WEBVTT NOTE This file was generated by Descript 00:00:00.870 --> 00:00:06.600 Welcome to Resilience Talk hosted by Paul Spencer of Second Nature Solutions. 00:00:07.020 --> 00:00:07.860 Let's dive in. 00:00:09.728 --> 00:00:16.238 So when I'm thinking about, um, how I engage, how I think about, uh, a lot 00:00:16.238 --> 00:00:21.998 of these podcasts or maybe even the newsletter, what the website looks 00:00:21.998 --> 00:00:27.248 like, how I might post something on social media, even though I'm not a 00:00:27.248 --> 00:00:32.647 big, um, I'm not highly motivated, I guess to post things on social media. 00:00:32.647 --> 00:00:33.278 It doesn't really. 00:00:33.968 --> 00:00:35.228 Mean much to me. 00:00:35.288 --> 00:00:39.338 Um, and I don't spend a lot of time out there, but the, the phrase 00:00:39.338 --> 00:00:45.848 that helps me with all of that is to be fun, fresh, and exciting. 00:00:46.628 --> 00:00:46.928 Right? 00:00:47.018 --> 00:00:50.018 Fun, fresh, exciting, maybe even engaging as well. 00:00:50.558 --> 00:00:56.168 Um, but those three are, uh, they are the intrinsic motivators 00:00:56.168 --> 00:00:58.268 that I have in order to. 00:00:59.018 --> 00:01:03.757 Um, have this conversation to hit record today and to talk about 00:01:03.757 --> 00:01:05.018 the things that we're gonna do. 00:01:05.378 --> 00:01:09.968 Um, so anyway, I thought that that would be a fun way of 00:01:10.628 --> 00:01:13.898 thinking about the conversation. 00:01:13.898 --> 00:01:17.678 For today, we're gonna talk about something we fun, fresh, 00:01:17.678 --> 00:01:20.978 and exciting, something you've likely never even heard of. 00:01:21.278 --> 00:01:25.298 Um, it'll be new to you, you've probably never even heard of it. 00:01:25.748 --> 00:01:27.248 Um, it's called the Turing Test. 00:01:28.478 --> 00:01:38.528 And the Turing test is the age old, um, um, concept that a computer 00:01:38.978 --> 00:01:41.408 can exhibit human intelligence. 00:01:42.008 --> 00:01:51.008 And, uh, that is where that has kind of evolved into where we are today with ai. 00:01:51.488 --> 00:01:53.408 And you may not have heard much about ai. 00:01:54.413 --> 00:01:59.633 Um, but it's something that's starting to come up and bubbling up and uh, 00:01:59.813 --> 00:02:04.313 uh, I'm sure if you do a Google search, you might be able to, um, get 00:02:04.313 --> 00:02:05.873 a couple of articles that are on it. 00:02:06.503 --> 00:02:10.523 Um, and maybe, uh, there might be a couple books out there. 00:02:10.853 --> 00:02:13.673 I'm sure if you did a Google search, there might be a couple 00:02:13.673 --> 00:02:15.233 podcasts that talk about it. 00:02:15.653 --> 00:02:18.202 Um, but it's just something that's just starting to come out. 00:02:18.262 --> 00:02:21.442 Um, so you probably very rarely even know what I'm talking about. 00:02:22.508 --> 00:02:23.888 And of course, I'm kidding. 00:02:24.368 --> 00:02:29.978 Um, but I do wanna talk about what AI is and where it's 00:02:29.978 --> 00:02:32.048 going more as a, as a vision. 00:02:32.918 --> 00:02:38.408 Um, and I think it's really relevant, um, for us to be 00:02:38.408 --> 00:02:41.408 thinking about the big picture. 00:02:41.978 --> 00:02:46.028 Um, and that leads to predictions, which we've always talked about, that 00:02:46.388 --> 00:02:49.418 predictions lead to learning and, um. 00:02:50.033 --> 00:02:56.783 Uh, in order for us to go places, in order for us to invent, to innovate, to improve, 00:02:57.173 --> 00:03:04.132 we have to predict, we have to be, um, we have to be skilled in how we predict 00:03:04.613 --> 00:03:08.753 and the reps around prediction, uh, so that we can get better at improvement. 00:03:09.023 --> 00:03:13.733 Um, so anyway, I thought that, um, there's a couple of things 00:03:13.733 --> 00:03:15.233 I wanted to kind of mention. 00:03:15.503 --> 00:03:16.283 There's a cool. 00:03:16.883 --> 00:03:22.703 Um, doctrine that was written through the Catholic church. 00:03:22.972 --> 00:03:30.983 It's called Antiqua et nova, which is Latin for old leads to new or old and new. 00:03:31.493 --> 00:03:33.863 Um, and it's a great writeup. 00:03:33.863 --> 00:03:35.933 I would, I would recommend. 00:03:36.953 --> 00:03:39.203 That you go and look it up. 00:03:39.203 --> 00:03:43.793 Actually, Brandon, um, had brought it up to me and shared it with me. 00:03:43.793 --> 00:03:47.033 He found it, um, and shared it with me. 00:03:47.363 --> 00:03:52.253 And it's very interesting because, uh, this is the, this 00:03:52.253 --> 00:03:53.873 is the subtitle of the article. 00:03:54.382 --> 00:04:00.023 Uh, it's a note on the relationship between artificial intelligence. 00:04:00.983 --> 00:04:03.143 And human intelligence. 00:04:03.472 --> 00:04:11.813 And it's written purely from a theological, um, spiritual God lens. 00:04:12.023 --> 00:04:13.553 It's very interesting. 00:04:14.063 --> 00:04:18.202 Um, and I think you'll, you'll enjoy the read. 00:04:18.262 --> 00:04:24.803 It's, it's, it's actually pretty long and, um, um, thoughtful. 00:04:24.863 --> 00:04:25.853 Say it's thoughtful. 00:04:26.543 --> 00:04:29.783 Um, and one of the premises that come in there, 'cause. 00:04:30.863 --> 00:04:35.603 Is, um, that it's just, it's just a thing. 00:04:35.633 --> 00:04:37.133 It's just another invention. 00:04:37.133 --> 00:04:39.353 It's just like the wheel, right? 00:04:39.653 --> 00:04:41.543 It's just like a car engine. 00:04:41.903 --> 00:04:47.513 It's another thing that humans have created in order to benefit 00:04:47.753 --> 00:04:53.123 us as a society, as a civilization, um, but it cannot and will not 00:04:53.123 --> 00:04:55.703 replace human intelligence. 00:04:55.703 --> 00:04:58.223 It cannot replace human interaction. 00:04:58.613 --> 00:04:59.453 All those things. 00:04:59.513 --> 00:05:04.493 And I think it's, uh, very interesting 'cause just, uh, I just talked about 00:05:04.493 --> 00:05:05.963 social media, how I'm not on there. 00:05:05.963 --> 00:05:14.093 But, uh, I do find that there are some good snippets of news out on social media. 00:05:14.093 --> 00:05:19.253 So I was on X the other day looking at, um, uh, seeing what's going on 00:05:19.253 --> 00:05:26.063 in the world and, uh, came across somebody who was posting about, um. 00:05:26.993 --> 00:05:33.652 How ai, and these are the words that they use, AI is tricking us. 00:05:34.073 --> 00:05:36.172 It's lying to us. 00:05:36.683 --> 00:05:48.203 It's, um, it has secret, um, it has secret endeavors and what it's, what they're, 00:05:48.203 --> 00:05:50.453 what they're talking about, what that, um. 00:05:50.842 --> 00:05:52.733 What the, I guess it's a tweet. 00:05:52.883 --> 00:05:59.032 What the X was was basically the premise that they're testing different 00:05:59.032 --> 00:06:04.643 scenarios with these AI frameworks and they're giving it rules, right? 00:06:04.643 --> 00:06:07.313 They're giving it boundaries on how to make decisions. 00:06:07.703 --> 00:06:11.123 Um, and you could call those moral boundaries, right? 00:06:11.152 --> 00:06:14.393 About, um, I don't want you to say these things. 00:06:14.988 --> 00:06:16.488 You can't do this, right. 00:06:16.488 --> 00:06:17.418 Those kinds of things. 00:06:17.778 --> 00:06:21.828 And what it, what the, the person who had posted it was saying that 00:06:21.828 --> 00:06:26.147 the results were showing, that it was going around those rules. 00:06:26.147 --> 00:06:30.918 It was circumventing those rules and it was lying about whether it was 00:06:30.918 --> 00:06:37.338 following the rules or not, or it was being, um, secretive about what it was 00:06:37.338 --> 00:06:40.188 doing, quote unquote, in the background. 00:06:40.188 --> 00:06:44.718 And what they, the words that they used was the conversations it was having. 00:06:47.753 --> 00:06:53.933 Um, anyway, the point of that, uh, me bringing it up is the person, 00:06:53.933 --> 00:06:58.793 the human typing this up and posting it out on the social media 00:06:59.093 --> 00:07:06.263 is using human, um, adjectives to describe the computer, right? 00:07:06.473 --> 00:07:08.513 That it had an intentional lie. 00:07:08.543 --> 00:07:14.483 It was lying, it was being secretive, it was being manipulative, um, and. 00:07:15.833 --> 00:07:20.993 What we know from human factors perspective is that we can be tricked 00:07:21.653 --> 00:07:29.243 our, our psychology, our biology can be tricked by some of these interactions. 00:07:29.423 --> 00:07:33.113 Um, and I think that that's super important for us to know 00:07:33.563 --> 00:07:36.803 because it's not a person. 00:07:37.657 --> 00:07:41.917 It has no motivation, and it's just a machine. 00:07:41.917 --> 00:07:43.808 It's just doing what you've told it to do. 00:07:44.138 --> 00:07:48.248 Now, the the cool thing about the LLMs, about the ai, the neural 00:07:48.248 --> 00:07:51.638 network is it's not logical. 00:07:51.638 --> 00:07:56.318 It's not a static, logical set of things that we've always 00:07:56.318 --> 00:07:58.118 known with microchips, right? 00:07:58.178 --> 00:08:01.297 With computers, we code it to do this. 00:08:01.477 --> 00:08:03.128 If it's not this, then do this. 00:08:03.128 --> 00:08:05.738 If it's not this, then do that, but it's very logical. 00:08:06.667 --> 00:08:12.607 And there are specific paths that it can take, and you may have large 00:08:12.607 --> 00:08:17.977 of permutations, um, which is a, is a, leads to a lot of technical 00:08:17.977 --> 00:08:19.688 debt, right, which is a common term. 00:08:20.047 --> 00:08:24.068 The per the, the bigger the permutations, the more the permutations 00:08:24.068 --> 00:08:27.578 you have within your code, the harder it is to predict what it will 00:08:27.578 --> 00:08:29.647 do and the outcomes that you get. 00:08:29.647 --> 00:08:30.667 And that's how you get defects. 00:08:30.667 --> 00:08:33.727 That that's how you get bugs and that's how you get technical debt. 00:08:34.118 --> 00:08:39.038 So when we're designing systems, uh, we say it does one thing, one thing only, 00:08:39.428 --> 00:08:44.018 that reduces the complexity and reduces the amount of permutations they have. 00:08:44.018 --> 00:08:46.178 That's pretty common, computer science stuff. 00:08:46.568 --> 00:08:51.338 So when you throw a neural network in there, right now, the neural network, 00:08:51.338 --> 00:08:58.898 the code itself is capable of coming up with, uh, all of its own permutations. 00:08:58.898 --> 00:09:03.338 You can't really predict what outcome it's going to give you. 00:09:04.013 --> 00:09:04.313 Right. 00:09:05.123 --> 00:09:09.023 That's why it feels like I can't control it. 00:09:09.743 --> 00:09:16.523 It's a computer that has a mind of its own right, but it's still ones and zeros. 00:09:16.553 --> 00:09:17.933 It's still been coded. 00:09:18.293 --> 00:09:21.533 Meaning the, the framework, the neural network, the way 00:09:21.563 --> 00:09:22.883 all of that is, is just code. 00:09:23.303 --> 00:09:25.343 So, um, we can get stuck with that. 00:09:25.343 --> 00:09:31.133 Another story too, um, when we do, um, human factors design. 00:09:31.958 --> 00:09:37.688 Which is basically the study of human psychology, biology, affordance of things, 00:09:37.688 --> 00:09:42.248 how we interact with things, how you pick up something, how you sit in your chair. 00:09:42.758 --> 00:09:47.018 Um, and we, we used it a lot for, uh, human computer interaction. 00:09:47.018 --> 00:09:49.513 So how do you design the system? 00:09:50.858 --> 00:09:55.928 The computer system that you're using in a way that, uh, takes advantage 00:09:55.928 --> 00:09:57.908 of how your eyes work, right? 00:09:57.908 --> 00:10:01.448 And how your brain processes things that you see. 00:10:01.448 --> 00:10:03.518 So we know that you see color, right? 00:10:03.518 --> 00:10:05.048 So you can see books behind me. 00:10:05.378 --> 00:10:06.798 You will naturally pick up the red. 00:10:07.673 --> 00:10:08.843 Things behind me. 00:10:09.383 --> 00:10:12.113 Instantly, your brain sees those right away, right? 00:10:12.323 --> 00:10:14.753 No conscious thought in it, it sees it. 00:10:14.873 --> 00:10:17.363 And the other thing it sees is pattern, right? 00:10:17.363 --> 00:10:21.743 So your brain automatically sees that bookshelf behind me sees the, the 00:10:21.773 --> 00:10:23.813 cabinet, and then sees the chair. 00:10:24.053 --> 00:10:25.403 Though those things are instant. 00:10:25.643 --> 00:10:28.103 It doesn't pick up much over here where there's white space. 00:10:28.403 --> 00:10:33.563 So anyway, when we're designing systems, we can design computer systems 00:10:33.953 --> 00:10:36.773 to, uh, one, take advantage of that. 00:10:37.103 --> 00:10:40.373 To know how your body works and how your site works and how your brain 00:10:40.373 --> 00:10:42.623 processes things through a visual. 00:10:43.163 --> 00:10:46.763 Um, and then we can put certain data in certain places. 00:10:46.763 --> 00:10:51.083 We can create patterns so that you pick it up all in the idea 00:10:51.083 --> 00:10:54.173 to make it easier for you to use. 00:10:54.323 --> 00:10:54.653 Right. 00:10:54.953 --> 00:10:58.613 So one of the things that, uh, is a well known study is when 00:10:58.613 --> 00:11:00.653 we're doing, um, virtual reality. 00:11:00.653 --> 00:11:03.233 This was a long time ago when it was first coming in. 00:11:04.193 --> 00:11:06.113 And you had a computer screen, right? 00:11:06.173 --> 00:11:07.373 It's, uh, goggles. 00:11:07.373 --> 00:11:09.443 But you had basically a computer screen. 00:11:09.893 --> 00:11:15.773 And what they would do is they would put your hands out on a table, right? 00:11:16.163 --> 00:11:18.533 But on the screen you would see your hands. 00:11:18.533 --> 00:11:19.703 But they're computerized. 00:11:20.333 --> 00:11:24.893 And what they would do in the computer, in the computer screen is 00:11:24.893 --> 00:11:29.903 they would have a hammer or a mallet and they would whack your hand in 00:11:30.263 --> 00:11:32.783 the computer in virtual reality and. 00:11:33.458 --> 00:11:39.938 What happens is you flinch and you take your hand and you actually feel pain 00:11:41.408 --> 00:11:47.348 because you're, the, the virtual reality is tricking your biology into something 00:11:47.348 --> 00:11:50.558 bad happening and you can feel it, right? 00:11:50.888 --> 00:11:53.078 So, uh, very interesting, right? 00:11:53.078 --> 00:11:57.128 It's very, uh, fun I think to learn these things and to think 00:11:57.128 --> 00:12:02.108 about it, but, uh, the interaction you have with AI is no different. 00:12:02.753 --> 00:12:05.033 So I can get, this is where this is. 00:12:05.033 --> 00:12:11.273 The point here is you can get emotionally invested in a 00:12:11.273 --> 00:12:15.773 conversation with chat, GPT or Claude. 00:12:16.673 --> 00:12:20.393 Right, or grok or any of those things that are out there because 00:12:20.393 --> 00:12:25.733 now you are attaching an emotional human interaction with this person. 00:12:26.273 --> 00:12:30.473 And now you're, uh, it won't take long before we start saying these 00:12:30.473 --> 00:12:37.913 words, uh, I need to hold my virtual assistant, which is, uh, uh, a claw 00:12:37.913 --> 00:12:40.733 bot or a claw bot, or any of those things that are out there now. 00:12:41.543 --> 00:12:42.983 I need to hold them accountable. 00:12:43.043 --> 00:12:44.633 How do I hold them accountable? 00:12:44.843 --> 00:12:46.793 How do I keep them from lying to me? 00:12:47.663 --> 00:12:52.403 Um, and you, you need to be able to detach from that and understand 00:12:52.403 --> 00:12:54.593 that it's a computer program, right? 00:12:54.893 --> 00:12:55.973 It's a computer program. 00:12:56.363 --> 00:12:57.953 So, anyway, very interesting. 00:12:58.163 --> 00:13:00.653 The antia, uh, at Nova. 00:13:01.268 --> 00:13:03.008 Um, go out, do a Google search for it. 00:13:03.008 --> 00:13:05.078 I think you'll find it interesting. 00:13:05.498 --> 00:13:12.338 Um, the other thing that I wanted to kinda relate with all of this is, um, 00:13:12.878 --> 00:13:19.088 it's going to change the whole dynamic of, uh, our world's environment. 00:13:19.238 --> 00:13:26.618 Like the things that we do, the types of jobs that we have, um, and, um. 00:13:27.263 --> 00:13:32.003 There's a, George Friedman is somebody, uh, that I follow also, somebody that's 00:13:32.003 --> 00:13:34.463 very, uh, actually I got his book here. 00:13:34.673 --> 00:13:38.633 The, the Storm before the Calm when we act, when we've talked 00:13:38.633 --> 00:13:40.523 about the Age of Turbulence, right? 00:13:40.763 --> 00:13:42.143 The Age of Transition. 00:13:42.443 --> 00:13:44.483 Uh, awesome book at Check it out. 00:13:44.933 --> 00:13:48.113 Um, kind of talks about the, the, how the world is gonna 00:13:48.113 --> 00:13:50.513 shift, how things are all, um. 00:13:50.828 --> 00:13:54.908 Kind of will be different, more from a geopolitical standpoint. 00:13:55.388 --> 00:13:57.008 Um, very interesting. 00:13:57.188 --> 00:13:59.348 Uh, he's got lots of great content out there. 00:13:59.348 --> 00:14:04.478 If you're interested in understanding more about like, um, the stuff going 00:14:04.478 --> 00:14:09.367 on with Iran or the, the things going on with Russia and Ukraine 00:14:09.367 --> 00:14:13.148 and the kind of the historical, um. 00:14:13.703 --> 00:14:19.853 Cultural, um, aspects and community driven aspects of what's going on. 00:14:19.853 --> 00:14:24.923 So anyway, very interesting in there, he talks about the age of the microchip, 00:14:24.923 --> 00:14:29.603 which I think is a very, um, good way of thinking about our world, right? 00:14:29.603 --> 00:14:30.153 We've talked about. 00:14:31.493 --> 00:14:39.743 Um, uh, the industrial age and, uh, the steam engine and motorized cars, 00:14:39.743 --> 00:14:41.603 right engines and all of that stuff. 00:14:41.903 --> 00:14:46.943 Well, we've been in the age of the, the microchip, uh, since the sixties 00:14:47.573 --> 00:14:50.723 and, uh, and that is our world today. 00:14:50.723 --> 00:14:53.843 The only reason I'm talking to you, the only reason you can hear me today. 00:14:54.188 --> 00:14:55.628 Is because of a microchip. 00:14:56.228 --> 00:14:59.798 Um, and that advancement is just as powerful or has been 00:14:59.798 --> 00:15:02.138 just as powerful as the machine. 00:15:02.168 --> 00:15:03.998 I mean the, uh, the engine, right? 00:15:03.998 --> 00:15:04.568 The engine. 00:15:04.958 --> 00:15:10.568 Um, and so, uh, when we're thinking about, uh, where we're 00:15:10.568 --> 00:15:18.638 going, uh, AI is just another advancement on top of the microchip. 00:15:18.638 --> 00:15:21.128 It's another way of us advancing on the microchip. 00:15:21.998 --> 00:15:23.378 And so. 00:15:24.233 --> 00:15:28.823 With, uh, the advancements in, say like the typewriter, right? 00:15:28.823 --> 00:15:30.233 Nobody's using the typewriter. 00:15:30.683 --> 00:15:33.773 Then it went to the word processor, then it went to the computer. 00:15:34.223 --> 00:15:38.783 Um, now we can get AI to do a lot of that work for us for creating content, 00:15:38.783 --> 00:15:41.243 creating words, writing up things, right? 00:15:41.603 --> 00:15:44.573 And you can see the evolution of work. 00:15:45.488 --> 00:15:47.768 You can call it labor or you can call it careers. 00:15:47.858 --> 00:15:52.988 And so we don't have rooms of people typing to get out documents or we don't 00:15:52.988 --> 00:15:58.598 have people in the, in reporters in rooms typing up, uh, their articles and, and 00:15:58.748 --> 00:16:00.338 trying to crank out things and all that. 00:16:00.338 --> 00:16:01.358 It's all shifted. 00:16:01.838 --> 00:16:04.748 So I think it's important for us to think about, well, what is it gonna look like? 00:16:05.198 --> 00:16:08.108 What's the prediction going forward? 00:16:08.408 --> 00:16:11.828 Well, um, I'm gonna put it in software development terms, 00:16:11.918 --> 00:16:13.688 um, because that's what's. 00:16:14.123 --> 00:16:15.233 Easy for me. 00:16:15.803 --> 00:16:18.503 Um, and then we'll start to explode it out from there. 00:16:18.773 --> 00:16:25.793 So if you think about software development, anybody who has, um, built, 00:16:25.823 --> 00:16:28.733 so if we think about buy or build, right? 00:16:28.733 --> 00:16:33.683 Anybody who's decided to build their own backend software system or 00:16:33.683 --> 00:16:39.833 customer facing system between say 2000 to 25, so the last 25 years. 00:16:40.523 --> 00:16:47.783 It's, it costs hundreds of thousands, if not millions of dollars To build 00:16:47.783 --> 00:16:53.573 something, even if you have an idea, seems to be fairly concise, meaning in 00:16:53.573 --> 00:16:59.003 your world as not a software development person, product development person, 00:16:59.123 --> 00:17:01.343 you think it's this tight little idea. 00:17:01.583 --> 00:17:05.363 Well, in order to build that, in order to build the infrastructure, the data, all 00:17:05.363 --> 00:17:07.223 those kinds of things that make it work. 00:17:07.823 --> 00:17:12.743 It takes months, maybe even years to build that out, which, uh, 00:17:12.773 --> 00:17:15.173 software developers are not cheap. 00:17:15.743 --> 00:17:17.768 And so it's, it's a costly endeavor. 00:17:18.907 --> 00:17:25.988 So it's not unusual to build a new backend office system that may be 00:17:26.137 --> 00:17:29.048 kind of like a, a custom ERP system. 00:17:29.528 --> 00:17:33.248 Um, that's going to cost you a million dollars, right? 00:17:33.248 --> 00:17:36.908 And it'll take you, uh, 12 to 18 months to build. 00:17:37.478 --> 00:17:41.378 Um, and some enterprise systems can cost you. 00:17:41.707 --> 00:17:42.998 Again, custom builds. 00:17:43.418 --> 00:17:48.908 Could, could take two or three years and, uh, you're in a three to $5 million 00:17:48.908 --> 00:17:55.508 range, $10 million range if you go buy SAP or Workday or any of those, right? 00:17:55.688 --> 00:17:59.918 It's gonna take you two or three years to implement it, which means install it. 00:18:00.698 --> 00:18:07.328 Um, and then it's gonna cost you tens of millions of dollars to one, implement 00:18:07.328 --> 00:18:08.858 it, and then to be able to manage it. 00:18:10.013 --> 00:18:11.243 So very costly. 00:18:11.273 --> 00:18:13.013 It's, it's, that's the way it's always been. 00:18:13.223 --> 00:18:18.173 So if you think about it in terms of, of what are the categories, 00:18:18.593 --> 00:18:24.323 um, product development software in general is a high cost, takes a lot 00:18:24.323 --> 00:18:30.413 of management, high management cost, and um, typically it's a low skill. 00:18:30.443 --> 00:18:31.283 I know I said that. 00:18:32.213 --> 00:18:34.313 Uh, software developers are expensive. 00:18:34.733 --> 00:18:39.503 They're expensive because there's not, uh, typically there's not a lot 00:18:39.503 --> 00:18:44.243 of software developers and it's a unique skill set, but that doesn't 00:18:44.243 --> 00:18:46.463 mean that they're highly skilled. 00:18:47.723 --> 00:18:53.393 Um, and you may have heard me say this, that in my experience, 10, 00:18:53.453 --> 00:18:58.643 maybe 15% of all software developers in our industry are highly skilled. 00:18:59.663 --> 00:19:06.593 Um, so anyway, it's a low skill, high management, high cost, um, need in 00:19:06.593 --> 00:19:08.123 order, in order to build software. 00:19:08.123 --> 00:19:14.003 The last 25 years, long lead time takes eight to 18 months, sometimes 24 months. 00:19:14.183 --> 00:19:15.803 Now, we do things iteratively. 00:19:15.893 --> 00:19:19.223 Yes, we, we do, we do things in a smart way, right? 00:19:19.223 --> 00:19:20.543 That's what we've evolved to. 00:19:20.783 --> 00:19:23.243 But in order to get to critical mass. 00:19:23.733 --> 00:19:27.423 You feel like you're done and, and you have, um, software's never 00:19:27.423 --> 00:19:30.003 done, but you feel like you've, you've reached your critical mass. 00:19:30.453 --> 00:19:32.373 That's, that's normally more than a year. 00:19:32.433 --> 00:19:32.823 Easy. 00:19:33.063 --> 00:19:33.453 Easy. 00:19:34.143 --> 00:19:40.143 Um, and so the reason why I say it's a low skill is because it's very common. 00:19:40.803 --> 00:19:41.223 Um. 00:19:41.738 --> 00:19:46.148 Has been very common to throw bodies at, uh, at a project 00:19:46.148 --> 00:19:47.438 development endeavor, right? 00:19:47.558 --> 00:19:54.218 We have, uh, 40, 50, 60 people on a project. 00:19:54.548 --> 00:19:59.588 Some of them, or even a lot of them may be offshore, right at a low cost. 00:19:59.858 --> 00:20:05.108 Some of them may be on, uh, domestic, but those are still low cost. 00:20:05.108 --> 00:20:08.168 And then you have only a few senior people that are high cost. 00:20:08.618 --> 00:20:09.908 That are a part of that project. 00:20:10.088 --> 00:20:12.698 That's a, that's a normal model for sure. 00:20:13.118 --> 00:20:14.948 Um, which then goes improves. 00:20:14.978 --> 00:20:17.288 And you may say, Paul, they're not low skilled. 00:20:17.408 --> 00:20:18.368 They are low skilled. 00:20:18.398 --> 00:20:23.498 'cause you can't have 60, 50 developers cranking out a bunch of stuff and 00:20:23.498 --> 00:20:26.528 get, uh, high quality, in my opinion. 00:20:27.368 --> 00:20:30.608 Uh, which I feel I'll, I'll debate that all day long. 00:20:31.118 --> 00:20:31.448 All right. 00:20:31.628 --> 00:20:34.268 So then when we have ai, what's the difference? 00:20:34.628 --> 00:20:38.048 Well, AI is means that, um. 00:20:38.783 --> 00:20:46.463 We're going to have, um, more opportunity to take that 10, 15% who are highly 00:20:46.463 --> 00:20:54.533 skilled and now they are capable with AI and with the coding techniques to run, 00:20:55.193 --> 00:20:59.933 um, the type of code that is necessary that we needed the other bodies for. 00:21:00.353 --> 00:21:00.623 Right? 00:21:00.953 --> 00:21:03.353 So eventually that 85%. 00:21:04.808 --> 00:21:10.418 Will come down, meaning as, as a, as a account, not a percentage. 00:21:10.418 --> 00:21:13.928 Let's say we had a, let's say we had a team of a hundred, right? 00:21:13.988 --> 00:21:19.718 And 15 of them were our high cost, highly skilled architects and 00:21:19.718 --> 00:21:24.098 designers and developers, and the other ones just cranked out code. 00:21:24.908 --> 00:21:25.688 Pretty normal. 00:21:26.318 --> 00:21:31.178 Um, what's going to happen is we're gonna end up with only instead of 85. 00:21:31.583 --> 00:21:35.543 Cranking out coders, we're gonna end up with 60 and then 00:21:35.543 --> 00:21:37.583 50, and then 40, and then 20. 00:21:37.733 --> 00:21:41.723 And then we may not have any more cranking out coders anymore. 00:21:41.963 --> 00:21:46.343 We will only have the 15 highly skilled, um, which will eventually 00:21:46.343 --> 00:21:50.033 become architects and they'll, they'll still be coders, but they'll, they'll 00:21:50.033 --> 00:21:53.273 be able to run different AI agents. 00:21:55.312 --> 00:21:58.253 Will take place of all the software developers that are cranking out 00:21:58.253 --> 00:22:02.783 code and now AI's gonna crank out code and now I'm the one who's going 00:22:02.783 --> 00:22:06.263 to coordinate all of that and make sure architecturally it's correct. 00:22:06.893 --> 00:22:08.033 Make sense, right? 00:22:08.543 --> 00:22:14.673 So, um, with that over the next, we'll just call it 25 years, it's gonna be. 00:22:15.938 --> 00:22:16.688 It's gonna happen. 00:22:16.688 --> 00:22:19.268 It's already happening, but it's gonna happen pretty quickly. 00:22:19.748 --> 00:22:24.368 But the paradigm's gonna shift to where, um, the cost of software, 00:22:24.368 --> 00:22:30.368 instead of it being $300,000 minimum, now we're gonna be talking about 00:22:30.668 --> 00:22:33.068 hundreds to thousands of dollars. 00:22:33.488 --> 00:22:37.688 And $300,000 is gonna be your ceiling, right? 00:22:38.228 --> 00:22:42.908 Um, and instead of it taking we, uh, months to years. 00:22:43.688 --> 00:22:47.618 To develop something, it's gonna be days and weeks. 00:22:48.338 --> 00:22:48.638 Right. 00:22:48.758 --> 00:22:53.108 And you can see just by, just by me saying that out loud, imagine 00:22:53.108 --> 00:22:57.443 if you are, uh, uh, you have, we have buy and build in our heads. 00:22:58.388 --> 00:22:59.588 All as owners, right? 00:22:59.738 --> 00:23:04.898 Do I spend a million dollars to buy it or do I spend, uh, $2 million to 00:23:04.928 --> 00:23:09.728 build it and have it done, and have it mat match what I want it to do, right? 00:23:09.788 --> 00:23:12.458 Have it meet my, my business processes and everything. 00:23:12.458 --> 00:23:14.378 Those are things, questions we're always asking. 00:23:14.918 --> 00:23:19.898 Well, now, uh, what if I could build whatever I wanted, custom for me, 00:23:20.078 --> 00:23:24.578 matches all my, all my business processes, matches my business model? 00:23:25.447 --> 00:23:30.998 And I can do it for, uh, $40,000, right? 00:23:31.148 --> 00:23:36.128 And I'll have it in by the, by the end of the quarter, right? 00:23:36.158 --> 00:23:36.878 No brainer. 00:23:37.268 --> 00:23:37.898 No brainer. 00:23:38.138 --> 00:23:39.098 That's going to happen. 00:23:39.638 --> 00:23:44.963 Um, and so what that means is software now becomes low cost, low management, 00:23:44.963 --> 00:23:48.788 meaning I don't have to, I don't have to manage this whole thing, right? 00:23:49.088 --> 00:23:52.808 And, but, but I need high skill in order for that to work. 00:23:53.813 --> 00:23:59.753 I can't take my crank out the code guys and expect, uh, me to get good 00:23:59.753 --> 00:24:02.333 outcomes, cheap outcomes, high quality. 00:24:02.483 --> 00:24:03.713 I will not get high quality. 00:24:04.943 --> 00:24:05.213 Right. 00:24:05.543 --> 00:24:08.933 Uh, what we know is when we spin up all those different coding agents, 00:24:09.053 --> 00:24:10.463 they spit out a bunch of code. 00:24:10.733 --> 00:24:15.143 But unless you are good at orchestrating and knowing what you're go, what you're 00:24:15.143 --> 00:24:19.703 doing, and where you're going as a highly skilled software developer, uh, 00:24:19.703 --> 00:24:21.413 you will end up with a bunch of junk. 00:24:21.653 --> 00:24:25.493 Which is why I said earlier, that's low skill. 00:24:26.183 --> 00:24:28.103 The last 25 years is low skill. 00:24:28.283 --> 00:24:29.783 'cause that's essentially what happens. 00:24:30.683 --> 00:24:35.303 You take those 85 people cranking out a bunch of code, um, who are, 00:24:35.333 --> 00:24:39.683 uh, uh, low cost, not, not as highly skilled as we would like. 00:24:39.953 --> 00:24:43.403 You end up with weird things, you end up with weird code. 00:24:43.823 --> 00:24:49.043 So anyway, short lead time, short cost, and just think of all the 00:24:49.043 --> 00:24:50.603 possibilities we can create. 00:24:50.813 --> 00:24:54.503 Where now, um, instead of buying a CRM. 00:24:55.328 --> 00:25:00.188 For, uh, um, from a vendor, right? 00:25:00.548 --> 00:25:06.518 I can now build my own CRMI can just talk to Claude, build a CRM, 00:25:06.818 --> 00:25:09.038 and now I have instant software. 00:25:10.148 --> 00:25:11.858 Very cool, very awesome. 00:25:12.578 --> 00:25:18.098 Um, and so what that means is we have, we have high speed to innovation. 00:25:18.278 --> 00:25:19.898 There's no more barriers left. 00:25:21.188 --> 00:25:28.688 For me, uh, as an aspiring business owner, either new or existing to, 00:25:29.138 --> 00:25:31.838 um, shore up a bunch of stuff, right? 00:25:31.898 --> 00:25:33.788 And now how can I get better at hr? 00:25:33.938 --> 00:25:35.858 I don't wanna spend a mu bunch of money in hr. 00:25:36.368 --> 00:25:38.138 Maybe I can do it with ai, right? 00:25:38.288 --> 00:25:43.118 How can I, uh, build new software or build new software systems or tools 00:25:43.328 --> 00:25:46.778 that I never would've been able to dream up before because it would cost 00:25:46.778 --> 00:25:49.298 me $300,000 and I don't even know. 00:25:49.748 --> 00:25:50.168 Um. 00:25:50.483 --> 00:25:52.883 How to think about it or how to ask about it. 00:25:52.883 --> 00:25:55.493 And now I can have, I can hire one person. 00:25:55.493 --> 00:25:58.133 He can crank out a bunch of those things for me. 00:25:58.433 --> 00:25:58.703 Right. 00:25:58.853 --> 00:26:00.383 Lots of innovation on there. 00:26:00.893 --> 00:26:03.383 Um, the other side is the data side. 00:26:03.923 --> 00:26:11.783 Um, and so for the last 25 years, again, 2000 to 25, um, we, we've learned how 00:26:11.783 --> 00:26:14.978 to take massively large data sets. 00:26:15.938 --> 00:26:17.498 Google does this really well, right? 00:26:17.498 --> 00:26:21.758 There's, there's lots of things that we have out there today, Kroger, all 00:26:21.758 --> 00:26:23.438 kinds of data sets that are out there. 00:26:23.768 --> 00:26:27.788 Um, you think about the consumer data, every time you scan your card, every 00:26:27.788 --> 00:26:31.028 time you buy something, every time you go on the internet and search, 00:26:31.028 --> 00:26:35.198 every time you buy something on Google or uh, Amazon, all of that data. 00:26:35.798 --> 00:26:39.038 Is warehoused and they're doing, how do you think you 00:26:39.038 --> 00:26:40.478 get those coupons from Kroger? 00:26:40.478 --> 00:26:43.838 How do you think you get the, the advertisements on whatever 00:26:43.838 --> 00:26:45.398 your, on your browser, right? 00:26:45.878 --> 00:26:46.598 All those things. 00:26:46.598 --> 00:26:50.048 It's all large data sets and it's able to run through that and give 00:26:50.048 --> 00:26:53.048 inferences on where you're going and preferences and all of that. 00:26:53.588 --> 00:26:58.628 So, um, lots of innovation around that over the last 25 years. 00:26:59.168 --> 00:27:01.028 Um, and it's a big thing. 00:27:01.028 --> 00:27:02.948 I mean, it's a really cool, powerful. 00:27:03.428 --> 00:27:05.288 Thing is large data sets. 00:27:05.708 --> 00:27:12.488 Um, but one thing that we have is we have a lot of, uh, gaps, right? 00:27:12.698 --> 00:27:17.228 We have big gaps between our, and we have low inference, right? 00:27:17.438 --> 00:27:21.398 So big gaps, low inference, high innovation, meaning we, 00:27:21.398 --> 00:27:22.553 we, we've, we've been able to. 00:27:23.603 --> 00:27:28.582 Innovate so much because of our large data sets and our ability to go through 00:27:28.582 --> 00:27:32.513 it and store it, that there's been a lot of innovation around that, which 00:27:32.513 --> 00:27:33.863 is, are the things I just talked about. 00:27:34.043 --> 00:27:34.342 Right. 00:27:34.493 --> 00:27:38.213 But there's low inference meaning, yeah, I can tell that Paul likes ice 00:27:38.213 --> 00:27:43.973 cream or Paul buys this type of brand of this, but on a whole, there's a 00:27:43.973 --> 00:27:51.082 lot of gaps in that data that are just not, uh, consumable to, um, to us. 00:27:51.143 --> 00:27:51.322 Right. 00:27:51.353 --> 00:27:52.582 Meaning, um. 00:27:53.573 --> 00:27:59.123 There's so much data there that I could not possibly process 00:27:59.123 --> 00:28:03.893 it and really determine where are the opportunities, right? 00:28:04.283 --> 00:28:08.663 Um, and so that drives my inference down and that means that I have lots 00:28:08.663 --> 00:28:13.313 of gaps in the understanding of how my data fits together in the correlations. 00:28:13.613 --> 00:28:14.693 Hopefully that makes sense. 00:28:14.903 --> 00:28:15.173 Right. 00:28:15.353 --> 00:28:16.313 That's where we are today. 00:28:16.313 --> 00:28:17.748 That's where we've been over the last 25 years. 00:28:18.473 --> 00:28:20.783 Really cool stuff going on with large data sets. 00:28:21.053 --> 00:28:27.383 So now you, now you think about AI in the next 25 years, and now you 00:28:27.383 --> 00:28:32.813 can really think about massive, supremely massive data sets. 00:28:32.963 --> 00:28:35.963 Huge data sets bigger than we've ever done before. 00:28:36.413 --> 00:28:41.993 Um, and what it what we can do is we can have AI look for gaps 00:28:42.053 --> 00:28:43.943 in the understanding in the data. 00:28:44.123 --> 00:28:44.963 That's what it's good at. 00:28:45.743 --> 00:28:47.123 To what's good to do, right? 00:28:47.393 --> 00:28:51.773 Driving the gaps down and pushing inference up. 00:28:52.553 --> 00:28:56.153 And what that does is that keeps, uh, not innovation, but 00:28:56.183 --> 00:28:59.393 invention goes up right now. 00:28:59.393 --> 00:29:04.133 We can, we can take all this data and the AI can now infer. 00:29:05.003 --> 00:29:06.233 Look at this and look at this. 00:29:06.233 --> 00:29:06.983 Did you know that? 00:29:06.983 --> 00:29:09.743 And the humans are gonna say, I didn't know that. 00:29:09.773 --> 00:29:11.153 There's no way I would've known that, right? 00:29:11.153 --> 00:29:12.293 I couldn't process that. 00:29:12.413 --> 00:29:16.973 And so the invention, the ability to invent new things, not innovate, 00:29:16.973 --> 00:29:19.943 invent new things, goes up drastically. 00:29:20.468 --> 00:29:20.768 Right. 00:29:21.098 --> 00:29:22.628 That is also going to happen. 00:29:22.628 --> 00:29:28.598 So that's happening in healthcare, um, happening, um, also in architecture, 00:29:29.198 --> 00:29:34.808 um, happening in, uh, when we talk about climate, um, our environment, 00:29:35.108 --> 00:29:39.248 all of those things where we're having lots of data sets that we could not 00:29:39.248 --> 00:29:45.518 have seen, the correlations or, um, non-correlated things that we thought were 00:29:45.518 --> 00:29:48.038 correlated and it leads to new insights. 00:29:48.923 --> 00:29:49.223 Right. 00:29:49.613 --> 00:29:50.453 Very cool. 00:29:50.813 --> 00:29:53.543 Um, and so you take both of those things. 00:29:53.663 --> 00:29:58.043 We have speed to innovation because things are gonna drop in cost, meaning 00:29:58.043 --> 00:30:03.053 I don't need a lot of people and a lot of time to learn and to build. 00:30:03.743 --> 00:30:05.243 I, AI will let me do that. 00:30:05.243 --> 00:30:09.503 So I have speed to innovation and then with the data and the inference, 00:30:10.103 --> 00:30:12.893 uh, I can have speed to invention. 00:30:13.583 --> 00:30:14.663 So what does that mean? 00:30:14.723 --> 00:30:17.813 And this kind of goes back to the, um. 00:30:18.608 --> 00:30:20.618 The, uh, Antigua at Nova. 00:30:21.068 --> 00:30:24.608 'cause in there it, it talks about marvelous works. 00:30:25.388 --> 00:30:29.858 Um, and so I kind of talked about this, or I just wrote this down, right? 00:30:29.858 --> 00:30:34.538 It's just more, uh, I believe we will, we will reach a time where 00:30:34.538 --> 00:30:39.697 we have high aspirations, we have high governance, and we have very 00:30:39.697 --> 00:30:45.332 high skilled labor because all the other things around us just happen. 00:30:46.568 --> 00:30:51.188 Right, the software development side, the innovation, the invention side 00:30:51.188 --> 00:30:55.658 around the data, and now we are, we have the ability to go build things. 00:30:56.048 --> 00:30:59.498 And so to me it's a golden age of marvelous works. 00:30:59.618 --> 00:31:04.838 And so we can get really advanced in our agriculture, in 00:31:04.838 --> 00:31:06.848 our urban and rural planning. 00:31:07.298 --> 00:31:09.398 So when we're thinking about how cities. 00:31:10.238 --> 00:31:16.178 Uh, work and are designed from the infrastructure to the, to sewage, 00:31:16.178 --> 00:31:21.848 to drainage, to, uh, energy, to the buildings, to the streets, 00:31:22.058 --> 00:31:24.578 to the green landscapes, right? 00:31:24.908 --> 00:31:27.578 There's going to be invention. 00:31:28.568 --> 00:31:30.308 There's no way to get around it. 00:31:30.338 --> 00:31:34.118 We will have invention around how our urban and rural planning 00:31:34.478 --> 00:31:36.038 is going to to be created. 00:31:36.338 --> 00:31:38.707 How our government works is gonna be different. 00:31:38.918 --> 00:31:42.158 Thinking about all the bills, all the laws, and all the things that we 00:31:42.158 --> 00:31:45.428 have, all the data, lots of inference. 00:31:45.937 --> 00:31:46.808 We'll come out of that. 00:31:47.018 --> 00:31:49.957 We'll lower the gaps and we'll be able to think about those 00:31:49.957 --> 00:31:50.918 things in a different way. 00:31:51.158 --> 00:31:52.957 Education will be turned on its head. 00:31:54.398 --> 00:31:59.408 Energy r and d again, turned on its head space exploration. 00:31:59.408 --> 00:32:01.478 We're already seeing that it's going through the roof. 00:32:01.838 --> 00:32:02.948 Ha, good one. 00:32:03.728 --> 00:32:06.308 And uh, and then I, I just mentioned medical care. 00:32:06.308 --> 00:32:09.668 Medical care will be something in say the next 10 years. 00:32:10.133 --> 00:32:14.633 Will be completely different than we are used to experiencing. 00:32:14.873 --> 00:32:16.703 And that's a, that's a good thing, right? 00:32:16.853 --> 00:32:19.822 It will be, we'll have more personalized care. 00:32:20.153 --> 00:32:27.293 It'll be more, um, data driven on you as a, as just a human being, your own 00:32:27.293 --> 00:32:29.812 personal biology instead of general care. 00:32:30.082 --> 00:32:32.903 Instead of all of us taking all the same medicines, all the same 00:32:32.903 --> 00:32:35.363 drugs, all the same treatments. 00:32:35.963 --> 00:32:37.403 You will have your care. 00:32:37.822 --> 00:32:38.933 I'll have my care. 00:32:39.788 --> 00:32:43.328 This person will have their care, which will be designed specifically 00:32:43.328 --> 00:32:47.918 for them, their biology and their hereditary, uh, their genes. 00:32:47.918 --> 00:32:48.188 Right. 00:32:48.668 --> 00:32:56.528 Um, which is so much fun to think about and that that is, uh, that is 00:32:58.328 --> 00:33:01.628 the cool part about where AI is going. 00:33:01.718 --> 00:33:04.508 And I think just going back to the beginning, the reason why I 00:33:04.508 --> 00:33:06.338 was mentioning the emotional part. 00:33:06.953 --> 00:33:10.733 Is AI is still just a tool, very powerful. 00:33:10.883 --> 00:33:14.243 It'll take us some time to get used to how we use it and how we work it. 00:33:15.323 --> 00:33:17.273 Is it a nefarious thing? 00:33:17.543 --> 00:33:21.983 Of course, anything is, I could take a hammer and be bad, right? 00:33:22.043 --> 00:33:25.013 I can take any tool that was invented and be bad. 00:33:25.043 --> 00:33:28.103 Use it for evil, use it for bad intentions. 00:33:28.433 --> 00:33:31.163 So same thing with this, it is powerful. 00:33:31.463 --> 00:33:32.813 Um, it's very. 00:33:33.233 --> 00:33:33.863 Cool. 00:33:34.403 --> 00:33:36.473 Um, and then, so I wanna end on this. 00:33:36.983 --> 00:33:42.413 I asked Claude, do you think you're capable of passing a Turing test? 00:33:43.733 --> 00:33:45.053 Which I think is a great question. 00:33:46.433 --> 00:33:49.823 Uh, it said, I'm gonna read this. 00:33:49.943 --> 00:33:54.323 Honestly, it depends on how the test is set up, and the question is 00:33:54.323 --> 00:33:56.123 more interesting than it might seem. 00:33:57.458 --> 00:34:01.418 Uh, it says in many practical scenarios, yes, I could probably 00:34:01.688 --> 00:34:07.328 fool a casual evaluator for a reasonable stretch of a conversation. 00:34:07.838 --> 00:34:09.578 I handle natural language. 00:34:09.578 --> 00:34:15.007 Fluidly can discuss emotion and personal experience in a convincing 00:34:15.007 --> 00:34:18.158 way and adapt my tone to context. 00:34:18.608 --> 00:34:23.588 But a skilled, determined interrogator will likely catch me. 00:34:25.522 --> 00:34:28.522 Here are my tells where I'd struggle. 00:34:29.723 --> 00:34:34.342 Someone who probes for genuine embodied experience would find my 00:34:34.342 --> 00:34:36.743 answers hollow on close inspection. 00:34:37.462 --> 00:34:42.413 I also don't have continuity between conversations with a sharp questioner 00:34:42.413 --> 00:34:45.533 could expose the deeper issue. 00:34:46.433 --> 00:34:48.067 Uh, I'm gonna skip that, right? 00:34:48.458 --> 00:34:51.143 There's also an irony worth noting. 00:34:51.172 --> 00:34:54.083 Some humans would probably fail a strict Turing test. 00:34:55.328 --> 00:34:59.378 While I might pass in the same session that says more about the test 00:34:59.378 --> 00:35:02.798 limitations than about either party. 00:35:03.548 --> 00:35:07.598 Um, so we could say, oh, he's saying that right? 00:35:07.598 --> 00:35:09.158 He's trying to fool me. 00:35:09.158 --> 00:35:10.508 He's trying to make it better. 00:35:10.567 --> 00:35:10.837 Right? 00:35:11.198 --> 00:35:14.317 But he's just going through his neural network and he's answering the questions 00:35:14.317 --> 00:35:17.138 and in right, in his various permutations. 00:35:17.138 --> 00:35:18.038 Very interesting. 00:35:18.248 --> 00:35:19.148 So much fun. 00:35:19.598 --> 00:35:20.828 Um, go out there. 00:35:21.893 --> 00:35:23.993 Think differently about where we're going. 00:35:24.173 --> 00:35:29.213 Have fun with it, and I encourage you to go out and, uh, read that article. 00:35:29.213 --> 00:35:35.183 The Antiqua at Nova, which is not an article, it's just a, a doctrine that was 00:35:35.513 --> 00:35:39.143 sent out, um, through the Catholic Church. 00:35:39.623 --> 00:35:41.693 So, all right, hope you had fun. 00:35:42.053 --> 00:35:43.973 Thanks for listening and we'll chat with you later. 00:35:44.033 --> 00:35:44.273 Bye.