AI Can Build It. You Still Have to Make It Good.

[00:04] Dave Pengelley: And we give it [00:05] Richard Webbe: What [00:05] Dave Pengelley: a little [00:05] Richard Webbe: do they [00:05] Dave Pengelley: minute, [00:06] Richard Webbe: use it [00:06] Dave Pengelley: and [00:06] Richard Webbe: for? [00:06] Dave Pengelley: then there's gonna be a bit of a gap before YouTube actually picks up the [00:09] Richard Webbe: Sorry, [00:09] Dave Pengelley: stream, and [00:10] Richard Webbe: I'm [00:10] Dave Pengelley: we are [00:10] Richard Webbe: I'm [00:10] Dave Pengelley: probably [00:10] Richard Webbe: back [00:11] Dave Pengelley: live [00:11] Richard Webbe: back on my phone because my [00:12] Dave Pengelley: now. [00:12] Richard Webbe: computer did die. Oh, The [00:13] Dave Pengelley: Welcome, [00:13] Richard Webbe: point I that's was making [00:14] Dave Pengelley: gentlemen. [00:14] Richard Webbe: good. about the call [00:14] Dave Pengelley: Yeah, action. [00:15] Richard Webbe: center stuff [00:15] Dave Pengelley: I've actually [00:16] Richard Webbe: Is is that [00:16] Dave Pengelley: I've got one [00:16] Richard Webbe: crackable? customer [00:16] Dave Pengelley: uh I've got [00:17] Richard Webbe: service [00:17] Dave Pengelley: one of those [00:17] Richard Webbe: used [00:17] Dave Pengelley: on my [00:17] Richard Webbe: to [00:18] Dave Pengelley: phone, [00:18] Richard Webbe: be done [00:18] Dave Pengelley: I'm [00:18] Richard Webbe: by [00:18] Dave Pengelley: pretty [00:18] Richard Webbe: people face [00:18] Dave Pengelley: sure. [00:19] Richard Webbe: to face. [00:19] Dave Pengelley: Um, [00:20] Richard Webbe: And [00:20] Dave Pengelley: from [00:20] Richard Webbe: then [00:20] Dave Pengelley: what [00:20] Richard Webbe: we [00:20] Dave Pengelley: I've [00:20] Richard Webbe: moved [00:21] Dave Pengelley: been doing, [00:21] Richard Webbe: that to [00:21] Dave Pengelley: like [00:21] Richard Webbe: a [00:21] Dave Pengelley: camera [00:21] Richard Webbe: mass [00:22] Dave Pengelley: testing [00:22] Richard Webbe: area [00:22] Dave Pengelley: and stuff. [00:22] Richard Webbe: to improve [00:22] Dave Pengelley: I can [00:23] Richard Webbe: it. [00:23] Dave Pengelley: never [00:23] Richard Webbe: And [00:23] Dave Pengelley: remember [00:23] Richard Webbe: then it [00:23] Dave Pengelley: what [00:23] Richard Webbe: became [00:23] Dave Pengelley: that's called, [00:24] Richard Webbe: people on the phone. [00:24] Matt: Clapper, [00:24] Dave Pengelley: like the name [00:25] Richard Webbe: Now [00:25] Dave Pengelley: of the [00:25] Richard Webbe: we've [00:25] Dave Pengelley: app. [00:25] Richard Webbe: got AI [00:26] Dave Pengelley: No, I just [00:26] Richard Webbe: people [00:26] Dave Pengelley: searched for clap [00:27] Richard Webbe: responding. [00:27] Dave Pengelley: and it's not [00:28] Matt: can gonna see that we're you able be get it. to [00:28] Richard Webbe: It's [00:28] Dave Pengelley: there. [00:28] Richard Webbe: not like we've [00:28] Dave Pengelley: Oh, [00:29] Richard Webbe: lost [00:29] Dave Pengelley: I [00:29] Richard Webbe: that [00:29] Dave Pengelley: know, [00:29] Richard Webbe: service. [00:29] Dave Pengelley: right? It's [00:30] Richard Webbe: It's [00:30] Dave Pengelley: a [00:30] Richard Webbe: just that we've [00:31] Dave Pengelley: they [00:31] Richard Webbe: automated [00:31] Dave Pengelley: have [00:32] Richard Webbe: it all [00:32] Dave Pengelley: board [00:32] Richard Webbe: in a certain [00:33] Dave Pengelley: snapper [00:33] Richard Webbe: way, but [00:34] Dave Pengelley: board [00:34] Richard Webbe: we're probably [00:34] Dave Pengelley: clapper [00:35] Richard Webbe: getting So better [00:35] Dave Pengelley: snapper. [00:35] Richard Webbe: service, video. which is not getting [00:37] Dave Pengelley: No, [00:37] Richard Webbe: as much [00:37] Dave Pengelley: I [00:38] Richard Webbe: humanity. [00:38] Dave Pengelley: have no idea. And like you got like so many little windows. It's like which which which app pages on. I used to be really good at putting everything in folders, right? And then as technologies evolved and search got better, even on my desktop, I became a little bit less worried about folders within folders within folders. But sometimes you go, ah, I wish I was a little bit more focused. [01:00] Richard Webbe: Engineering [01:00] Dave Pengelley: On that. [01:00] Richard Webbe: you that does the folders within folders within [01:02] Dave Pengelley: That's [01:02] Richard Webbe: folders. [01:02] Dave Pengelley: the DOS. It's the DOS in me. Back in the old days where you had eight character file name limits, you kind of needed to have good folder structures in order to store everything. [01:12] Richard Webbe: Very true. [01:13] Dave Pengelley: Um [01:13] Matt: I I've gone the other way. I I've stopped foldering uh uh everything's flat for me now and yeah, like I've leaned heavily into search for everything. And I try and teach other like non-techie people this stuff. too because I get I see people struggling kind of like what you're doing right now but um and like surely there's a way to search that and if they haven't I don't know provided those keywords for their thing you know however the search actually works with apps I have no idea always got it [01:48] Dave Pengelley: And and you [01:50] Matt: oh [01:52] Richard Webbe: I [01:52] Matt: amazing [01:52] Richard Webbe: like that. [01:53] Dave Pengelley: uh it's good when you're trying to actually like when I've been setting up like OBS for recording and there is a delay on the mic and the video and you Do that and then you drop it into an app and you can count how many frames you can do the maths and go, okay, I need like a point seven second delay on my audio. And I can never work out is it which way to move it? Like, do I shift it backwards or forwards? I can I my brain can never comprehend am I taking point seven or adding point seven to the audio? I just but um can be really handy because you get the like as the colors line up, you get the the the audio spike and you're just gonna make sure they line up. Um [02:26] Matt: Yeah, [02:26] Dave Pengelley: that's [02:26] Matt: that's [02:26] Dave Pengelley: why [02:27] Matt: actually [02:27] Dave Pengelley: they use [02:27] Matt: cool. [02:27] Dave Pengelley: them. That's why they use them in film and TV, if anyone didn't know. Because they like this I don't know what they're doing nowadays, but certainly traditionally they record all the audio completely [02:35] Richard Webbe: By [02:35] Dave Pengelley: separately [02:35] Richard Webbe: default, [02:36] Dave Pengelley: from [02:36] Richard Webbe: I [02:36] Dave Pengelley: the [02:36] Richard Webbe: mean all [02:36] Dave Pengelley: video. [02:37] Richard Webbe: that Facebook [02:37] Dave Pengelley: And so [02:37] Richard Webbe: stuff [02:37] Dave Pengelley: in post, [02:38] Richard Webbe: was self-flagellation [02:38] Dave Pengelley: they need to line up [02:39] Richard Webbe: for the [02:39] Dave Pengelley: the [02:39] Richard Webbe: companies. [02:40] Dave Pengelley: two [02:40] Richard Webbe: They [02:40] Dave Pengelley: bits [02:40] Richard Webbe: were publicly [02:40] Dave Pengelley: of footage, [02:41] Richard Webbe: shaming [02:41] Dave Pengelley: especially [02:41] Richard Webbe: I themselves [02:41] Dave Pengelley: multi-camera [02:41] Richard Webbe: had a big by [02:42] Dave Pengelley: and stuff. [02:42] Richard Webbe: I giving had a people big the video ability week to this get week. on their My Facebook son was page invited and complain to about audition their service. for a It was like Tropfest a public film, uh which is [02:50] Matt: Yeah. [02:50] Richard Webbe: kind of big. He's he's a keen actor and he's pretty good at it too. And he's done a few you look on my Facebook page and see one of his ads. And so my idea of doing an audition for Him was to um set up the all the text, get words of people's names on bits of paper and stick them on the heads of seats, and he could walk in the room and act as though all the actors were there. And and um this took me about an hour of him complaining to me, Dad, this is not how it's done. And he said, Dad, we just do additions with a white background, please. So we went back to that. But it was very funny, the fighting and the so-called, but he's a good actor, my son. It's amazing. I don't trust him anymore when I asked him. question I think he's acting it's [03:32] Matt: True. [03:32] Richard Webbe: like back to team america don't trust them they're acting [03:37] Matt: Oh, [03:37] Richard Webbe: I knew [03:38] Matt: Team [03:38] Dave Pengelley: So [03:38] Matt: America. [03:38] Dave Pengelley: the [03:38] Richard Webbe: like that preference Dave [03:39] Dave Pengelley: We're not we're not quite we're not quoting [03:41] Richard Webbe: Yes. [03:41] Dave Pengelley: Tim America on this show. [03:42] Matt: I never thought about what hear a Team America reference out of you. That's that's um impressive. [03:47] Richard Webbe: I love that film the best bit was uh john uh Kim John Hill singing I'm so lonely with a W. It was [03:54] Dave Pengelley: Now [03:54] Richard Webbe: awesome [03:54] Dave Pengelley: the the the thing we quote the mo the most often is I think he's saying kiss me, kiss me. [04:02] Richard Webbe: Okay, everything AI. What was the big thing about AI for you this week, Dave? You said you had a lot happening. [04:08] Dave Pengelley: Well, you said the letters, you know what that means. [04:24] Matt: Actually have no idea what happened this week. What did happen? [04:30] Dave Pengelley: opinion and other bits and pieces before we get to the news. Um it's actually interesting on the topic of AI and AI for good and AI for society, which I want to talk about a little bit today. And we won't play clips from them, but I watched two different interviews, one with Sam Aldman and one with Elon Musk talking about their views on AI and the future of AI. And it's really interesting hearing their different points of view. And I haven't watched anything of Dario's, but obviously the big thing from them lately was you're not allowed to be mean and abuse the AIs in case you hurt their feelings. [05:01] Dave Pengelley: Which we'll we'll we'll we'll loop back into. But just on the um some of the good that's coming from these advanced technologies, I was at a um meeting on Friday night. It's a group called Faith Tech, it's sort of Christians and technology and stuff. And they had this um guy from a group called Destiny Rescue come, and they rescue people from human trafficking and child trafficking, like really heavy stuff. But they he told told the story about this one girl who was trapped, this like young girl, like five or six. It's really tragic um circumstance. But they had a photo of her in front of a green fence, and they knew she was in Manila. And so they spent two years of volunteers and workers, Google Maps, traversing all the rabbits, then Maises, Manila, until they finally found this green fence. And then they were able to go in, rescue the girl, arrest the uncle and the mother, and find all the computer details, and you know, launch a whole bunch of investigations all to all the people that were. Consuming the media and huge, but took two years to rescue this girl. [06:04] Matt: Wow. [06:05] Dave Pengelley: Now, fast forward a little bit with some of the new advances in technology and image recognition and so on. They had another case where they had this picture of this girl on a boat and just random boat with some coastline behind her and some, you know, masts and sails and whatever. They found her in a week. [06:24] Richard Webbe: Wow. That is and that's t that's the part of IOI. These dipsticks complaining about data centers and AI don't talk about it's the advancements for human benefit and things like that. I know that I think I might have reported it last week, that what would take up to a year to two years to diagnose a young lady with endodimetriosis could now take 12 seconds. [06:54] Dave Pengelley: Yeah, so I think I thought that was a really interesting um use case of Where these tools are being applied for good, because there is always the dooming stories and the issues, and um that can go in either direction. But I just want to start with, you know, that was a really positive story of this technology actually helping people. Because unfortunately, there's a lot of cases of this technology helping the nefarious people hide from authorities and move money and crypto and all these other bits and pieces. Um so I mean, they they talk about how do you redeem the technology, how do you how you use it to do good things. Uh yeah, so I thought I thought that was really interesting because uh And that leads me into the difference between Sam and Elon. And I know obviously Elon thinks very little [07:36] Richard Webbe: Did [07:36] Dave Pengelley: of [07:36] Richard Webbe: you remove [07:37] Dave Pengelley: Sam if [07:37] Richard Webbe: all [07:37] Dave Pengelley: he [07:37] Richard Webbe: my [07:37] Dave Pengelley: thinks [07:37] Richard Webbe: swear [07:38] Dave Pengelley: of him at [07:38] Richard Webbe: words, [07:38] Dave Pengelley: all. [07:38] Richard Webbe: Dave? [07:38] Dave Pengelley: Um, you know, calls him Scam Altman, etc. Obviously. [07:41] Richard Webbe: Didn't I start opening IR together though? [07:43] Dave Pengelley: Well, that that's the thing, right? And then Elon claims that Sam took it in the wrong path and cut him. It's all very, you know, he said, she said toddlers throwing their toys at each other, kind of thing, a little bit. Feels like it from the outside. [07:56] Matt: It's [07:57] Dave Pengelley: Um [07:57] Matt: weird. I don't know if we'll ever actually truly know the story. And like I I've I've had my tinfoil hat before where I think that all of these things are very orchestrated where, you know, let's let's drum up this kind of drama because it will cause this opportunity. Like [08:11] Dave Pengelley: But but I think I think it's actually interesting. Listening to both Sam versus Elon in interviews talking about AI, the future of AI, what it's for, humans' role with AI versus under AI, etc., they have philosophical disagreements about the future. Future of AI and where it sits in the sort of food chain of humans. And I think that would [08:35] Richard Webbe: Well [08:35] Dave Pengelley: explain [08:35] Richard Webbe: done. [08:35] Dave Pengelley: a lot of the differences. [08:36] Matt: Yeah. [08:37] Dave Pengelley: The fact, the fact that they have underlying different theses. Like from for instance, Sam Elon, he strongly believes at some point [08:47] Richard Webbe: It is [08:47] Dave Pengelley: the [08:47] Richard Webbe: who [08:47] Dave Pengelley: computers [08:47] Richard Webbe: you are, [08:48] Dave Pengelley: are [08:48] Richard Webbe: Dave. [08:48] Dave Pengelley: so smart [08:48] Richard Webbe: I agree. [08:48] Dave Pengelley: that [08:48] Richard Webbe: I [08:49] Dave Pengelley: we will [08:49] Richard Webbe: think [08:49] Dave Pengelley: just do [08:49] Richard Webbe: your [08:49] Dave Pengelley: what they [08:49] Richard Webbe: ability [08:50] Dave Pengelley: say, [08:50] Richard Webbe: as [08:50] Dave Pengelley: or [08:50] Richard Webbe: an [08:50] Dave Pengelley: they'll [08:50] Richard Webbe: interviewer, [08:51] Dave Pengelley: think of us. [08:51] Richard Webbe: an [08:51] Dave Pengelley: He [08:51] Richard Webbe: executive [08:52] Dave Pengelley: said, [08:52] Richard Webbe: producer, [08:52] Dave Pengelley: like, [08:52] Richard Webbe: and a narrator [08:53] Dave Pengelley: we [08:53] Richard Webbe: is [08:53] Dave Pengelley: are [08:53] Richard Webbe: second [08:54] Dave Pengelley: we [08:54] Richard Webbe: to [08:54] Dave Pengelley: we [08:54] Richard Webbe: none. [08:54] Dave Pengelley: don't try to to to micromanage chimps, we just let them run around in the forest and then they can do whatever they like. Like AI will probably Like, not really be too concerned about us. It's so super intelligent, we'll let it run things and it will just let us exist. But he sees this future of you know the AIs basically running the show, humans semi-subservient. [09:12] Richard Webbe: But you you you [09:13] Dave Pengelley: Um, [09:13] Richard Webbe: may [09:13] Dave Pengelley: he [09:13] Richard Webbe: need [09:13] Dave Pengelley: doesn't [09:13] Richard Webbe: to get [09:14] Dave Pengelley: seem [09:14] Richard Webbe: a [09:14] Dave Pengelley: locked [09:14] Richard Webbe: haircut. [09:14] Dave Pengelley: in on that. He did he did vacillate a little bit on the role of humans, um, but generally thinks you know, AIs will become so stupendously intelligent that we'd be stupid to not give them the keys. Versus Sam, and I and I like the the really clearest comparison. And then I'll let you guys chip in that Sam [09:32] Richard Webbe: Yes, [09:32] Dave Pengelley: made [09:32] Richard Webbe: please. [09:32] Dave Pengelley: is he thinks that all this AI stuff, he doesn't want it to be a new industrial revolution, he wants it to be a new renaissance, and that the AI should always be for humans and it's always around the humans, and this benefits humans versus is better than or is controlling or taking over. So very different [09:50] Richard Webbe: Entry [09:50] Dave Pengelley: philosophical [09:50] Richard Webbe: point is [09:50] Dave Pengelley: points [09:51] Richard Webbe: low. [09:51] Dave Pengelley: of view between the two. [09:52] Matt: Do you reckon there's any lens [09:55] Richard Webbe: Yeah. [09:56] Matt: that needs to be removed from either of their statements to [09:59] Richard Webbe: Priorities. [09:59] Matt: do with Actual PR and VC standing. [10:05] Richard Webbe: Always. They're both they're both they're both very shoot from the hip. I don't think I sometimes think they plan their PR and they have teams that do that. Yeah. But then again, I think when you look at Elon with, you know, his self-admitted ADHD autism and free speaking free thinking spirit, uh, Yep. he will always say things that And are running that means through you'll his head be more before empowered, he's running through I a filter. think. And I think that's kind of exciting in a way. I think I have a bit of that personality myself. And I think that's good because we get a bit of honesty from him. I mean, I love his interviews. I mean, he was talking the other That's day the about human condition. how he thinks money's gone. Money will I go agree. and value will be exchanged in another way. Because if you think you're, he says, if you think you're going to Great work section. for money for the value of the Yeah, knowledge you have, thanks. that won't be valuable anymore. And I can see that collapsing in certain areas around the world, in some cases, like with lawyers and the like. Oops. Um, it's a good thing, right? But all AI still needs is humans to guide it. [11:06] Dave Pengelley: It it comes back to what we talk about all the time around creativity and taste and the human element, uh, that the AIs, at least in the shorter immediate future, just do not have the ability to do. [11:20] Richard Webbe: No, [11:20] Dave Pengelley: Like [11:20] Richard Webbe: not for ten or twenty years, agree. [11:23] Dave Pengelley: I mean, I might wouldn't wouldn't put a a time time horizon on it, because that it's moving very fast. But But yeah, I think like and I think that's why I like Sam's idea of the Renaissance thing. And for me that that really fits, right? Like I'm the amount of different things I'm doing, being able to create really interesting different websites and applications and um video games and comic strips and all the other content and things that I do that were just out of reach for me. I had these ideas and was always frustrated that I didn't have the skills, ability, or team to execute them. And now I can actually make headway on some of these things. Um and if many humans Uh that's one one of the reasons we're on the show. That's why I do field notes to try and innovate and inspire like people with creativity because you can do so much and people are just getting it to rewrite their emails or check their inbox. [12:11] Matt: Oh, that hurts so much. Uh, I still think like it'd be so good for us to to create like a a tiering. You know, it's almost like the um almost like the OSI layer of of um people's understanding of what AI actually is. Like I [12:27] Richard Webbe: Oh, [12:28] Matt: don't [12:28] Richard Webbe: I think [12:28] Matt: think [12:28] Richard Webbe: it's on my model. [12:30] Matt: Yeah, like you know, like you you your very first interaction is you Google search or you know you're you're on a web browser and you're searching in a search engine and there's an AI result and you're like, oh what's this? And you might be able to add like a follow-up statement there. Or you hear of people talking about JPD and Claude or whatever, and you download it and you know, you start trying to do stuff there, but you don't really know what you're doing. Like those people, like the comparison to that, or like the delta between that and and people that are actually using their agents to Do real work every day autonomously, it's almost incomprehensible to those people. And I find like that shift is going to continue to expand as well. Even with these realities that Altman and Musk are talking about, there's going to continue to be that crazy divide, like the people that really understand it and the people that don't. It's kind of like if you take um, if you take 10, like a spattering of 10 different uh current generation teenagers and you give them all Uh an M5 MacBook and it's like go go do a thing. Uh, you know, the differences in results there are going to be huge. Like you said, Dave, about like taste and everyone's individual opinionated approach to things. You know, if we all have the same tools, if we all have that sumnet average, Richard, of the AI that's available to us, regardless if it's super intelligence or not, there's still going to be that opinionated approach from the individual. That's where the value is gonna come from. [14:02] Richard Webbe: And I agree. And I want I I would like people who watch this. I I think like this all the time. I I know you do, Matt, and I I know Dave does. I'm not sure everyone does. A lot of people, when it comes to AI, shut down their comprehensive comprehension and look at it like a black box scenario. It's this magic in the black box. And I want to unbuckle that for everyone so they understand what's going on. It'll be less fearful and they'll know how to apply that creativity that we're Alluding to. So I want people to think for themselves when they've got a problem, how do they solve that problem? Well, they pull in data, they run scenarios in their head, and they go, you know what? I'm going to do it that way. And that's all AI is doing based on the data it's got, the capacity it's got to focus and it's focus and away it goes. And if people start to think like that, they'll see how it can be like a third brain or an extension to their body. And I think that's the really important one. [15:00] Matt: Some people just don't know how to think though. That's that's the problem. They don't they don't it's unconscious competence. They just do what they do and they continue to do it. And if you try and tell them otherwise, they get confused. [15:10] Dave Pengelley: I [15:10] Matt: Like, [15:10] Dave Pengelley: mean look at look at shows as as overproduced as they are, but look like like for maybe it's a bad example. Maybe Lego Master is a better option than the block. But you give people the same task with the same tool set and look at the different results you get. If you gave five blokes a set of Makita tools and and a pallet of wood, you're gonna get five completely different. [15:29] Matt: you know. [15:30] Dave Pengelley: Results of how they solve the problem and what they build with the same tools and the same materials. [15:35] Matt: Yeah. [15:35] Dave Pengelley: Um, and and AI, I think the thing is like AI has so many millions, billions of data points to leverage, it's not going to bring all of them to bear on every single problem. That's where things like the weightings come in, like and the probabilities of I should grab this one next or that one. Just like humans, you know so much accumulated knowledge, and in any given situation, you will not leverage everything you know because not all your memories will come to the fore, your brain won't. Connect all the neurons in the right order in the time frame you need in order to give you the optimal thing every time. Biases exist for that reason to sort of shortcut things and make make it easier for you to mostly get it right most of the time. Um, but you know, then when you reflect or you someone else says, But what about such and such? Then your brain goes, Oh, bang, bang, bang, bang, everything clicks into place, and you go, This is what it should be. Um, and the AI is no different, except the AI just is running a much with a much bigger data set. And so the probability of it getting the right five things every time. Is actually from a taste point of view, even less. If you say make me a website, like it's got so many websites to choose from. [16:39] Richard Webbe: I love doing that with a good LLM. Give it a very vague instruction and look at the variations on what it'll come up with. I just think that's exciting. [16:46] Matt: So uh there's uh there's something really fascinating that um I can't remember who it was. Somebody said something recently that uh we all talk about hallucinations, you know, regardless of if it can find the correct context or not, like you know Uh it makes me think that you know we are we're the biggest hypocrites here. You know, we talk about context rot and you know various ways of context management with with the genetic engineering. And what Dave just described is like pure context rot in in the human brain. Like, do you remember what you had at lunch in year five in you know, August? Uh like no. Like the the it's impossible. And we we have so much expectation on these these agents and these models to actually do that stuff. But Richard, what you said. When you when you ask a vague instruction, the thing that has reduced or very, very much eliminated hallucinations is the ability for these agents to actually verify, to actually check. You know, you you mentioned what do you do when you go to do something? You check the data. Well, what does the agent do when it answers your vague question? It's like, what could they be asking? What do I have access to? How can I go check this? And should I ask a follow-up question first to clarify? You know, that that kind of stuff is insane. And As soon as you expose the right tools to allow it to get the right information, all of those billions of parameters, Dave, don't necessarily matter anymore because it can just go find the information it needs. [18:13] Richard Webbe: Oh, by the way, in August in grade five, I would have been eating a meat pie. That's all over eight when I went to school. [18:19] Matt: Okay, yep. [18:19] Richard Webbe: Just so you know. [18:22] Matt: Perfect. Actually, maybe me too, actually. I had them in in primary school. [18:26] Dave Pengelley: No, no, [18:26] Matt: People [18:27] Dave Pengelley: I [18:27] Matt: used to [18:27] Dave Pengelley: I'm [18:27] Matt: peel the [18:27] Dave Pengelley: I'm [18:27] Matt: top of their meat pie off and like flip [18:29] Dave Pengelley: Yeah. [18:29] Matt: it upside down. Yeah. [18:30] Dave Pengelley: Yeah. [18:30] Matt: So, [18:31] Dave Pengelley: Uh I'm I'm a sausage. I I don't mind a peat [18:34] Richard Webbe: You're [18:34] Dave Pengelley: pie [18:35] Richard Webbe: the [18:35] Dave Pengelley: occasionally. [18:35] Richard Webbe: guy. You're the guy. [18:36] Dave Pengelley: I'm I'm the sausage all guy. Um no in in in primary school, I mean I was actually bad at having regular lunch anyway. Um, or if I take sandwiches, I wouldn't always eat them, and then they'd end up my lunchbox thrown under my bed somewhere, and you'd find a [18:52] Matt: yeah. [18:52] Dave Pengelley: month later, I was a terrible child. Um yeah, [18:55] Richard Webbe: So here's [18:56] Dave Pengelley: so [18:56] Richard Webbe: an interesting [18:56] Dave Pengelley: great. [18:56] Richard Webbe: one. So he's an interesting one. So Microsoft have announced That they're laying off 27% of their workforce, almost 60,000. And everyone's wanting to say it's AI. And I don't think the Microsoft press release said it's all AI, but certainly they're saying it's to do with efficiencies and things like that. So let's everyone's going, oh my God, I'm going to lose my job, and everyone's panicking. But have a look at the amount of jobs created for business and individuals helping businesses complete their digital transformation. It would exceed that 60,000. by a factor of four, I would think. And it's just the menial tasks that are done in the background that are getting off. Like again, as I always say, we don't have the typing pool anymore, but we've still got as much employment and still got as much quality of [19:43] Dave Pengelley: But [19:43] Richard Webbe: life. [19:44] Dave Pengelley: there's there's a transition period with that, right? That might be true in the long run, but in the short run, the new jobs haven't been created yet. People haven't worked it out, like we're just talking about with the innovation and trying new things and building new ideas with the tools and reskilling and so on. That doesn't happen overnight. And so you can get This gap in the middle of these transformations [20:03] Richard Webbe: Yep. [20:03] Dave Pengelley: where jobs vanish before new jobs are created. Um, so [20:08] Richard Webbe: Yeah, that's true. I ag I I totally agree. And that transition period can be very tough. [20:14] Dave Pengelley: yeah. Yeah. So it's it's one of these things where I mean like I I like I said, I think the role of people like us is beyond, you know, getting paid work to implement and sell AI and do other bits and pieces, but it is that the role of a podcast is just uh Try and get people thinking bigger around these topics and not to be scared of them and to try things and to experiment and to challenge the AI to do more. I saw someone, I've seen a few posts lately, basically saying if you're still asking Claude or whatever to prompt things like you did two years ago, you're not being ambitious enough. Just ask for bigger things. Just [20:50] Matt: Yeah. [20:50] Dave Pengelley: just ask it to do more because it can. [20:55] Dave Pengelley: Like yesterday. [21:00] Matt: The bigger [21:00] Dave Pengelley: We'll [21:00] Matt: picture [21:01] Dave Pengelley: just [21:01] Matt: rather than the inter the granular piece. [21:03] Dave Pengelley: We'll [21:04] Richard Webbe: That is [21:04] Dave Pengelley: just dream [21:04] Richard Webbe: so [21:04] Dave Pengelley: dream [21:04] Richard Webbe: true. [21:05] Dave Pengelley: dream it and see how it goes and then iterate from there. But but don't don't shortcut it and go, I think I can only ask for this. Like [21:11] Matt: Correct. [21:11] Dave Pengelley: ask for the big thing or ask it to build the plan for the big thing and then break the plan down and execute it. But but go bigger. And I know this feels like they Dave's always trying to plug his own stuff. Um you guys can plug your own stuff too. But um that's all can it's it's all contextual. So I I I launched I launched I launched Tin My Games last week. I I sort of previewed a few bits and pieces over the last few shows, but I actually went live with three games on itch.io last week. And I put up together a basic website. [21:44] Richard Webbe: I saw [21:44] Dave Pengelley: And [21:44] Richard Webbe: that. I did [21:45] Dave Pengelley: the [21:45] Richard Webbe: say [21:45] Dave Pengelley: web [21:45] Richard Webbe: that. [21:45] Dave Pengelley: the website, tinmygames.com, sucked. And compared to all my other web properties, and you guys know I put a bit of effort into animations and different looks and feel and just polish on my websites. This wasn't up to standard. Uh, and I'd seen these things saying, just ask for more, I'll do this, do that. So I was like, you know what? What? I want the site to feel a bit more like a game. I want you to use assets from the video games. Um, you know, Timmy Games, let's let's make the homepage a kennel, and you go into the kennel, and then the experiences inside the kennel, right? And so I kind of one-shot that last night. And it went and looked up the repos, pulled the video game assets, did all that work. And so I went from a really plain, boring website to now I've got this guy. And it's fun, it's funny, the little helicopter flying across the top was flying. Flying was flying backwards. So talking about the AI, not really know what's going on. [22:34] Richard Webbe: Let's get context [22:36] Dave Pengelley: All [22:36] Richard Webbe: right. [22:36] Dave Pengelley: the helicopter sprites were actually mirrored and just like flying backwards. It was hilarious. But yeah, so now I can I go in and zooms in, and then I've got more little assets just flying around on the on the pages. And just [22:50] Richard Webbe: I [22:50] Dave Pengelley: like [22:50] Richard Webbe: think and [22:50] Dave Pengelley: is [22:51] Richard Webbe: I think that's wonderful, Dave. That's a great story and very indicative [22:54] Dave Pengelley: is it perfect? [22:54] Richard Webbe: of what's [22:55] Dave Pengelley: No, [22:55] Richard Webbe: happening. [22:55] Dave Pengelley: but like it's it's just it now feels a little bit more like it's on theme with what the What the games are, the whole thing, the helicopters fly the right direction now, paratroopers are dropping in. Like it's just, yeah, I think, but I was well prepared to say, just do this. Like, just give it a go. And it just went and did the work. So I'm trying to bring us back to camera. It's planner. Uh and that was Grokbot. Grokbot did all of that last night while I slept. Just [23:21] Richard Webbe: Go [23:21] Dave Pengelley: amazing. [23:21] Richard Webbe: drop. Well, don't tell everyone that just tell them you were slaved, worked hard. You know, they say, you know, you guys know I come from an artistic family. uh not autistic family but maybe that too and um people connect over creativity that's how we connect people don't realize that we do you go to an art show and you look at paintings and it evokes something in you that the artist was thinking about and all of a sudden you're connected to the artist by looking at what they created when we're using AI to create either entertainment or productivity the same thing happens and it's kind of a magical process I think it is probably we're probably heading towards the most fundamental raw natural creativity we could imagine because the production of what you want to create is being moved and it is pure and applied creativity of what you want to achieve. I think that's just an exciting time in in human history. [24:18] Matt: That's actually a really fascinating point. And I I think I was connecting with what you were saying there. It it sounds like you were saying that, you know, previously you needed a level of skill before you could actually play the game. Before you could actually engage your creativity, you know, because creativity kind of requires unconscious competence, um, you know, so that it's intuitive and then you can you can just feel it out and make it better and better. So it's like when you're learning guitar, you know, you have to think about one hand and then the other hand, and then what does this finger do? But then when you don't have to think about that anymore, it's kind of like touch typing. You then are just doing it by ear and by feel and and you know, how is that kind of resonating with you at the time? And yeah, i that's what you're saying, right? [25:03] Richard Webbe: That's exactly what I'm saying. [25:05] Matt: That's interesting. I like that. I really like that. [25:08] Richard Webbe: Dave, Dave just demonstrated it. I looked at Dave's creativity then, and I don't know what games those little parachute people and helicopters look like, but I identify with them and I can remember playing games in my past that looked like those characters and did stuff like that, whether it was Load Runner or you may remember Load Runner and It's all got that same pixelation, so immediately when I looked at your website, I go, Oh, this looks like a place I belong. [25:35] Dave Pengelley: Well, and then those those three launch games were based on old 80s type arcaders like Paratrooper, Sopwith Camel, um, yeah, Lunar Lander. Like they were they were inspired by those old things. Now the next game I'm building, um, and you talk about you know the AIs, I'd back and forth and planned out a whole bunch of stuff and then went and built a thing. It was terrible. It's taking me days and days of iteration and back and forth and break this down and fix. That and know that that interface sucks. We need to fix this, make that anime like that, make it look like this, um, in order to get something that's now starting to feel like it's got some meat on the bones and feels really good. And even then, still needs so much like you, you, UX designers and user interface designers are gonna be in high demand for a long time because it's something the AI doesn't innately do well, understand how humans see things, even visual models and stuff. Like you give them a video clip of something, and they've got to try and break that down frame by frame. Interpret what's happening frame by frame, that understand the feel of it from a human, from a timing, from a how long it takes us and our eyes to see something and and process it and recognize it and use it. And so I think I I know Matt, sometimes you think, well, all the user interface is just gonna be the chat going forward. Um but not always, right? Um yeah, so I think that that comes to that cognitive understanding of design skills and user understanding, that user experience. I think that is an area where humans and from a development designer, you might not need to know how to write the code anymore, but you certainly need to know how to guide it, instruct it, and coach it to get things where they need to be. [27:14] Richard Webbe: And it comes back to libraries in context and MCP and all that sort of stuff. I know there's a number of organizations now using visual AI to read the faces of the people, either at the entertainment they're at or at the company or where they are, to read their feedback. I can say, Is this pleasing the humans around me? And that takes it to another level. [27:36] Dave Pengelley: I I saw some of that sentiment analysis stuff like a decade ago for like expo booths where someone had a little camera at like your trade show and it was looking at all the people that were watching the little thing happening at the trade show and it was like trying to gauge sentiment analysis on are they engaged, are they happy, are they dissatisfied, in order to feed that back to to what you were doing. It was a bit of a tech demo. It was early days of sort of machine learning visual um categorization. But all these things like I was working on on an actual getting away from games. I was working on a a business um uh request and going through some technical issues and had to get back to the back with a sort of short presentation to this um this pro prospect around solutions and the easy thing to do would be say hey agent here's the here's the discovery tell me what to do but I'm like that's not gonna give me the result I want so I gave it the the the document but then all Also gave it a whole bunch of commentary from me, like just turned the microphone on, Matt. And it was just like, uh, yeah, I think this. And uh uh based on the thing that they're saying there, maybe we could go in this direction. And so I actually, you know, applied some my human learnings against it. Uh and then it goes, okay, but think about that. And what about this? I was like, okay, cool, that extends that. And I I like that aspect, but uh no, I think we need to loop back, and this is the main thing, the main thing. And then cool, came up with a spec. And I tried three or four different things to try and build the sort of the deck for it. And none of them are complete. Like none of them are right. Too many words, or they miss the point, or they're they're just. And so again, from a human creativity and human communication point of view, these tools are great and they're an accelerator, but they're not 100% there yet. And if [29:15] Richard Webbe: They're [29:15] Dave Pengelley: you're trying [29:15] Richard Webbe: not [29:15] Dave Pengelley: to [29:15] Richard Webbe: human. [29:15] Dave Pengelley: communicate clear [29:16] Richard Webbe: They're not [29:16] Dave Pengelley: messaging, [29:16] Richard Webbe: human. [29:17] Dave Pengelley: you need to still be the expert on how to communicate as a human to humans. [29:21] Richard Webbe: Experience creativity context. That's what it's all about. I mean, you could have one of those image things looking at people's faces to see if they have And they'll say everyone's happy, and really most people were just constipated, and it wasn't a real smile. You still need a detail. Oh, it's a cheap joke, sorry, but you still need the detail about what that experience is behind about what's happening. [29:45] Matt: Context, context is king. Like I uh I find it really, really interesting that we've sort of gone we've had a regression. I think like all of us that use these tools, regardless of like what level you're using them at, we've had a Regression on our our understanding of them, and we've increased our expectation. Like that continues to scale. It's almost like every every new success or every new little like piece of um magic that we see that they can do, our expectation has now just reset the thermostat. And you know, every every little gripe that you have with it is so much worse than what you used. to uh you know doubt it for or whatever. Like I know myself, I fall victim to this all the time. I get so stressed and so angry and I just sit here yelling through my voice to text. And you mentioned before about yeah you mentioned before about uh Dario you know talking about don't hurt the machine's feelings. I it's it's for me it's not it's not so much abuse. It's just listen to what I'm saying. Like like [30:57] Richard Webbe: Yeah. [30:57] Matt: we've been through this and you know the whole context is King, uh I I don't understand [31:04] Matt: I don't understand why we keep forgetting that. Like I don't understand why we expect this thing. Like you said at the very, very beginning of this, Dave, like how you know we have that context rot over time. And there's almost no way to solve it. And it's the same with people. You know, like I I could go and ask the perfect example actually is my partner coming and asking me about that thing that we talked about last week, and I have no idea what it was. You know, and she said, Are you gonna do that thing? And I'm like, What thing? You know, so and I'm sure many of us can relate to that. Like [31:37] Dave Pengelley: No, [31:37] Matt: that that's the kind of expectation though that we [31:40] Dave Pengelley: no, [31:40] Matt: have [31:40] Dave Pengelley: no, no. [31:40] Matt: on these machines. It's like, why don't you remember this? And it makes you so angry, but you know, this is where we need to try and think of other ways to enhance this context, remind the machine about what you're doing, [31:54] Richard Webbe: So let's let's go back working because at the end of the day, for the human condition in my humble [31:59] Matt: yeah. [32:00] Richard Webbe: opinion. It is about having challenges and solving them. That's life. And those challenges are around keeping the human race going, being loving and caring and kind to the family and the people that are important to us and filling the idle times between those concerns with fulfilling pursuits. That's why we do sport. We create a challenge and then try and solve it. It wasn't one that's going to save the planet. And that's what sport really is and that's what creativity really is. And I think if people go back and think about how Live 200 years. They probably worked 150 hours a week just to stay alive. [32:36] Matt: Yeah. [32:36] Richard Webbe: Now you can work 20 or 30 hours a week to stay alive, and the rest of the time is filled with more creativity, more fulfilling, and almost um more charitable exercises. And I'm hoping that's the way we're [32:50] Dave Pengelley: This [32:50] Richard Webbe: going. [32:50] Dave Pengelley: is this one of the things that Sam was saying in the interview around like it's possible if we were happy with a lifestyle that people had a hundred years ago, we could get away with working one. Or two hours a week versus what they had to do in the sort of feudal um subsistence farming kind of environments. But you know, we have uh this extra time, so we use it, we maximize it. And talking about that scale of people that are using AI and the people at the very top, they're not getting a lot of they're not getting a lot of extra time back. They're just doing a lot more. [33:20] Matt: Yeah. [33:20] Dave Pengelley: Um, we all fall into that category, right? Like we're working nonstop. And it's not that like, yeah, the AI, we can save time, so we go, cool, I've got more time. What can I do with that? Like, how can I be more creative now? Like [33:32] Richard Webbe: Yeah, [33:32] Dave Pengelley: what [33:32] Richard Webbe: you work [33:33] Dave Pengelley: what [33:33] Richard Webbe: hard. Has [33:34] Dave Pengelley: yeah? [33:34] Richard Webbe: any ever read Michael Crichton's um Timeline or Time Zone? It's a really [33:39] Dave Pengelley: Nah. [33:39] Richard Webbe: great book because it gave me, so he's a guy that did ER and all that sort of stuff. [33:43] Dave Pengelley: Jurassic [33:43] Richard Webbe: But he and [33:44] Dave Pengelley: Park. [33:44] Richard Webbe: that yeah, but no, it wasn't, yes, but it wasn't the the book that he wrote, I forget that it's it was a time something, and they made a movie of it which was rather B-grade. But what when I started reading the book, what got me so dramatically involved is humans. weren't stupider. He he he's written about how people lived in medieval times. Humans were actually really smart. The only thing they didn't have was technology. And it gives you a whole new perspective because you think, oh, if I went back in time, you know, 500, 1,000, 2,000 years, I'd be really wealthy because I'd be the smartest person now. No, you wouldn't. And you see how the and he writes in detail in these books about how people used primitive technology to achieve the same things we achieve today, with a lot. Less efficiency. And it was, it was, I found it such an eye-opener that we are not different to the people who were alive 30,000, 100,000 years ago. The only difference is we have technology. And that's [34:41] Dave Pengelley: Well [34:42] Richard Webbe: been the change. [34:42] Dave Pengelley: one one of the big things that we have is is, you know, this thing that I've got shining on me right now. Uh artificial light is actually a huge, massive game changer. I did um there was a a charity that made little solar kits to give light for um people in Countries where they don't have a a main grid everywhere. Um and some of the stats they talk around around how if you give people a light source that they can then do homework with at night, they can do with they can cook at night, it's a complete game changer for them. Um [35:13] Richard Webbe: Yeah. [35:14] Dave Pengelley: completely. Speaking of game changers, I'm gonna have to jump off for a second, go get another package. So you guys keep talking, and I hate to do this. Sorry, live show, gotta do it. [35:23] Richard Webbe: He's got the the doorbell's come and done it to him. You know, speaking about light, I mean cooking and Pre-digesting our food, which is what cooking actually is, is what allowed our cranials to grow, our brains to grow so quickly, once we learnt to not just be, you know, foragers. And you look about basic inventions, Matt, one of the ones that I saw that got a worldwide recognition was a guy in Africa in all the uh all the lean to, what do they call the shanty towns. Now, what you don't realize is when those poor people are living with corrugated iron for houses and it stretches like a deck of cards. Used to do as kids for miles and miles. One of the problems I have in there is light. And this one guy was collecting all the leftover um uh uh soda bottles, filling them with water, drilling a hole in the corrugated iron because it's so dark in these places, right? It's not like you can put a skylight in or you can run electricity. And he he won a World Uh Invention Award for this. So he cut a hole the width of one of those, and he just stuck it in the roof and stuck it there with water. In it and the top sticking out. So then the light hit it, and the water on the round bottom of the soda pop bright distributed the light, and it was just amazing. And such a simple thing. [36:40] Matt: That's clever. Like that. [36:43] Richard Webbe: Totally. I thought it was amazing. It won some World Invention Award from there. So we're cracking along with AI, companies like Microsoft. And I find that's the other thing. There's a few data companies that are getting into a bit of trouble recently, I know, because of the Overestimating earnings. We've got the government trying to shut things down. We've got more AI software tools appearing every day of companies from Silicon Valley coming into all over the world to harness and help and grow all the AI. One of the interesting things I find is how long before those tools are absorbed in a more bigger macro, right? And everything just becomes ubiquitously accessible. [37:27] Matt: Yeah, I [37:28] Richard Webbe: What [37:28] Matt: was having [37:28] Richard Webbe: will [37:28] Matt: this chat recently about Um, you know, people say SaaS is dead or you know, software is is no longer the moat, you know, if everyone can create whatever they need, like you know, where where is that next plateau? Like what do you aim for? Uh like because the distribution doesn't really exist. And you know, if everyone is using the same tool, then is it just the people that own the tool that are the ones that are gonna succeed? Like, you know, where is this next kind of economy, you know, this next series of businesses and how are they all gonna start working together? Cause like uh I I literally speak to people every day. Like my my prospects are the ones that are people that are started to build something or they have some sort of business that they're running on and they want to do something different, but they're petrified, sorry, petrified, not patrified, um, about where the industry's going. They're like, well hang on, if I keep going down this path, am I just gonna hit a dead end and am I going to get absorbed by everything else around me? Like curious to know what you guys think about that and how your what your kind of takes are. [38:33] Richard Webbe: I don't think it's going to happen that quick. I don't think it's going to happen that quick. First of all, let me ask you quickly, Matt, how's that new keyboard of yours going? [38:41] Matt: It's good. [38:42] Richard Webbe: Yeah. And you're used to it now. And we're talking about most people, you know, using voice to text and visual to program their environment and how they do things. And I'm I'm watching people, including myself, wake up in the morning and have my phone talk to me about all the all the emails I've got coming in. That's evolving. Slowly, right? I ran an AI company ten years ago before people could spell AI. And really people could not conceive about what that joke still works, doesn't it? [39:09] Matt: That's [39:09] Richard Webbe: So [39:10] Matt: never [39:10] Richard Webbe: people [39:10] Matt: gonna [39:10] Richard Webbe: could [39:10] Matt: not [39:10] Richard Webbe: not. [39:10] Matt: work. [39:14] Matt: That's [39:14] Richard Webbe: So [39:14] Matt: good. [39:14] Richard Webbe: uh so uh it's part of our new talk, Dave. So um so it's changing slowly and it is gonna have a big impact, but it won't be like I get up and go to work tomorrow and they say, sorry, you are no longer leader. Needed and people like you are going to be expunged from the earth because you're a useless person and you've got a useless skill. A bit like we say to our old people, it is not going to be like that, right? I'm glad you got that joke down. It's going to be a slow evolution. So, what's going to be more fulfilling for people? Keep doing what you know best. Keep thinking about how you can improve what you're doing. And then as you see things changing, think one step ahead and go, gee, I wonder how could I? I improve that? Like you know what I mean? I mean, no one wants the person to sit there and just keep laying bricks without improving it. And we see today we can print a house. What a fabulous thing. Now, all the builders broke now. No, they're in more demand than ever. But we still have machines out there that are actually physically printing a house. I think that's going to continue to evolve and it's going to be more opportunities. [40:22] Dave Pengelley: I think that that that comes back to the broader education and critical thinking, but also growth mindset and that we that we need to the people in a society to really take on. Um there's a lot of that's just the way it is. And you know, sometimes that is the way it is, and sometimes there's good reasons. And you don't just sort of burn the house down for the sake of it because you think houses are dumb now. But I was talking to a a young kid uh the other other day and he's just graduated after doing, you know, software engineering, computer science, thought he was going to be a programmer. And he graduated, you know, December just as clawed code and agentic coding and everything kind of sort of Really started ramping up and um found a hard to find a junior role. Now I don't know all his backstory, but he he's had trouble finding jobs as a junior dev. And he's ended up, you know, as a receptionist in a medical practice somewhere, like just some random office job just to pay the bills. Um, and he said, I'm they're not really interested in in building any apps or anything. So, but I I quizzed him further and I was like, You're doing this this day job. Are there tasks in your job that are repetitive? Build sort of Solutions, building that that was what I built, and I told him the story. That's what I built my early career on building VBA in the back of Excel and building access databases and stuff because not because the business said, hey, you should build a database, but because I'm like, I hate doing this copy and paste from this app into that app. I'm gonna build a thing that does it for me. I'm gonna write a script, I'm gonna do this. Um and just look at where you're at and what you're doing and build solutions. Don't ask for permission to build the solution. And if they go, what do you do? What is that? You say, Oh, I build this thing to make my job easier. And faster, so now I can do some other things. Like, what else is there to do? Um, but people, even someone who's just studied software engineering, comes out into the workforce and isn't thinking in that way of just build solutions to improve work, productivity. [42:13] Matt: Must [42:14] Richard Webbe: That's [42:14] Matt: have [42:14] Richard Webbe: that's exactly why I get upset about the narrative at the moment. It's shutting everyone down when we should be giving people empowerment, which is what [42:22] Matt: a [42:22] Richard Webbe: you were doing for that young man. Most [42:24] Matt: lot of [42:24] Richard Webbe: people get a degree, don't end up working in this discipline [42:28] Matt: sense. [42:28] Richard Webbe: that they did their degree in. [42:30] Dave Pengelley: No, I [42:30] Richard Webbe: What [42:30] Dave Pengelley: didn't [42:30] Richard Webbe: the degree is taught them how to think, solve problems. And if I'm a programmer and I'm coming out of college now, I just got better questions for my AI. Sorry, Matt, I'll cut you off. [42:40] Matt: No, that's okay. I I [42:41] Dave Pengelley: know. [42:42] Matt: um I don't remember what I was saying. I I think I was saying that most devs don't think that way. They they're very um I feel sorry for you know the quote unquote traditional dev, you know, the the person who went and learned some compute some computer science. Picked a particular path, you know, they learnt some language and they got good at some niche area. Um, you know, the the ones that are feeling that existential crisis the most right now, I really feel sorry for them, not because of the crisis, but because they haven't learned how to think, you know, wider, [43:18] Richard Webbe: Yeah, [43:19] Matt: you know, like you're just like you're describing there. Like what is [43:21] Richard Webbe: so [43:21] Matt: value? You really, really need to try and think about what is value and why are these people undervaluing you now? Like that it's not. [43:29] Richard Webbe: I'm not sure. [43:30] Matt: Anything to do with your ability. It's just your ability and your capacity to think about this stuff. [43:34] Dave Pengelley: Yeah, I mean, [43:35] Richard Webbe: Yeah, [43:35] Dave Pengelley: I [43:36] Richard Webbe: so [43:36] Dave Pengelley: I've [43:36] Richard Webbe: I got [43:37] Dave Pengelley: always said I went into computer science but never wanted to be a programmer. And so then when I got out in the job market, I'm like, what do I do? I didn't want to go for software engineering roles. I didn't didn't want to be a programmer. I just kind of defaulted into computer science because I had a tech background as [43:48] Richard Webbe: a [43:49] Dave Pengelley: a kid. Um but then when I got into the workforce, I went, Oh, hang on. I'm I have skills, I can solve [43:56] Richard Webbe: lot [43:56] Dave Pengelley: these [43:56] Richard Webbe: of people. [43:56] Dave Pengelley: problems, I can do things that will make my life easier in the Role. [44:00] Richard Webbe: The [44:00] Dave Pengelley: Um, [44:00] Richard Webbe: best example [44:01] Dave Pengelley: and [44:01] Richard Webbe: of transferring, sorry, that experience is the pilot, I believe, if I'm correct, that landed the stricken A380 that had the gas chamber go off and almost removed all the uh pneumatics. And he was flying one of the Australian quantits, one of the first A380s. And [44:19] Dave Pengelley: I'm [44:19] Richard Webbe: uh [44:19] Dave Pengelley: not sure. [44:19] Richard Webbe: he talks about, you know, they've got a little tiny joystick and a couple of buttons to fly the biggest flying machine in the world. And when everything went wrong, he had Test pilots with him, he had instructors with him, he had certifiers with him, and they all said at the outset, I believe, correct me if I'm wrong, that they would have failed him after he landed this craft, that by the way, someone says should never have been able to land. They would have failed him because he didn't follow procedure. But what he did do, because he was a 60-year, 40-year veteran or whatever, he knew how to fly a plane without hitting a button. He knew what the underlying value and physics and airflow was all about. about. So when things didn't perform, he just tried other things that he knew and he saved what 400 lives [45:06] Dave Pengelley: Yeah. [45:06] Richard Webbe: and a you know billion dollar aircraft and that's exactly what you're talking about. Dave, you come with skills. You may not apply them how you think you would, but they're there. [45:16] Dave Pengelley: Yeah, and so so this this this guy, I looked at his LinkedIn and he's got a featured post which is his GitHub repo, and I clicked in and he's had one commit all year. And so I said to him, Dude, take [45:28] Richard Webbe: You [45:28] Dave Pengelley: it [45:28] Richard Webbe: gotta [45:28] Dave Pengelley: or leave [45:28] Richard Webbe: do [45:29] Dave Pengelley: it, [45:29] Richard Webbe: more [45:29] Dave Pengelley: but [45:29] Richard Webbe: than that. [45:29] Dave Pengelley: but uh I've got like two thousand commits on mine and I'm not looking for software engineering roles. Um you should be doing projects, you should be doing something to show you're active in the space and current with the the tool set. Um [45:42] Matt: That's [45:42] Dave Pengelley: it's [45:42] Matt: what happened [45:43] Dave Pengelley: it's [45:43] Matt: to me, like when I was trying to look for uh help, you [45:48] Dave Pengelley: yeah. [45:48] Matt: know, in in in my business, in my professional contracts and and and stuff, my my uh my cognitive load was just getting to that point where I'm like, I think I need some help. And I started to look around And everyone that I spoke to, or one of the most promising candidates that I had chosen, they weren't doing that either. They showed me a couple of little bits and pieces here and there. And it became this weird philosophical thing for me. And I'm sure there's business owners out there that kind of feel this way as well in this time. Um, where you go, is it in my court or is it in theirs? Like who's hitting the ball here? And am I supposed to keep prodding you for evidence of life, or are You're supposed to be coming to me and kind [46:32] Dave Pengelley: Yeah. [46:32] Matt: of showing off and like showing, you know, that you can do this stuff. Um, yeah, because [46:37] Dave Pengelley: And [46:37] Matt: I I honestly just shut the door. I gave [46:39] Dave Pengelley: and [46:39] Matt: up on [46:40] Dave Pengelley: yeah, [46:40] Matt: it. [46:40] Dave Pengelley: in the in this era of coding agents, I I feel if you're going for software engineering roles, especially as a junior dev, you've got to show that you're using the current tool set, that you're not afraid of it, that you can be highly productive and come up with those personal projects and whatever that is, just build stuff. I mean, that's you know, we we talk about him sometimes. Peter Steinberg. Built 20 things before he released OpenClaw. Um, he was just building little things to help people in his life, to help himself, to solve micro problems. I showed my little uh MC cards app to help me like do little talks the other day. Like just all those little things, just build stuff and show that you are not giving up. [47:19] Matt: Yeah. And and [47:21] Dave Pengelley: Um [47:21] Matt: like anyone can set up a a a scheduled routine in in any of their AI tools to just you know send a prompt that says go build a thing. Um like there there's that still that differentiator um between building a thing and building the thing, or you know, building a thing that solves a problem or provides a value or you're receptive to somebody's pain and then this is the painkiller. You know, that that kind of thing is what I was looking for and expecting. [47:52] Dave Pengelley: Yeah. [47:53] Matt: I I if they had have done nothing, like literally nothing, and all they did was ask me endless questions. I would have been more interested. [48:01] Dave Pengelley: Yeah. [48:01] Richard Webbe: Yeah, yeah. It's all it's I think what we're talking about is not revolution. It's not a fall off the cliff. It's evolution. And it's been on since man first learned to stand. It is evolution. And I I just again, I'm not happy with the media and the hyperbole in the marketplace. And it's probably fueled by our first technology of social media that's painting doom and gloom. Nothing was solved. Doing that. And you just get what you got and away you go. You know, you you look at the person who who used to, I don't know, they used to deliver, you know, for the local supermarket on his pushbike, and then the guy next door buys a car and he's never seen a car before, and he goes, Oh, my life is ruined. No, it's not. Get a job with the guy with the car, save up and buy a car, or say that you can deliver better on a bike because you can get down smaller streets. It's [48:54] Dave Pengelley: Yeah. [48:55] Richard Webbe: going to be a slow evolution. [48:56] Dave Pengelley: You capital's pig. How dare you? [48:59] Richard Webbe: Well, here's the That's the thing. When Elon talks about money disappearing, everyone calls him a capitalist pig. But he is he's designing, not a communist, but he's certainly designing what uh Gene Rodenbury used to write about uh in [49:14] Dave Pengelley: Yeah. [49:14] Richard Webbe: Star Trek, and that is a world where money is not important, but value is. So we go straight to your actions relate to value rather than your actions relate to money, which translates to value. [49:30] Dave Pengelley: Does that devalue you as a human and are we getting away from the uh [49:33] Richard Webbe: Let's [49:33] Dave Pengelley: the [49:33] Richard Webbe: end up to the fifth dimension, you know [49:35] Dave Pengelley: Yeah. [49:35] Richard Webbe: that. [49:36] Dave Pengelley: Um, [49:37] Matt: It [49:37] Richard Webbe: Sorry, [49:37] Matt: is true. [49:38] Richard Webbe: that's Superman. I got it all wrong. [49:41] Matt: It is weird though, right? Like I I don't I I don't know where workforce appraisals are gonna go, you know, in this evolution, Richard. Like I don't know what they're gonna be marking. [49:53] Dave Pengelley: I a good idea. that's think [49:53] Matt: Like how do you mark it? Is it somehow, you know, uh a metric to do with the evolution? of the company [50:00] Dave Pengelley: Yeah. [50:00] Matt: and the culture and the the energy and the thinking styles and and how things are actually getting done rather than KBIs. You know, [50:09] Richard Webbe: Well, [50:10] Matt: like [50:10] Richard Webbe: yeah, no, that's true. That's true. And we so I think the problem is the solution. And I remember someone talking foolishly on the radio on the television when I was young, when they're promoting these really fast V8 engines in Australia, some of the most powerful and fastest cars in the world. And I remember some I won't. Say that word. Some silly person got up and said, This is so dangerous, crime is going to go up because no one will be able to catch them. Because that person wasn't thinking that the police might have these cars. And so that's not going to be a problem. So, what I'm saying is the more complex AI becomes, the more it'll be able to measure what is truly real, and more it'll be able to measure the value of an individual. And we shouldn't feel like that's putting pressure on us. We should feel more embodied. And I mean, I'm still trying to hope that the world's a meritocracy, but it'll become more of a meritocracy if we're measuring what we're doing better, not the popular kid or the good-looking kid or you know, the kid that was born into money. It'll be the person that adds the most value. [51:17] Dave Pengelley: It's it's funny that that comparison talk about the cops and the V8 and all the crims will have an advantage. As a sound of my kids the other day going, like they they should teach martial arts in in school. Like that should be part of and it's like that's a terrible. Then all the bullies will know how to beat people up. And I'm like, but they won't, because I'll know that everyone else knows how to defend themselves as well. He's like, yeah, but then they'll escalate and they'll just get knives because knives are harder to defend against. I was like, bad guys are already going to use knives, they're going to use knives. Like, I don't think that changes anything. But you know, [51:46] Richard Webbe: No, [51:46] Dave Pengelley: like it's just that equalizer and thinking, you know, like, okay, if one side's got something, the other side's probably going to have that as well. [51:54] Richard Webbe: no. [51:55] Dave Pengelley: Um, and how do we lift everyone, the rising tide lifting all ships? Um [52:00] Richard Webbe: My battery's running low and I haven't plugged it in, but I will die right at one o'clock. [52:04] Dave Pengelley: All right. Don't don't die. Um, I mentioned Peter Steinberg. I thought this was funny on X this week. Uh, and I know sorry, sorry, sorry, regulars, we haven't thrown any bumpers up for news or wins of the week. We're very just like discussion show this week. But uh Open Claw, Peter Steinberger claimed he spent 10 million on AI tokens in a single month this September. That'd be enough to pay for an office with hundreds of workers for an entire month. I don't love noose research, the guys that make Hermes doing what? [52:31] Richard Webbe: That is [52:31] Dave Pengelley: Which [52:31] Richard Webbe: very [52:32] Dave Pengelley: pe [52:32] Richard Webbe: funny. [52:32] Dave Pengelley: which which Peter didn't take very well. Pretty [52:35] Richard Webbe: No. [52:35] Dave Pengelley: disappointed to see this from the official news research account. You've trained models yourselves, you know how much compute research takes. These personal things are getting old. [52:44] Richard Webbe: That is ironically hilarious. [52:46] Dave Pengelley: 10 million dollars on AR tokens in a single month. I think that he's saying about training, but it's anyway. That's that's crazy money for [52:53] Richard Webbe: That [52:53] Dave Pengelley: what open [52:53] Richard Webbe: is [52:53] Dave Pengelley: core [52:53] Richard Webbe: crazy, [52:54] Dave Pengelley: is actually [52:54] Richard Webbe: man. [52:54] Dave Pengelley: delivering. [52:55] Richard Webbe: That is [52:55] Matt: Do you [52:55] Richard Webbe: that is. [52:56] Matt: I I'm curious if you guys have come across like looking at I don't know who runs the noose account, if that's tech or what, but um like that that comp that kind of PR, you know, that kind of uh comes across as rudeness. You know, have you guys come across that as well after working with agents for so long? Because I know personally, like, you know, anyone who's listening or watching this that has spoken to me in the last sort of month or so probably knows that my uh my ability to communicate somewhat softly. has probably diminished. [53:32] Richard Webbe: I haven't noticed any. [53:34] Matt: You haven't? [53:35] Richard Webbe: No. [53:36] Dave Pengelley: I [53:36] Matt: I [53:36] Dave Pengelley: know [53:36] Matt: genuinely. [53:37] Dave Pengelley: I'm I I I I I noticed I'm just gonna go, okay, Matt's Matt's not being at aggressive or attacking, Matt's just being Matt. [53:45] Richard Webbe: This [53:46] Matt: Yeah, [53:46] Richard Webbe: is where [53:46] Matt: it [53:46] Richard Webbe: it [53:46] Matt: it's [53:46] Richard Webbe: is. [53:47] Matt: a shame. Like I, you know, people people worry about like um seeing [53:52] Dave Pengelley: Like [53:52] Matt: AI slop in their newsfeed. Whereas I think the bigger problem is us projecting our own Like um bastardized version of ourself because we've just lost our humanity [54:06] Dave Pengelley: Like [54:06] Matt: in the [54:06] Dave Pengelley: like [54:06] Matt: communication [54:07] Dave Pengelley: when when [54:07] Matt: process. [54:08] Dave Pengelley: when I send you that funny, that that funny John Stewart clip, and like I said, John Stewart take him or leave him, like broken clocks are right twice a day. But he was calling out Muse, and he had this little comedy sketch thing where he had this little fake CG Muse doll being nefarious, but being all cute about it, so that way you'd it's disarming that uh that meta would launch this sort of um cute little mascot to you know hide us nefarious. Potentially nefarious purposes. But like, yeah, Matt, you were like, I don't watch anything from mainstream. Why didn't they just say that? I'm like, well, they did say that, but then [54:37] Matt: Yeah. [54:37] Dave Pengelley: they just also augmented it with comedy. [54:38] Matt: It was at a very bad time. [54:40] Dave Pengelley: Yeah, [54:40] Matt: And [54:41] Dave Pengelley: no [54:41] Matt: like [54:41] Dave Pengelley: worries [54:41] Matt: I I [54:41] Dave Pengelley: there. [54:42] Matt: can't I can't stand sort of reporter style language stuff at the best of times. Like I know that was more of like a talk show, you know, late night show kind of [54:52] Dave Pengelley: Yeah, [54:52] Matt: vibe. [54:52] Dave Pengelley: yeah. [54:52] Matt: Um but I don't know, it just makes me cringe now. But yes, my my my terseness in that exchange was probably a little bit too [55:00] Dave Pengelley: Over man. [55:01] Richard Webbe: Matt, I think most people understand the and this is the important thing. Whether someone's swearing at you, screaming at you, hugging you, or whatever, they understand, because this is humans, right? And this is the human condition. We understand the passion and the meaning behind your words and actions. And so while someone could be animatedly angry, the people around them go, yeah, they're not really angry at me. They're just in a hurry or frustrated or they want to help us, right? And I think that's that nonverbal communication that can be misinterpreted in the emotional tone of our voices in there. I always think about. [55:39] Matt: He [55:40] Dave Pengelley: Oh, [55:40] Matt: died. [55:40] Dave Pengelley: his battery died. He always thinks about n [55:43] Matt: What [55:43] Dave Pengelley: charging [55:43] Matt: a [55:44] Dave Pengelley: his laptop more effectively. [55:45] Matt: what a perfect cut as well. That was amazing. [55:48] Dave Pengelley: Yes. Um what was I gonna wrap up on? [55:52] Matt: What [55:52] Dave Pengelley: Uh [55:52] Matt: about the IPO, the anthropic. [55:56] Dave Pengelley: well yeah, like there's people saying they're financially better off than they thought, but Um yeah, [56:02] Matt: No, [56:02] Dave Pengelley: I [56:02] Matt: the [56:02] Dave Pengelley: don't [56:02] Matt: actual [56:02] Dave Pengelley: know. [56:03] Matt: the the the public act like the actuals um I don't forget all the investment [56:06] Dave Pengelley: The S [56:07] Matt: go-to-market [56:07] Dave Pengelley: one IPR, [56:07] Matt: terms. [56:08] Dave Pengelley: yeah, yeah. [56:08] Matt: Um apparently like all of the anthropic employees are gonna liquidate into about six million dollars each. Um which kind of means that like you know there's gonna be a whole bunch of angel investing and you know just distributing that wealth around after this, which you know, what's that gonna do? Like it's fascinating how that keeps coming up. [56:30] Dave Pengelley: Cursor and SpaceX AI employees apparently uh making good money on things and stuff like that. Just go there's money out there. Money people, some people have money, some people don't, and and they'll they'll use it for whatever they use it for. And dude, I just yep, tourways, mate. Um [57:14] Matt: Hmm. Yeah, I know we spoke about that before with like the self-scan checkouts and stuff. Um like I I I quite often actually I had this, I actually had this exact situation where I I use Starlink, you know, for my internet. It's just the water it's It gives me the best result in where I live right now. And I had a bit of an issue. And it was like I very rarely ever go to support, or I very rarely ring a company to complain or ask for a refund or whatever the case may be. But I was curious about some support scenario. And instead of just trying to email their support at email, um, like they have Grok wired into their support system. And that enabled me to ask a whole bunch of these extra questions that I wouldn't have even bothered. And I got way better service. than what I would have if I had have tried to go through the official channels. So yeah, that was interesting. [58:06] Dave Pengelley: Yeah, it's it's funny how as new channels evolve, sometimes they can be overpowered compared to traditional channels. It depends on over time. I I remember back in the day, Qantas, when they first launched their Facebook support, and you could message like for frequent flyer support through to Facebook. You got a special dedicated social team and they would bend over backwards. They were the best. If I had any frequent flyer issues, like I just like once I was. Was like a thousand points short of my renewal to stay in silver or whatever. And they just bumped me and went, No, no worries, cool. Yeah, you can stay silver. And then, like two years later, as that all got conglomerated back into their primary support systems, and you're just getting automated or like knowledge base articles and stuff was broken. I was like, this sucks now. But I remember when they first adopted the new channel, that had a special team, had a special focus. And I think with some of this AI stuff, if they're employed, especially somewhat like SpaceX or Grok and stuff, they're leveraging these technologies at The best way they can with the latest features and the latest context information to drive a great customer service because they want that to be the primary thing and legacy systems take a while to catch up. [59:30] Dave Pengelley: You meant Matt, you mentioned like your tone and interaction with humans being affected by the way you talk to agents. I set up one of my agents that does like try to do trading activities for me and it's all you know make-believe money at the moment. But I've got it emailing me daily reports. And I think it's the Lunar model that's doing it. And it says, Oh, I found this thing and this didn't work and this one was broken. I'm like, so fix that thing and do this and set this thing to this. And then you know it reads that email and comes back to me because I'm just doing right. Email like old school, like it's another person. And he goes, Um, okay, so you want me to reset this thing. Can you confirm that you want me to reset this thing? I'm like, I gave you a direct instruction. No, I'm more caps. I gave you a direct instruction. That was the confirmation. Stop asking for confirmations. Switch to the Terra model. I want to try Terra. Um, and it's like, oh, sorry, I can't change, like I can't let you do that, Dave. Um, I can't I can't I can't change model via the email. Go to the terminal and do it straight into Hermes. I'm like, damn. I will. I will um you know, news Terra six point one dropped. So um we thought [01:00:35] Matt: Yeah. [01:00:35] Dave Pengelley: Terra was dead. Long live Terra. [01:00:37] Matt: Yeah, I I really liked Ben Davis's take on that. And he he uh he said it so well. He's like, this if you're reading this message, it this model isn't for you. Like, you know, this this model is for people who are on cheaper plans and they need a more efficient, you know, cheaper costing agent, or it's for people that are building agents and they're wiring it into agency systems for their clients. And it's just gonna become a more of a cost effective option where you're not just grabbing the rocket launcher when you need a pistol, you [01:01:07] Dave Pengelley: Yeah, [01:01:07] Matt: know. [01:01:08] Dave Pengelley: I'm I'm trying Terra just to see whether it speaks more human because Luna and Sol are a bit robotic. Astra has been disappointing me a little bit later. Again with that confirmation thing. Like I'm building, building or not good sort of diagnosis of the issue, and it's inferred that I wanted to fix it, but just says, yep, that's right, that's the problem. I'm like so. [01:01:30] Dave Pengelley: Fix it. Like [01:01:31] Matt: So [01:01:32] Dave Pengelley: that's [01:01:32] Matt: what's what's your take so far on that that how it feels like you know compared [01:01:36] Dave Pengelley: I [01:01:36] Matt: to [01:01:36] Dave Pengelley: I have [01:01:36] Matt: soul? [01:01:37] Dave Pengelley: haven't used it heavily enough yet to have strong feelings for it. Um still I mean it only dropped yesterday, right? So I I haven't spent a lot of time with it. Hamilton is still probably being backed by Astra. When Hamilton's okay, Hamilton's a different kettle of fish to Codex, um, mostly. Grockbots. That's my dot. Yeah. Um my my grogbots are still doing a lot of heavy lifting on lots of different things. So like I said, I'm I'm I'm using and abusing Hamilton as much as I can for game game development at the moment because that's a lot of tokens otherwise and they're free. [01:02:09] Matt: Okay, like [01:02:09] Dave Pengelley: So [01:02:10] Matt: was one of my um one of my client projects, we we did a bunch of evaluations and we found that the five-six Terror actually made a better result than what Five Six Soul did, and that was even comparing to um uh sonnet uh one of the Sonnet. Sonnets recently, [01:02:30] Dave Pengelley: Yeah. [01:02:30] Matt: it would have been probably Sonnet 5. Um, and yeah, Terra literally came out with a better language, you know, style that these particular this company needed for their work. Uh and their work that they were doing was actually client-facing for them. So my thing that I built for them was a pipeline where research was done, um, proprietary information was woven into it, and then that would get presented to their clients. And yeah, we found that 5.6 Terra did a meaningfully better job than. The other lineup, so it'd be interesting to see if Six One Terror does as [01:03:03] Dave Pengelley: Well, [01:03:03] Matt: well. [01:03:03] Dave Pengelley: yeah, I'm gonna I'm gonna give 6.1 Terra a shot at doing the comic strip this week uh for the show. Last week I think Sol did it and it wasn't great, and I got Groc47 to review and fix it. Um and didn't did a much better job on the storyboarding. Uh and so we'll see how 6.1 Terra does on extracting the actual moments in context because sometimes it pulls key moments and attaches a uh a mismatch quote to a point, and you go, hang on, no, what I said that like I was wrapping the show up, or I was wrapping that segment. That was a like a quip, not a key point for you to highlight as the the moral of the story, you idiot. Uh, whereas Grok Grok kind of got that and was like, Okay, cool, yeah, I see what you're doing there. So we'll we'll see, we'll see how 6.1 Terra does with the comics, um, which is a good segue. We're gonna wrap the show there. But uh, of course, AI operators, this is the show. This is the website, ai operatorspod.com, and now on the website. Website, not only do we have our full episode catalog with all our previous episodes, but the field notes now have a spot on the website, which is fun, as do the operator stories. So shipped. What's the words? No, you the the guy, the the he I haven't haven't pushed the uh the latest uh story from Monday up onto here yet. But uh our New York friend Greg Rosner, who was in the operator stories on Monday, he uh he didn't mind the occasional cast word. So um but Great chat. And Matt, you'd you'd love this one. I don't know if you've had a chance to watch or not. But um, really, his thing is he's backed into AI from a marketing narrative consultancy agency. And so all he's doing is he's applying those traditional storytelling mechanics and techniques and then getting AI to execute them in the businesses. So uh Greg was a great chat. Uh published that one up on Monday. But yeah, all the stuff there. Mondays is operator stories, Wednesdays is the live show, Fridays field notes. Um, I've got two more operator stories. stories that I'm filming later this week. So we're gonna actually have a bit of a backlog and start building up a bit of a a hopper um for the stories, which is great. [01:05:07] Matt: This [01:05:07] Dave Pengelley: Just [01:05:08] Matt: is [01:05:08] Dave Pengelley: lots [01:05:08] Matt: awesome. [01:05:08] Dave Pengelley: of interest. [01:05:08] Matt: I I I I wouldn't I want you to start like we we need we either need sponsors or some sort of revenue stream to just be able to pump into you to just keep you doing this because it's where you're strongest, I think. You know. [01:05:31] Dave Pengelley: You're all very kind. Um, you know, if it's the will of things, it will happen. But for that, we need it, all of you lovely viewers, like, subscribe, hit the bell icon. Let's get some traction on there so we can get some people uh more eyeballs and get some funding to continue to grow and invest in this show. Cause I I love it. [01:05:50] Dave Pengelley: No, what are you talking about? It's it's I'm uh that's [01:05:54] Matt: Just [01:05:54] Dave Pengelley: like golden [01:05:55] Matt: cut aside, [01:05:55] Dave Pengelley: locks. [01:05:56] Matt: sleep the rest. [01:05:57] Dave Pengelley: Um, anyway. Had the mention men's grooming. We were trying thought we nearly got a show without it, but uh [01:06:04] Matt: I've got a nice closing statement for everyone. [01:06:07] Dave Pengelley: Yes, [01:06:08] Matt: It's [01:06:08] Dave Pengelley: please. [01:06:08] Matt: kind of pithy and it kind of probably sounds like slop, but um you know we've made it pretty clear that you know the idea of AI here makes starting a lot cheaper. You know, the [01:06:20] Dave Pengelley: Mm-hmm. [01:06:20] Matt: actual starting of something [01:06:22] Dave Pengelley: Yeah, [01:06:22] Matt: is basically [01:06:22] Dave Pengelley: it it [01:06:23] Matt: free. [01:06:23] Dave Pengelley: breaks the inertia, clears the blank page. [01:06:25] Matt: Yeah. Your job now is picking what needs to get finished. [01:06:33] Dave Pengelley: Yeah. [01:06:36] Dave Pengelley: Yeah. [01:06:36] Matt: There you [01:06:37] Dave Pengelley: Just because [01:06:37] Matt: go. [01:06:37] Dave Pengelley: it's everyone can do more things now doesn't mean that everyone's going to be doing the best things and there is room for your creativity to differentiate. [01:06:43] Matt: Yep. [01:06:44] Dave Pengelley: Um the bar's higher, but so is the floor. You're stepping up well and starting from a higher point. [01:06:49] Matt: And that pressure and expectation, like you know, the the the dip, you know, the lulls you'll feel that lull more because you're doing more at once. So you're just gonna keep pushing through it. [01:07:00] Dave Pengelley: Brilliant. [01:07:02] Matt: In the [01:07:03] Dave Pengelley: Yeah. [01:07:03] Matt: long run. [01:07:04] Dave Pengelley: What one one thing that Elon said that he wants his models to do is seek the truth and to stay curious. And I think [01:07:14] Dave Pengelley: that's that's what we need to be doing as humans. Yep. Brilliant. All right, gentlemen. Thank you very much. See you next week. [01:07:25] Dave Pengelley: Thanks all for watching us.