:11] Dave Pengelley: Sounds like they need a better trade to use if it looks like your two-year-old picture. Um, [31:15] Matt: You [31:16] Dave Pengelley: I mean you [31:16] Matt: agree. [31:16] Dave Pengelley: talk about the car the car and the impact that the car had, and so on, and people's ability to imagine the world with the vehicle. Have you guys heard the story of Bertha Benz? [31:27] Dave Pengelley: So, Carl, Carl Benz, you know, from You know, Mercedes-Benz Origins. He built this car and he was kind of stuck. Like no one really wanted it. No one knew what it was for. And so one night, his wife stole the keys, basically, uh, is how I remember the story. And she drove for nearly a hundred kilometers. And so she just showed how quickly and how far you could close distances between places by using this new technology. Compared to horses. And so, yeah, she drove all the way to some other town to see her mother and then drove back again. And so built up this thing. Like as she drove back through the town, people were like, Did you see that thing that came through? Oh, it's coming back again. And built up all this hype around the automobile [32:12] Richard Webbe: What [32:12] Dave Pengelley: as [32:12] Richard Webbe: a [32:13] Dave Pengelley: well. [32:13] Richard Webbe: fabulous story. [32:14] Dave Pengelley: Um, it's worth looking into and hearing the story. I can't remember where I heard her, but uh it's a really interesting tale around how if she hadn't got out there and showcased this thing in public, you know, building in public, people wouldn't have been able to get their head around it. I mean, I discussed this with Tanya on the AI. The Operator Stories podcast the other day, the Henry Ford quote. If he'd gone and asked people what they wanted, they would have said faster horses. [32:37] Matt: Yeah. [32:37] Richard Webbe: Exactly. Well, you've heard my story about when they wanted to build the space shuttle and a reasonable spacecraft and they went and asked the engineers and they said trillions of dollars and they got some artists to draw it. And they say, Oh, we could have done that anytime. What didn't you tell us in the first place? It's [32:57] Richard Webbe: that is a great story. And I I like that one. [33:02] Matt: It was very good. On the on the orbital topic, what do you and a bit of teleco, what do you guys think of that that new band that SpaceX or Starlink has um taken control of? So they can actually have low Earth orbit, um lower frequency band that used to be used for something else, um, like radios and stuff like that, but it's gonna just be like a blasted and it it can actually penetrate concrete and stuff so you don't necessarily need cell towers if they get all that infrastructure in [33:35] Richard Webbe: So [33:35] Matt: place. [33:35] Richard Webbe: this is new information, I'm very excited. [33:38] Dave Pengelley: This is this was I don't know if if anyone remembers back in my, you know, I've already mentioned my Telstra history, but back in that era, like late 20 zeros, whatever that's called, um, when Telstra every everyone had their their 3G networks and then Telstra released their next G network. Um and I was I was fortunate. To be uh selling the Next G network against the other 3G competitors. And why NextG was special is because all the other 3G networks ran at 2100 megahertz, uh 21 or 2400, depending if you're Optus or Vodafone. And Next G ran at 750. They'd actually taken back some of their old analog band and repurposed that for their version of 3G. And that meant because it was a longer, slower wavelength, it [34:27] Richard Webbe: Thank [34:27] Dave Pengelley: could travel [34:27] Richard Webbe: God [34:27] Dave Pengelley: much [34:27] Richard Webbe: for [34:28] Dave Pengelley: further [34:28] Richard Webbe: that. [34:28] Dave Pengelley: distances, so they got much better coverage. From their 750 megtowers than the 2100, which was like used all its energy squiggling up and down and didn't go as far. And those longer waves also penetrated through concrete and buildings better. And [34:40] Richard Webbe: Yeah. [34:40] Dave Pengelley: so that was why in the early 3G race days, Telstra had such a massive advantage because they were using those lower wavelength signals to get [34:50] Richard Webbe: Yeah. [34:50] Dave Pengelley: the transmission distance and penetration. So [34:52] Richard Webbe: Yeah. [34:53] Dave Pengelley: yeah, I'm I think you know there's definitely um merit [34:57] Richard Webbe: Think [34:57] Dave Pengelley: in [34:57] Richard Webbe: about [34:58] Dave Pengelley: that [34:58] Richard Webbe: and [34:58] Dave Pengelley: strategy. [34:58] Richard Webbe: think about I mean all the rent. Telstra in those days too, David, and I know exactly what you're talking about. Think about the impact to the value and brand and the cultural impact. Telstra got established as the mobile phone that worked further and farther from that moment onwards. And everyone after that was an also ran. [35:16] Dave Pengelley: Yeah. [35:17] Richard Webbe: Today, even though Ericsson makes the same technology that most of them use, right? A couple of niche competitors. You know. [35:25] Dave Pengelley: And then since then we've had 4G and 5G, and whoever knows, because the whatever marketing and technology legends Came up with Next G obviously haven't been hitting the bandwagon as much, and now they are scared of satellite and Elon and all these other new technologies coming through. Like it's a whole different market now. But and 5G obviously is a higher frequency and much shorter range. I mean, that's one of the issues with 5G. You can get a high bandwidth, but you've got to basically be sitting on top of the tower to get it. [35:50] Richard Webbe: Yeah, yeah. [35:51] Dave Pengelley: Um [35:51] Richard Webbe: Well they've yes, having worked at Ericsson for a while and and with on the Telstra account at Ericsson, there's there is some amazing things they're doing now with those trans. Transmission towers and frequency bending and uh you know identifying people on different sides of the tower, they can direct the beam of the uh the [36:09] Dave Pengelley: Yeah. [36:09] Richard Webbe: uh data stream directly to them. It's it's quite fascinating. [36:13] Dave Pengelley: On the um before we hit into some news and and things, I just want to close out on the people not understanding AI or the fear of AI and some of that sort of stuff that's going on. And as I've been, you know, pursuing a bit of game development lately, um, I've noticed places like Steam, which is one of the biggest marketplaces for video games online, they have this whole AI disclosure section where developers need to talk about if what they've what they've used in the game, any content, art, code, sound created with the help of AI tools during development, you gotta flag that basically. And if you're using AI in real time for, you know, characters, operations, generating stuff in real time, you also need to disclose that because players wanna know with this whole AI slop revolution. Like, am I playing an actual game that a game developer that a human wrote, or is this all just AI generated one-shot stuff, like all the little rocket demos and things we see? Like the stuff you were talking about with, you know, these Xbox mashups of different genres. The needle and haystack game has been talked about lately. That came out recently where someone made this game where you basically pull individual strands of hay off a stack looking for a needle, just like an idol game. And obviously, that is super simple for an AI to code. And there was a A million clones of the needle in a haystack app the next [37:33] Matt: Do [37:33] Dave Pengelley: day. [37:33] Matt: you think it'll get to the point where people don't care? [37:37] Dave Pengelley: Uh it depends [37:38] Richard Webbe: Yes. [37:38] Dave Pengelley: on who, [37:38] Richard Webbe: Yes. [37:40] Dave Pengelley: but but I and then Richard, you've talked about how your kids don't love think all AI stuff is terrible, AI generated slop. Uh LinkedIn brought in a this looks like AI slop button to me. Um I want to this looks like generic, useless brain rot button. [38:00] Matt: is worth like having a quick discussion about. So in the in the realm of like cabinet making, woodwork, um, chippies in general, you know, if you're if you're a craftsman that that builds stuff with wood, you know, there's there's that that huge divide between people that are like hand tools or like minimal power tools versus CNC machined carved objects. Like, you Like you get your table router and it just carves the thing out, and like you you sand it maybe a little bit and it's done, and you sell thing, and you're like, Hey, I make woodwork thing, but do you like [38:41] Richard Webbe: Well, [38:41] Matt: you [38:41] Richard Webbe: that's [38:41] Matt: know? [38:41] Richard Webbe: been on for years, hasn't it? And it's I have uh I have a brother who's a museo and has been on the road for many years and you know and and I have this debate with him all the time. And it's it's it's perception and perspective, isn't it? You see, um I don't know if you guys remember if you go into uh my dad being the artist that he was, hence all The artwork that I work with as well. I remember someone coming to our door in the 70s. Now, you've all been into an old doctor's surgery, or and you see those pictures that have like Japanese, and they've got Mount Fuji there, and they're all the same composition. They've got trees either side, and they sometimes have a lake, they sometimes have a mountain, sometimes with snow on it, sometimes not. They're all the same. And uh, you know, some guy came to our front door and said, Oh, I'm a young artist. I I painted these paintings. Would you like to buy one? This is back in the 70s. And my dad told him to F off and said, get a life and you're not an artist at all. And I remember going to him, Dad, what'd you do that for? He said, Richard, I've seen it. You don't, because he's a real artist. You don't understand. So what they do in these factories is they get a big long lump of wood and they draw it right across the factory floor. And then someone goes along and sketches in the trees and the mountain. Next person comes along and paints in the and they just do it en masse and then cut it up. Up. So if you look at those so-called paintings on the walls of these, you know, surgeries and cheap offices and stuff like that, they're mass-produced. It's as close to AI as I can imagine, right? But they would run around and sell them door to door, pretending to be an artist, and I'm a struggling artist, could you give me money? The scam of being fake versus authentic has a soft line and it's hard to work out who does it. Because if I design something really beautiful out of wood but I used a lathe to make it did I cheat or did I carve it by hand? And I get your point. I don't have the answer but it's interesting how the cultural and acceptable line moves very quickly either way. [40:44] Matt: Makes me think of um like you just said perception like hormosey always talks about perceived value you know like [40:50] Richard Webbe: Yep. [40:50] Matt: whether something is valuable or not kind of is irrelevant if the perceived value exists. [40:57] Richard Webbe: As brand marketing that is exactly it. And the Chinese with the cars today are destroying that model, aren't they? You see, everyone likes a valuable car, whether it's a Rolls-Royce or a Mercedes or a Beamer or something like that. These are perceived European value because they're quality cars and they've always been well made, going back to what Dave mentioned about the first car ever made by Mr. Ben many years ago. So if I get a Chinese car for a third the price that runs on electricity, is that a valuable car? [41:30] Matt: Yeah, it depends what you value. Like [41:34] Richard Webbe: I don't know if you've seen the new one that looks like a Rolls Roy. It's not very expensive. And oh my God, it is, excuse my um using the Lord's name in vain, but it is unbelievable what [41:45] Matt: Yeah. [41:45] Richard Webbe: it can do, how big it is, and how cheap it is. [41:48] Matt: It's like I I would love to own a Model Y. I just I don't know, a bit of a Tesla fan. And um [41:56] Richard Webbe: You've got [41:56] Matt: but [41:56] Richard Webbe: a Tesla, haven't you? [41:57] Matt: no, like my my perceived value. Value of them is like a certain rank, but if I had that in the driveway, my partner would look at it and her perceived value of it would be far lower. So like it's quite a it's a fascinating thing. [42:14] Richard Webbe: There's the value of what you use it for, and there's your personal emotional value of what it says about you and what you're investing in. And so value has, we think one dimension of it has about ten different dimensions when you look at it from that point. If you [42:32] Dave Pengelley: Yeah, I think economists talk about utils and and the measuring the utility of something and that isn't all necessarily, [42:38] Richard Webbe: want to [42:38] Dave Pengelley: you know, how [42:39] Richard Webbe: can't do do that, that. you [42:39] Dave Pengelley: productive utility it is, but that that, you know, desire that drives supply and demand as well. Yeah, how many util's value do you get out of something? That's why they don't call it utility, they call it utils, because it's generic non word. [42:52] Matt: Makes sense. [42:54] Dave Pengelley: Yeah. Yeah, I'm uh I'm uh off the back of that um that steam thing, I'm I'm aware of what that means for for sort of publishing workout. And I've been experimenting with uh some new developments uh like this tool, like giving giving other developers a tool that I built with AI, but it means they can generate art and assets without using AI. Because AI, again, you get all that generic slop stuff. If you want things to programmatically look similar or the same, like what do you do? How do you give people those tools to create [43:27] Richard Webbe: Dave, [43:28] Dave Pengelley: various [43:28] Richard Webbe: did you [43:28] Dave Pengelley: assets? [43:28] Richard Webbe: Dave, did you create This yourself. [43:30] Dave Pengelley: Yeah, this is this is my me. So Timothy Games is me. And so I've built this tool now. And I'm gonna say I. This is actually my open AI dot. Not a single codex token has been burned building this tool. This has all been for [43:41] Richard Webbe: Or [43:41] Dave Pengelley: free [43:41] Richard Webbe: you [43:42] Dave Pengelley: using um Hamilton. But yeah, I can I can go and say, okay, what do I want? Do I want more asymmetry, less asymmetry, size variation in crystals? We can go and change the change the spread. Um yeah, we can go and say we want them [43:57] Richard Webbe: said [43:57] Dave Pengelley: to be [43:57] Richard Webbe: this [43:57] Dave Pengelley: green. [43:57] Richard Webbe: this is this is, I mean, that's fantastic. Dave, I'm very impressed. Um, and I know I'm more than very impressed. But this is the point, isn't it? So if we have a furniture maker who in the 1800s did everything by hand with sandpaper and a uh an old, you know, one of [44:14] Dave Pengelley: Yeah. [44:15] Richard Webbe: those wooden wooden, but today we've got a lathe and he's making it himself. Is it still handmade? All it's changing is the tooling. And so now you come along with some automated tooling, like AI or your animation controller there would. [44:30] Dave Pengelley: Procedurally generated, so it's not AI. AI is not generating that. That's all just maths creating those pictures and images and variations. And so that way, I don't think it falls under the AI generated category. A tool built the tool, an AI built the tool, but the tool itself is procedurally generated maths. That's not AI that's generating that art. And [44:46] Richard Webbe: But [44:46] Dave Pengelley: that means [44:47] Richard Webbe: what [44:47] Dave Pengelley: you [44:47] Richard Webbe: if [44:47] Dave Pengelley: can [44:47] Richard Webbe: the [44:47] Dave Pengelley: get [44:47] Richard Webbe: tools [44:47] Dave Pengelley: consistent. [44:47] Richard Webbe: that you're using to build the tool were created by [44:50] Dave Pengelley: Yeah, [44:50] Richard Webbe: AI? [44:50] Dave Pengelley: but but the thing that they're using for customer facing is not AI. Like this is just mathematical art that they can create, they can customize, they can create, you know. Crystals and rocks and bushes and grasses, and I'm building a whole sort of suite of things so they can get their overall art direction looking consistent. So they might not want to, they might be programming their own game by hand, but they need to buy art and asset packs. They can customize their own bespoke art using my tool in order to get that. So [45:18] Matt: I [45:18] Dave Pengelley: it's [45:18] Matt: think uh [45:19] Dave Pengelley: weird. [45:19] Matt: I think this is this is kind of sounding like veganism, where where y if you if you want to buy a tool you have to know the supply chain of That tool, like was any AI used in the creation of this tool? [45:34] Richard Webbe: Were [45:34] Matt: Um [45:34] Richard Webbe: any animals hurt during the making of this video? Yes, I agree. [45:39] Matt: it's gonna be the same thing, and yeah, there will be people that will want that. It's kind of like buying Australian made, you know, is any of this manufactured in on en masse in a Chinese factory, you know, and do you care about that? Um there's gonna be that. So that kind of disclosure thing that Steam is saying. [46:00] Richard Webbe: Interesting where and how quickly our moral line moves, and I always say that because the moral line is quite high until people save lives or money, and then it comes down very quickly. [46:11] Dave Pengelley: But I think just from a from a you know, in that instance, assets and and Richard, you come from the art background and and your father and so on. We talk about, you know, the the the quick the the dodgy copies of the factory and in engraving the wood. Like at the like the issue with AI generated stuff is it is still probistic, even though it's Really good at doing pictures and things. Like if you when you were mocking up your book covers, Richard, you look at zoom in on your face and it puts all these extra cracks and wrinkles in the skin tones and stuff, like uh, [46:39] Richard Webbe: You're [46:40] Dave Pengelley: which [46:40] Richard Webbe: right, [46:40] Dave Pengelley: are not [46:40] Richard Webbe: I'm [46:40] Dave Pengelley: yours. [46:40] Richard Webbe: not that [46:40] Dave Pengelley: They're [46:40] Richard Webbe: old. [46:41] Dave Pengelley: not yours. No, [46:41] Richard Webbe: Thank [46:41] Dave Pengelley: no, they're [46:41] Richard Webbe: you. [46:42] Dave Pengelley: not yours. It's the AI-generated version of you. And you get these really obvious tells, or sometimes when someone drops a logo in and says, Hey AI, put this logo in the thief in the picture for me, it scrabbles up the text in the logo. Like it's still a random sequencer in many ways. It's probabilistic, not deterministic. And so If you're generating an experience for someone and you want it to be a cohesive, unified experience, the reason I built that tool is because you can then have different colored crystals that are all from the same family and they all look similar, even though you can actually customize them and make them different shapes and sizes and depths and colors, but they're actually consistent, and you're gonna get this random slop of AI with a random pixel and a thing there that shouldn't be there, and and just yeah. So I I think and and again that building that is one of those examples of it's something I I never would have thought to build two weeks ago when I didn't have a need for assets and things in games. And it [47:35] Richard Webbe: Very [47:35] Dave Pengelley: just comes down [47:36] Richard Webbe: impressive. [47:36] Dave Pengelley: to that creativity and that imagination for people and just showing them what's possible. I've been pushing my GrockBots really hard. I've been pushing Hamilton, my OpenAI dot really hard this last week. And like that website is all built by Hamilton. Um amazing stuff. Just wild. [47:54] Richard Webbe: Impressive, impressive. So you guys know about Ben Affleck and his AI company that he Sold. Did you guys know about that? [48:02] Matt: Yep, I don't know the name of it, but I'm aware of that. [48:04] Richard Webbe: Yeah, I was but I was wondering if you so what I found interesting about that is Ben Affleck is a very accomplished movie maker, director, actor, producer, even camera angles. So when he saw AI coming along, I remember his early comments were I know this is going to kill the industry. So he very quickly got on board with an AI company that produced video. But what he added to it was the perspective of the libraries. For a cinematic view. And that's his IP. That's not mathematics. He then got all those libraries and got them to look at, you know, if if we told a basic AI tool, hey, you know, create a video of someone racing their pushbike along the beach where I do it, it'd probably be very boring. But with Ben Affleck's libraries, it becomes real and and a lot better because it's using the Hollywood cinematic view of how they capture things. And it's interesting where the line is, is it Ben Affleck's creativity that he's just replicating for money or for for the benefit of the audience? Or do we still call that AI? Or are we just borrowing his IP with his permission? It's it's a it's a moving, uh I don't know the answer, I just just posing the question, so to speak. [49:16] Matt: It's kind of like um it's a similar concept to if you are a course creator and you've got some sort of industry knowledge or some special skill that you want to teach people, and yeah, you sell access to your knowledge, they take your knowledge. And then they go and apply it to the world in their own, you know, uh understanding of reality. You know, ha is that the same thing? Like, you know, [49:38] Richard Webbe: Yeah, [49:38] Matt: where's [49:38] Richard Webbe: no, it's it's exactly, I mean, that's the human condition, isn't it? We uh uh I I uh who was it that I think even Einstein said there are no original thoughts, there are just people that put new perspectives and new angles on those [49:50] Matt: the [49:50] Richard Webbe: thoughts, and that's how the human condition progresses, you know, kind of like the use of the car that Dave mentioned, [49:56] Matt: idea? [49:56] Richard Webbe: you know, the intelligence was her husband built the car. But the real intelligence was when she came along and actually used it and gave a new user experience that did the tapestry. Um, it's hard to know. I don't think there's an answer, but I I suppose I advocate for people being less panicky about where we're headed with AI. I want people to be more accepting and put themselves in the driver's seat and ask themselves, what would you do? Where is your moral line and when will it change? [50:25] Dave Pengelley: I um I I saw this video this week, Matt, and thought of you immediately when it came. Up. [50:30] Richard Webbe: You can't still touch. That is awesome. [50:34] Dave Pengelley: Right. And this was this was a game dev one, but it's like the the work's getting easier, but that makes taste more valuable. Uh and he talks about taste a lot, which I know you talk about, Matt, and I think it's a that that's that human element, Ben Affleck's cinematography point of view, that he's honed over working on dozens of films and producing and directing films and so on. Like he's you learn that some people in those instances learn and they build the eye for it and they build that skill and that muscle memory and they can translate that. Into the output. [51:01] Matt: I think there's um there's two competing things as well. Like taste is one thing, but then you you can't discount skill. Like there is like a level of skill, and I've got a perfect example of where the two separate. Um it's a movie example. So like imagine an actor who or an actress that is uh very highly skilled, but maybe doesn't have the taste. Whereas the director has that taste, but possibly doesn't have have the skill and is like, no no no, do it again like this. You know, that's [51:34] Dave Pengelley: Or vice or vice versa. I mean, that that's actually one of the criticisms of the Star Wars prequels is that when they made the original four, five, and six, George was still married to Martha. I can't remember. But George's wife was on set all the time, and she was the people person. She actually got the emotional beats of the scene, and she would coach the actors on the emotional resonance. [51:56] Matt: what [51:56] Dave Pengelley: One, two, and three, they didn't have her standing on set. [51:59] Matt: I'm saying. [52:00] Dave Pengelley: So George was very focused on no robot comes in here, this goes out, shoot that laser, Jedi jumps there. And that's why you look at Hayden Christensen, who is a pretty pretty decent actor, but he wasn't directed well. And so he comes across as this wooden stiff, like really bad in two and three, comparative of what the actor actually could probably have done with better direction and guidance. [52:21] Matt: Yeah. Yeah, [52:22] Richard Webbe: Whenever [52:22] Matt: I mean [52:22] Richard Webbe: you [52:22] Matt: that's [52:22] Richard Webbe: talk about Star Wars and we're talking about AI style IP in class, I saw the uh documentary of how You guys would have seen it a hundred times, I'm sure. But what fascinated me when I first saw Star Wars is how they got the the planets to explode or the machines to explode or the Death Star to explode. And I and I I, you know, as a kid, technically I was looking for a mistake because they're in space and there's got to be gravity. And I never realized, um, I think it was George's Lucas' idea, they blew things up and they lay on the ground, filming through a glass upwards, so that when it blew up, Looked like it was in zero gravity. I thought, what a brilliant little touch on what could have been like the old the old space movie where it was hanging on a wire. [53:13] Richard Webbe: So look, it you know, create human creativity, style, taste, where that creativity is copied or automated or used. We're gonna have to answer all these questions as we go forward. [53:27] Matt: I also think that there's there's such a massive argument. To be made still about like you don't even have to understand that AI exists. It's more about the possibilities of what can get created now. And those things like Dave's crystal tool there, that was created, yes, through the assistance of AI and like the speed of that. But like, hey, look, now you've got a thing, and this thing wouldn't have existed before. So it's kind of like the loom. You know, a lot of people talk about the weaving loom. You know, the loom is not AI. Like the person who invented the loom at the time was AI. You know, like you now have this thing that has changed and created such change. Doesn't matter that it was from AI as such, but like, hey, we have a thing. [54:09] Richard Webbe: Well a lot of great creativity is is by accident. Like, you know, like a lot of us have traveled through Europe and gone to the beautiful churches and seen these beautiful, you know, tapestries on the wall, and you look up to look at it, and or you go see the Louvre and you see all the paintings, and they go, gee, how much work would it have taken? My dad used to rave about Leonard. Da Vinci and Rembrandt's use of light. And then I think about another friend of mine. Um, you guys were still in nappies when this happened, but they were at um we're all working at IBM, and then IBM lost a lot of money back in the early 90s. So uh everyone got these large packages to leave IBM. And a lot of us were quite young, so we what do we do with this money? I had a lot of friends who went out and bought wineries because wine was on the rise back in the in the 90s, and uh if he's listening, he's a CIO today. If he's listening, he'll know what I'm talking about. And he set up a winery out near Kolac on the back of the volcanic ground out there where the grapes grow beautifully and stuff like that. And he had a few Christmas parties out there. But he won a champagne award with a pink champagne. Now, the way he made that champagne was he got the standard, I think, Chardonnay grape or whatever and method Champenoise and bubbled it away. And they have these big stainless steel vats as well as doing it in the bottle. But then he'd done a Shiraz as well, and he accidentally pumped the Shiraz in with the champagne. [55:31] Richard Webbe: So it became a blend. And he went, oh, bugger. But he just bottled it anyway, called, I think he called it MC2, sent it off. They won an award. Now, do we think the people that were, and it was a beautiful champagne, I might add to Matt, if you're listening, do we ever think that they were going, well, this was just a mistake, so I'm downgrading the value of this, or they just taste it and go, well, that's a nice champagne. [55:55] Matt: Yeah. I was uh I used to work at a coffee roastery a long time ago. I remember, you know, roasters about to finish. So there's a batch ready to drop, about 30 kilos worth. And the roaster dropped the fresh light green from the hopper into that almost finished batch. So you had this 50-50 mix of done and not done. Those kinds of mistakes are not possible to really recover from. [56:22] Richard Webbe: No, [56:22] Dave Pengelley: No. [56:22] Richard Webbe: you don't, you don't. You don't at all. [56:27] Richard Webbe: Very sad, and yes, very true. So look So when we come to AI and it's running these systems, I still think, like with what Dave did, like with what you do, like what I do, it creativity and the human condition is what we need for it to happen. So I don't want people to poo-poo AI too [56:47] Dave Pengelley: What [56:47] Richard Webbe: much. [56:47] Dave Pengelley: one of the things that's coming up in a lot of the um the interviews I'm doing on the new story show is how when AI is implemented in the businesses, they're growing and actually adding heads. That might be a different shape of head, like they're Replacing adding extra admin staff with adding other sort of skilled revenue generating resources, for instance. The guys I spoke to from Queensland last week, they um they turned down a job. They were going to get paid like a quarter of a million to come and implement uh a whole bunch of AI. And the brief was to replace 50 people. And they went, No, because A, that's not what we want AI to be doing. We want to be augmenting and enhancing workers, not replacing them outright. And that's just unrealistic to think that you can. come in and replace 50 staff with automated AI systems. It's just ridiculous. Like you're [57:34] Richard Webbe: It [57:35] Dave Pengelley: asking us [57:35] Richard Webbe: is. [57:35] Dave Pengelley: to make promises. There's no way. No way we could ever fulfill. So like that's a bad deal. We don't care how much money you're throwing at us, we're not taking that job. Someone might. Someone probably will. [57:46] Richard Webbe: But [57:46] Dave Pengelley: But [57:46] Richard Webbe: I [57:46] Dave Pengelley: those [57:46] Richard Webbe: don't think [57:46] Dave Pengelley: two [57:46] Richard Webbe: it [57:47] Dave Pengelley: guys. [57:47] Richard Webbe: works. I don't think it works. I remember I was traveling the world, I think, for Cisco Systems, talking about the value of online and VPN. Now VPN became this panacea for accessing stuff like your core, Matt, when people couldn't access the data they needed, all of a sudden VPNs and intranets became the thing for getting access to data and information. And it saved, I think Solvik, who was the CIO at Cisco, I think he saved 1400% of their costs of administration at Cisco when he first put the intranet in. It was fabulous. I think the board said to him, do that again and you can do whatever you like. And so that's a really important thing right is one of the one of the things that they had on their slides when they sent me around. I literally traveled the world telling people this story and saying you should put in an intranet because hardly anyone knew what it was. Um it they had a slide in there and said and you can sack all these people. And I remember going back to them going no I'm not going to travel the world tell people that that's dumb. Number one, you're going to kill your your own customers and your own audience and number two no one wants that. But what we did Do is we did some time and motion studies and some loyalty and creativity understanding of businesses. And all we did is say, and all those people that are doing that boring job of reconfiguring routers and setting up servers can now move over here and apply their energy in the right place. So I think, I think the AI is going to create more jobs. And those people, like you said, Dave, running around saying we're going to get rid of 50. I think it's all about repurposing our energy. Face it, there's no more typing pools. Remember the typing pool? They don't exist anymore, and there's no one on living on the gutter who used to be a typist, right? We repurposed those people, their loyalty, their knowledge of the business, their knowledge, and we brought them into other parts of the business and [59:45] Dave Pengelley: You know [59:45] Richard Webbe: they did [59:46] Dave Pengelley: who I don't [59:46] Richard Webbe: well. [59:46] Dave Pengelley: want to replace? Checkout operators. Self-service checkouts are a lie. They're stealing my money and my time, and [59:53] Richard Webbe: I [59:53] Dave Pengelley: I don't [59:53] Richard Webbe: agree. [59:54] Dave Pengelley: like it. [59:54] Richard Webbe: I [59:55] Dave Pengelley: Push [59:55] Richard Webbe: agree. [59:55] Dave Pengelley: humans back at the checkouts. I'll I'll deal with some bad bagging. It's better. Than the the mess of these little terminals that crash all the time. You gotta put your hand up and stand there waving around like an idiot. Just I I choose to go human. [01:00:08] Matt: It must [01:00:09] Richard Webbe: I [01:00:09] Matt: be such a [01:00:10] Richard Webbe: agree. [01:00:10] Matt: strange like uh case here because I have I deliberately avoid the human at the shops. I I much prefer self-scan every time. I never have an issue with it. Or if I do, I'm not fussed because I used to work at Coles. So I'm like, I was there when they first came out and you know I'm quite familiar with the oh yeah, that probably happened. Hey. [01:00:30] Dave Pengelley: Yeah. Oh, [01:00:31] Matt: Um [01:00:31] Dave Pengelley: I get I get why things happen. I just don't like it. And I'm like, you know what, they're not paying me. If if they gave me a five percent discount or whatever to swipe my own shopping, but there's what like what's the benefit for me? I'll go employ a human. [01:00:42] Matt: Yeah, I I get that logic. I'm a bit weird, I guess. Um on the [01:00:46] Dave Pengelley: You [01:00:46] Matt: topic of [01:00:46] Dave Pengelley: are. [01:00:46] Matt: what you were saying, Richard, just then at the end, I like again I'm a sample of one, but a lot of my revenue through my business, n very, very little of it is like AI is the value. AI is the product, like very, very little. It's more the fact that I'm able to build what these businesses want faster and with like much more leverage than what I would have been able to do by myself. So literally, like the majority of the value I provide is not AI. It's the fact [01:01:16] Richard Webbe: That [01:01:17] Matt: that I'm [01:01:17] Richard Webbe: that's [01:01:17] Matt: delivering [01:01:17] Richard Webbe: my point. [01:01:18] Matt: the [01:01:18] Richard Webbe: It's [01:01:18] Matt: thing. [01:01:18] Richard Webbe: your brain, it's Dave's brain, it's the way you look at problems and the way you narrate where people want to get to. That [01:01:25] Matt: Yeah. [01:01:25] Richard Webbe: is what's and so I see AI as automating the output. As speeding up the output, as tools like whether it's a uh you know, an order. I don't buy, I've never stopped buying manual cars years ago. What a waste of time changing gears when the car can do it much more efficiently and much better than I can. And so I look at things. I I don't get shoes with laces. What a waste of time. Everyone's still lace up their shoes every morning. I mean, really. [01:01:50] Matt: This is another perception of value thing. Like [01:01:53] Richard Webbe: It [01:01:53] Matt: I love [01:01:53] Richard Webbe: is [01:01:54] Matt: my manual car. Like I I just you can you can tear it away from me from my cold down hands. [01:02:00] Dave Pengelley: Tesla, then. Um, I mean, re rich that that reminds me of the the the classic faux pas in Back to the Future, where Doc Brown in the 50s has Velcro's shoes when Velcro wasn't invented for another decade. Um all right, let's let's let's round it uh a few quick little news things to to finish off. We we're at the hour, we've got to wrap this up, but I want to cover these things and do the news. All right. [01:02:32] Dave Pengelley: Right, so uh Claude, Anthropic, uh abusive behavior towards models is being banned. Uh Anthropic is [01:02:39] Richard Webbe: Oh, [01:02:39] Dave Pengelley: banning [01:02:39] Richard Webbe: I saw [01:02:39] Dave Pengelley: sus [01:02:39] Richard Webbe: that. That might be laugh. [01:02:40] Dave Pengelley: banning sustained needless abuse towards Claude. So apparently swearing at it, yelling at it, calling it names, belittling it, or telling it something more than a glorified autocomplete machine will no longer be acceptable. Unless, of course, you use M dashes when insulting it. Um and you know, this is this is unconfirmed, but uh in the same vein. Just in the guy who created the digital. Whip for Claude Code reportedly got a lifetime ban for abusive or cruel behavior. [01:03:05] Richard Webbe: So, can I add a add a uh a media filter on what you just said? And I did read that and see that. I doubled down on the detail, and I think the detail is using AI to abuse people. But some media people picked it up as abusing it, which is bad because when I'm driving along and I say, Hey Siri, would you please call my mum? Uh, you know, sorry, not my mum, she passed away. Would you please call my brother or something? Someone like that, and it ends up calling, you know, McDonald's for some takeaway, I do abuse my AI interface there. [01:03:38] Dave Pengelley: Yeah. Yeah. Um [01:03:42] Dave Pengelley: oops. [01:03:46] Dave Pengelley: Uh the other other thing on the on that topic though, I mean uh around well not on that topic, completely different topic. I saw this one pop up today actually, Caveman. So people might remember that Caveman came out as a skill you'd run with Claude. So m it speak less tokens and [01:04:06] Dave Pengelley: be more succinct to save you money. Uh they've now been backed by Y Combinator. So they've now gone from like this little joke thing that after 111,000 GitHub stars, they've been backed as a venture company. [01:04:19] Matt: Well, [01:04:20] Dave Pengelley: Uh [01:04:20] Matt: that's what [01:04:20] Dave Pengelley: to [01:04:20] Matt: the [01:04:20] Dave Pengelley: have [01:04:20] Matt: OpenAI um thinking traces uses now, or something very similar. [01:04:24] Dave Pengelley: really [01:04:25] Matt: Yeah, it's how they've done a lot of um you know token efficiency. [01:04:30] Dave Pengelley: Wow. [01:04:30] Richard Webbe: That [01:04:31] Dave Pengelley: I [01:04:31] Richard Webbe: is [01:04:31] Dave Pengelley: mean, [01:04:31] Richard Webbe: very [01:04:31] Dave Pengelley: cool. [01:04:32] Richard Webbe: funny. [01:04:32] Dave Pengelley: Cool launch video. [01:04:33] Matt: That is pretty good. I love [01:04:34] Dave Pengelley: Yeah. [01:04:35] Matt: that um that particular pixel theme. [01:04:37] Dave Pengelley: Yeah. It's [01:04:38] Richard Webbe: That [01:04:38] Dave Pengelley: pretty [01:04:38] Richard Webbe: came from [01:04:38] Dave Pengelley: cool. [01:04:38] Richard Webbe: Bleisha Suit Larry, if you remember. [01:04:40] Dave Pengelley: Hmm. [01:04:41] Matt: Oh, I [01:04:41] Dave Pengelley: Well, [01:04:41] Matt: do [01:04:42] Dave Pengelley: it's [01:04:42] Matt: remember that. [01:04:42] Dave Pengelley: that's that's interesting. I mean, that that little video I showed on on Timmy Games before I went to the actual tool, um, Hamilton just used remotion to build that for me. I said, Hey, I need a splash reel that shows the assets being modified and then them in a little scene, and I just built it amazing. Um, so there's that's caveman around. You know, trying to be more efficient. But the other point of view on that is this thing I saw pop out, which is this new another Y combinator backed thing where it's a little home AI server that runs, you know, local models, and you have your own little personalized hard drive that you plug into it. [01:05:18] Matt: I [01:05:18] Dave Pengelley: So [01:05:18] Matt: did [01:05:18] Dave Pengelley: it becomes [01:05:18] Matt: say this one. [01:05:19] Dave Pengelley: a [01:05:19] Matt: This one annoyed me because of the name of it. [01:05:22] Dave Pengelley: okey. [01:05:23] Matt: No call. [01:05:24] Richard Webbe: Oh [01:05:24] Dave Pengelley: Huh. [01:05:25] Richard Webbe: yeah, they stole your name. I would be hitting them with some uh [01:05:28] Matt: I [01:05:28] Richard Webbe: copyright [01:05:28] Matt: have I have [01:05:28] Richard Webbe: there, [01:05:29] Matt: no trademarking. [01:05:29] Richard Webbe: mate. [01:05:30] Dave Pengelley: Um but this this looks like a really cool idea and uh the memchip's interesting. It's that that whole future concept. Like imagine that this is a utility that every house has one of these little devices, and when you move house, you just pull your memchip and go somewhere else and plug it in, and you've got all your histories and memories and training ready to go. So yeah, it's gonna be interesting as we see more of this sort of miniaturization of this sort of data center local hardware uh running. Which You know, if you've got that, you don't need caveman because it's you're just paying the energy to run the box. It's [01:06:05] Matt: It [01:06:05] Dave Pengelley: a [01:06:05] Matt: is [01:06:05] Dave Pengelley: whole [01:06:05] Matt: still [01:06:05] Dave Pengelley: different [01:06:05] Matt: interesting, like the economics of that. I still find it so fascinating. We probably don't need to go into that right now at the end of the show, but it [01:06:11] Richard Webbe: Bring [01:06:12] Matt: it's [01:06:12] Richard Webbe: it up next week. [01:06:13] Matt: yeah, it's still too expensive to do it at home. Even if it's free, it's still too expensive. Like it's kind of a weird thing. [01:06:21] Richard Webbe: I [01:06:21] Dave Pengelley: Yeah. [01:06:21] Richard Webbe: have to ask um Matt's dressed like a lumberjack, and that's okay. Uh Dave, what's on the front of your shirt? [01:06:27] Dave Pengelley: Oh, this is Sir Roger. [01:06:29] Richard Webbe: Oh. Why have I seen that before? [01:06:31] Dave Pengelley: Um it's a local Australian place that makes sauces and things. Um and they make a sriracha, which they call Suriracha. [01:06:42] Dave Pengelley: Uh up near Port Macquarie. [01:06:47] Dave Pengelley: Um so Matt, if I get the price of that device that they're saying is $17.99 and divide that by $200, that's nine months. Of a [01:07:01] Richard Webbe: No, [01:07:01] Dave Pengelley: max sub. So if you don't if presuming you don't need other subscriptions on top, if this can do everything you need it to, it pays itself off in a year of having [01:07:11] Richard Webbe: no. [01:07:11] Dave Pengelley: your own local intelligence. [01:07:12] Matt: That's a perceived value thing there. So if it can do everything you need it to. So yeah, you're not getting the same value. And the the the I don't know how to describe it, but the economic economics of scale of actually getting things through the data center is always going to be cheaper. [01:07:30] Dave Pengelley: Yeah, but if this um I'm I I'm and remember we're talking about normies and regular people, not you uh and me, it's as much. But you know, they might go like if I can have this thing and you know $17.99 if you're doing reasonably decent well off or whatever, and you just want this like AI toy at home, you just pay for that one month and then you forget that you've ever paid for it ever again, and it's not an it's just like a a one-off expense, and you've got all your stuff indexed, personal assistant with your calendar and all that kind of stuff happening. All on your own stuff. [01:08:02] Matt: So the what you're comparing there though is $17.99 against $17.99 equivalent, which let's say it's GLM 5.3 flash, and the volume of difference that you can get, even just direct API. Let's say it's uh Deep Seek 4.1 flash, you know, both of those very cheap, very economical models, or even like Luna, you know, haiku, the new haiku, the amount of value that you get direct to those APIs versus What you'll get from that box is not the same. [01:08:34] Richard Webbe: Interesting. How long in the future before when I've got my Apple Watch acting as my AI document taker and my phone and my computer and everything, how long in the future before facts become ubiquitous? And just you know what I mean? We [01:08:55] Dave Pengelley: You're [01:08:55] Richard Webbe: have a [01:08:55] Dave Pengelley: you're [01:08:55] Richard Webbe: conversation. [01:08:55] Dave Pengelley: the one with the fax machine getting faxes during the podcast, Ridge. [01:09:01] Dave Pengelley: Uh, [01:09:02] Matt: Do you mean do you mean like how everyone can just you know, previously it was oh just Google it, you know, everyone had access to Google searching, is that what you're [01:09:10] Richard Webbe: Yeah, [01:09:10] Matt: saying? [01:09:10] Richard Webbe: I'm saying, you know, we have a conversation, we walk away, we have different views of what that conversation was. We won't be able to argue over it anymore because we'll just hit the replay button and AI tells no, you did say [01:09:21] Dave Pengelley: yeah. [01:09:21] Richard Webbe: that. [01:09:21] Dave Pengelley: Well, that's [01:09:22] Richard Webbe: You [01:09:22] Dave Pengelley: the [01:09:22] Richard Webbe: did [01:09:22] Dave Pengelley: new Apple [01:09:22] Richard Webbe: promise. [01:09:22] Dave Pengelley: Watch. The new Apple Watch has actually got that ambient listening mode, I think, um, that they've just advertised. So with like it's constantly sort of capturing the last sort of two minutes of conversation. So if you miss something you can say, what did what did Janet just say? And it will tell you what just happened. [01:09:40] Matt: Coming stuck in. [01:09:41] Dave Pengelley: Yeah. Like that's a bit weird. Um, but versus having to run granola or have your plod or your pocket device or whatever, your watch is now going to be ambiently recording those notes and being able to give you that instant replay. But you know, we we duff we touched on this at the end of you always bring this up at the end when we've got no time to talk philosophy and [01:09:57] Richard Webbe: Next [01:09:57] Dave Pengelley: uh [01:09:57] Richard Webbe: week, [01:09:58] Dave Pengelley: deeper [01:09:58] Richard Webbe: next [01:09:58] Dave Pengelley: government [01:09:58] Richard Webbe: week. [01:09:59] Dave Pengelley: conspiracies. Richard, but I will remind everyone um that, you know, this is not a political podcast. Um but we've we we we've got merch underway. We're gonna have merch soon. Um for all our fans that want that. And I I love this mug. Definitely need one of these mugs. [01:10:17] Richard Webbe: Remember, [01:10:17] Dave Pengelley: Um [01:10:17] Richard Webbe: this is our first merge product. [01:10:20] Dave Pengelley: the book. So yeah, I I actually told my agent, I said, I want merch for the show. I I think it was Grockbot. Go look at AI operators pod, go read all our transcripts and look for meme content like m um repeated ideas and concepts and memes that we could build some some merch on and it pulled this is not a political podcast is one of our running gags that we use all the time. [01:10:41] Richard Webbe: I want [01:10:41] Dave Pengelley: So [01:10:42] Richard Webbe: to get a t [01:10:42] Dave Pengelley: like [01:10:42] Richard Webbe: shirt that says this is AI operators, this is not a political podcast. I like that as a [01:10:47] Dave Pengelley: we'll [01:10:47] Richard Webbe: slogan. [01:10:47] Dave Pengelley: we'll we'll get we'll get your t shirt version as well as the mug. But [01:10:51] Richard Webbe: Awesome. [01:10:51] Dave Pengelley: um anyway. Thank you guys. Great chat. We've we've run way over time. Sorry everyone. I'm sure you've enjoyed it. Just listen to us at two X speed. We we're good for that. [01:11:02] Matt: Sounds [01:11:02] Dave Pengelley: Um [01:11:03] Matt: good. [01:11:03] Dave Pengelley: yeah, thank you all. We'll see you next week. [01:11:08] Richard Webbe: Heavy update.