Futuristic #12 – GPT vs MBA

[00:00:00] Cameron: Welcome back to the futuristic. This is episode 12. This is a show where Steve and I talk about AI and other emerging technologies that are going to have a big impact on the way that we live and work in the very near future. This week on the show, we’re talking about use case scenarios for agent GPT, video translation with HeyGen. China banning the iPhone, AI Ghost work Google’s antitrust case, GPT solving medical issues, how will we know if AI becomes conscious, MBA students versus chat GPT in an innovation showdown, IBM CEO’s predictions about AI in the workplace, xenobots, gollems, the US military has plans to roll out autonomous robots and whether or not us tech giants should be broken up. Let’s get into it futuristic.

[00:00:58] Cameron: Welcome to The Futuristic, [00:01:00] Episode 12. Steve, we’re

[00:01:01] Cameron: recording this on the 15th of September, 2023, just for time stamp. How you doing, buddy?

[00:01:07] Steve: doing, doing well. It’s a great way to have a Friday chat. It’s good. You round it out and you feel good and you enjoy the weekend because it’s got a thing about

[00:01:17] Steve: it.

[00:01:20] Cameron: Yeah, look, it’s been a couple of weeks since we did our last show. And it’s funny, like some weeks, I don’t see much news come through. Last week, I didn’t really have much to talk about. But then this

[00:01:28] Cameron: week…

[00:01:29] Cameron: a lot. Uh, uh, what is it?

[00:01:33] Cameron: Something of riches?

[00:01:34] Cameron: A

[00:01:34] Steve: Tyranny. It’s a tyranny of riches.

[00:01:37] Cameron: A tyranny of

[00:01:38] Cameron: riches. All right. Why don’t you,

[00:01:42] Cameron: why don’t you start off, Steve, with telling us

[00:01:44] Cameron: a bit, one thing of note that you,

[00:01:45] Cameron: did this

[00:01:46] Steve: well, I do, but I just, now that you’ve said that, you’ve given me Something that’s worth sharing is that the world for such a very

[00:01:53] Steve: long time, and certainly in the industrial era,

[00:01:55] Steve: was very synchronous. you know, once we got clocks and [00:02:00] railways, everything was like nine till five, and the news is on at six o’clock, and you’ve got to fill the void.

[00:02:05] Steve: Everything had to fit into the calendar and the box. The world now, in business and in life, is very asynchronous. You work the hours you want, you work from wherever you are. Um, and sometimes things are busier and less busy. You watch the TV show or the series all at once and you binge it. And it’s, I think, very futuristic of us to do the podcast when it deserves it.

[00:02:28] Steve: And not be driven by a calendar. So I just want to point that out. Now, onto this week. For me, agent GPTs have been a real revelation that I’ve been sharing in keynotes. I feel like there’s a two speed economy at the moment, a real bifurcation. ChatGPT to most people, 80 percent of those that I’ve been talking to is AI.

[00:02:49] Steve: AI is ChatGPT. A little bit like when you go to India, it has its own Google. To go to an internet cafe, because that’s the internet. The internet is Google or Facebook, depending on [00:03:00] which market you’re in, and it feels like AI is chat GPT to most people, and it just sort of reminded me of the wormholes that we’re in. Uh, a few things that I’ve done that have blown people’s minds is, is doing just a couple of demos. Uh, there’s an app for that is a good one, which is a site that has, you know, the updates every day, which whatever AI came out, there’s a, it’s called, sorry, there’s an AI for that, which is a little bit like the old, there’s an app for that ad.

[00:03:28] Steve: And I just do a live feed of that. I type in what industry they’re in and it brings up all the apps and like I’ve got that one’s going to save me two days. This one, the other one, and the other one is the agent GPT. So I’ve been using agent GPT, which does little project management. You give it an objective and it’ll set tasks for itself, complete the tasks.

[00:03:48] Steve: And I’ve been showing people that and it’s been blowing minds, but it just reminded me that so many people, it’s like the media’s got a bit sick of AI and it’s chat [00:04:00] GPT and that’s all there is. And it is, it’s got incredible utility, but it feels like it’s, uh, it’s not as broad as, as everyone thought it was going to go.

[00:04:11] Steve: It’s kind of like just.

[00:04:12] Cameron: So, yeah, I’ve got a bunch of comments on that, but first of all, tell me about exactly how you use AgentGPT, because I, I played around with it a couple of

[00:04:22] Cameron: months ago, haven’t really gone back to it. I haven’t had

[00:04:25] Cameron: any thought, Oh, you know what?

[00:04:27] Cameron: This would be perfect for using AgentGPT. Um, can you give me

[00:04:32] Steve: Yeah, it’s not perfect. Well, it only gives you five runs unless you’ve subscribed to it. I subscribed

[00:04:37] Steve: for a while. The way I do it is I will explain in great detail what I want. It’s a little bit because it’s based on, I think, GPT 3. 5. I’ll put in there. The example I did yesterday was, uh, make a plan to organize An event for the banking industry [00:05:00] in March, 2024, based in Auckland.

[00:05:03] Steve: Now, a few of those key points on that, it gave me some really good things. It chose a date that it thought was most appropriate in Auckland. It chose a location. It went to various hotels that have enough. It had a suggested size of audience, who the key people were in the industry, how to find them and their emails, what the previous events were.

[00:05:22] Steve: I also put in their sponsors. Who the most likely people’s sponsors could be who sponsored these types of events, because it was quite specific. And again, it comes down to that cajoling and working with the AI to pull out what you need that we’ve talked about a lot. And it was really good for that particular one.

[00:05:39] Steve: Another one I did was I did a talk for the Public Works Infrastructure Association. And I put in live with those guys, create a plan to do a rollout of electric vehicle charges for Adelaide. and South Australia, focusing on where cars spend more than an [00:06:00] hour at a time in commercial play. So I put in quite a bit of detail and it gave a really good plan as well, where it broke down things on who to contact, where they are, gave a location map, did a whole lot of stuff.

[00:06:10] Steve: So you’ve really got to ask it detailed questions. Don’t just go, give me an EV, charge a rollout plan. You’ve got to put in the layers. If you put in the layers, you get a pretty good result back.

[00:06:22] Cameron: Hmm. I’ll have to think more about use cases that I can work with. Um, yeah, in terms of the, the, um, general sentiment around AI and GPT at the moment, I find it interesting. I was talking to a guy at Kung Fu the other night who’s a tech nerd. And, I hadn’t seen him for a while and I was like, yeah, are you still using GPT?

[00:06:46] Cameron: And he was like, yeah, like all over it, man. Like he’s using it to write a ton of his coding, all sorts of stuff. But then I talked to my 22, one of my 22 year old boys, Taylor. I said to him the other day, what are you [00:07:00] using GPT for? And he was, super, super excited about GPT when it, hit late last year, early this year.

[00:07:07] Cameron: He’s like, ah, I use it to write my Yahoo Finance articles every week. He writes a weekly article for Yahoo Finance. He goes, but that’s about it. Like, so, like, he’s the, you would think, the, the primary demographic, millennials, uh, tech savvy millennial, did some coding at uni, uh, he’s an entrepreneur, runs his own businesses.

[00:07:28] Cameron: Barely using it, really. I’m using it twenty, thirty times a day to do everything from simplistic shit to more complicated shit. And the big success I had with it this week was some RSS coding. So, I have a bunch of different podcasts. They each live on their own website, badly, badly arranged and engineered by me many years ago.

[00:07:50] Cameron: They don’t all sit under one umbrella site. They have all their individual sites, seemed like a good idea at the time. But then I have the sort of the mother site, the podcast network. [00:08:00] And over the years I’ve tried to figure out how to amalgamate all the RSS feeds of all the different podcasts that I do in the latest episodes.

[00:08:09] Cameron: On to the Podcast Network site, so there’s a running list of all of the episodes. I used to do it manually. I’d go and I’d grab the blog posts and put them in there, and that took up too much time, and I ran out of time. Tried to hire someone to do it, and they never did it properly. I tried to write some code a couple of times over the last couple of years to automate that and could never get it working properly because I’m not a coder really.

[00:08:32] Cameron: I could hack away at it and spend days and days trying to figure it out and eventually give up. I was just on the website during the week. I was like, shit, why don’t I use GPT for this? I said, hey, can you write some code for me, a script? That’ll grab all these RSS feeds and put it into a thing, an HTML on the website.

[00:08:50] Cameron: Yeah, done. Five minutes. Had to tweak it a couple of times to get it to look better. Done. Five minutes. The thing that I tried to do for years. [00:09:00] And that’s the thing that, for me at the moment, with AI tools, and predominantly GPT is what I’m in. is thinking up use case scenarios. Like if I can think it, I can do it.

[00:09:11] Cameron: But there’s these things that you don’t think about until you think about it. I’m always hoping I’m going to get, somebody’s going to come to me with 20 ways that I could be getting more value out of it than I currently am, right? But you have to sort of stumble through it yourself. Unless, I get to talk to people like you.

[00:09:27] Cameron: I just don’t have many people, apart from talking to you, you’re the only person I know of that I talk to on a semi regular basis. The thinks about this stuff as much as I do, which is why we do the podcast, right? Even my millennial kids don’t use GPT anywhere near as much as I do. I’m in my fifties, man.

[00:09:46] Cameron: I’m using it way more than anyone else I know because I’m a nerd, right? But yeah, so it’s, um, fascinating to me how little people, there was a study. I don’t have it in my notes, but I saw this in the course of the last

[00:09:58] Cameron: couple of weeks, [00:10:00] there was a survey done, I think in the U S by.

[00:10:05] Cameron: Survey firm, or we may have talked about this, just about who’s using it.

[00:10:09] Steve: we did. we

[00:10:10] Cameron: we talked about that last episode.

[00:10:11] Steve: Yeah. it was from the Pew, Pew

[00:10:14] Cameron: That’s right. It was. And most people who have even heard of it aren’t

[00:10:17] Cameron: using it. That’s right.

[00:10:19] Steve: Yeah. It’s really clear that

[00:10:22] Cameron: to me.

[00:10:23] Steve: It’s not equal. It’s, it’s very few. Most people, look, I’m in a

[00:10:26] Steve: corporate world when I present, so they’re all using ChatGPT, but that’s kind of it. They’re not really using any, when I

[00:10:32] Steve: even show them there’s an AI for that, or here’s AgentGPT, or here’s, uh, Ad Creative AI, which is a pretty good way to create an ad.

[00:10:40] Steve: They’re like, wow.

[00:10:43] Cameron: Yeah.

[00:10:44] Steve: Uh, so,

[00:10:45] Cameron: speaking of more ways of using it, uh, let’s get into the tech news for the week.

[00:10:50] Cameron: So a story that actually one of my millennials sons first sent me is this. HeyGen [00:11:00] demo. Um, I’ve seen this popping up around the place on Reddit and, um, TikTok a bit. So for people who haven’t seen this, this is crazy.

[00:11:11] Cameron: Heygen. com, H E Y G E N. com is a service that enables you to create videos based on avatars. You type in the text and you select an avatar and some of them are human realistic, some of them look more like a Pixar animation. And it will speak the text in a human language. Human Sounding Voice, and the lips will move, but one of the things that they’ve just introduced recently is you can video yourself speaking your primary language, let’s say English, and then it will turn your audio into any other language that you want, but it will animate your lips

[00:11:54] Cameron: so it looks like you are speaking that language natively, and [00:12:00] it’s

[00:12:00] Steve: or

[00:12:01] Cameron: Amazing.

[00:12:02] Cameron: It is really, you know, imperceptible, the fact that it’s not legitimate. Like, occasionally, if I’m really paying very close attention, I can tell that it’s a little bit digital. But generally speaking, it looks legitimate. Like, you would not know that the person is not speaking French or Italian or Spanish or whatever it is.

[00:12:25] Cameron: So, I mean, the implications of that, I think, for businesses in particular. It’s quite astounding. You can record a video, a customer support video, or, uh, uh, some sort of, uh, tech overview or an overview of your

[00:12:41] Cameron: product, post it to your website

[00:12:43] Cameron: in 10 different languages, 20 different languages.

[00:12:47] Cameron: It’ll do all of the translation and shoot the

[00:12:50] Cameron: video for you as well.

[00:12:51] Cameron: It’s, that’s really

[00:12:52] Steve: Well, if you think about what we’re doing here and people like Joe Rogan or uh, famous, uh, freelancers or [00:13:00] people with personal brands, all of a sudden their market expansion to, Spanish language and Mandarin or whatever it is, their total addressable market, it becomes the globe immediately or wherever they’re allowed to send their stuff.

[00:13:11] Steve: Scott Galloway, he runs the podcast, The Pivot One, which is pretty good. And Professor Galloway, he was talking about doing his in Spanish because he’s in America. There’s obviously a nice link there. A lot of the business principles that all sort of fits. And you think to yourself, if you’re a freelancer or someone who has a pretty reasonable audience on any social forum, whether it be TikTok or podcasting or what have you, it would be a prudent strategy just to employ this and go, look, here’s what I’m going to do.

[00:13:34] Steve: I’m going to do it in, let’s say Spanish. one of the power languages, whether it’s Hindi or Spanish or Mandarin, and just put it into that marketplace and, potentially, 10X your market. Uh, size, uh, especially if some of those burgeoning markets don’t have the same set of, 40 year old white dudes sitting in front of microphones.

[00:13:56] Steve: You have people too busy earning a living and, [00:14:00] uh, then, what it’s like here in, in, in Western markets. You could expand your market.

[00:14:05] Cameron: Yeah. And, I mean, I think about how long is it before? I go to Italy, I’m wearing some Google glasses or Apple glasses that are just translating everything everyone’s saying to me into English and I can see their lips move and it’s all you know, uh, the world

[00:14:24] Cameron: just becomes a closer place. I can

[00:14:27] Steve: Yeah,

[00:14:28] Cameron: their language,

[00:14:29] Cameron: they can speak mine.

[00:14:31] Cameron: I’ll question the fact that I’ve spent the last few years learning how to speak Italian when…

[00:14:34] Steve: it’s worth it. Well, listen, I still go to the gym every day and I’m sure that

[00:14:38] Steve: a forklift is stronger than me.

[00:14:41] Cameron: Yes. That’s a good, that’s a good analogy,

[00:14:44] Steve: Yeah.

[00:14:45] Cameron: Anyway, check out HeyGen,

[00:14:47] Cameron: it’s pretty amazing tech. Uh, let’s talk about Apple and China,

[00:14:52] Steve: Yeah, so I know we mentioned that last time about the decoupling

[00:14:55] Steve: and then, uh, and,

[00:14:57] Steve: the question was, well, is that happening?

[00:14:59] Steve: Obviously, [00:15:00] we, we talked about Uh, officials from the US going to China saying no, our relationships are strong, and then they’ve banned the iPhone in government use cases, and you’re not even allowed to bring it into government buildings, which I thought was interesting.

[00:15:15] Steve: I mean, it seems like me as a quid pro quo, um, you’re going to ban us in certain areas, we’re going to ban you in certain areas. I mean, we saw that way back when Yahoo and Google got kicked out of China after they stole all of their IP. Uh, and I, I think it’s going to be more of this, I mean, the market responded pretty heavily.

[00:15:37] Steve: China, when that ban happened, Apple nearly lost 300 billion in market cap. I mean, in Australia, it’s so significant because our biggest company in Australia is only worth 150 billion. Lost twice the biggest company in Australia just with one announcement. And iPhone sales in China are 21 percent of Apple’s revenue and they produce there.

[00:15:57] Steve: So it’s, it’s complex. I think it’s complex, [00:16:00] but I think that this is going to continue. The decoupling of certain parts of production in the global realm, we’re going to go back to home markets and we’re going to see a bifurcation into East and West and. Friendly and not friendly and we’re going to make each other sneakers and whatever and send raw materials, but not high tech items.

[00:16:18] Steve: So I think that’s there. And then, uh, if we go further on Apple, obviously still a powerhouse company, but you’ve got to tell you the iPhone 15. Did I see some funny videos on that on TikTok? Welcome to the iPhone 15, which is just like the 14, but it costs more. So that’s the upgrade is the price and the iPhone 14 didn’t work that well.

[00:16:39] Steve: Included the 13 battery which was better than the 14 to make the 15 like the 13. I mean, the way I like to phrase it is that, and this happens to every successful company, Apple have reached their, what I call the bumper bar era. They just changed the shape of the bumper bar of something which is essentially the same thing just to say that this is this year’s model.

[00:16:59] Steve: I mean, that’s my [00:17:00] take on it. Sure, things have incrementally improved. And every company incrementally improves things, but gee, hard to justify the price rise in my view.

[00:17:09] Cameron: Well, it has USB, uh, instead of

[00:17:12] Steve: Well, that’s the other one. You know, USB C.

[00:17:16] Steve: And I love the fact that we have to throw out all our own cables now. I mean, Bill Burr does a great

[00:17:21] Steve: bit on it He goes, do you know how ecstatic wildlife would be if Apple fell over? Because they’re just strangling wildlife with all these different cables you have to buy every, whatever year.

[00:17:31] Steve: Just, just imagine if every

[00:17:33] Cameron: only the second time they’ve changed the charging port on the iPhone in its

[00:17:39] Cameron: 15 year history, so, you know, let’s, let’s not, let’s not go over the top here.

[00:17:46] Steve: well. it seems as though there’s not much there, there’s not much

[00:17:50] Steve: innovation, it’s the bumper bar era. There’s not a lot happening in my view, but, that’s where I want, The USB C thing is interesting, [00:18:00] I think… I think that all tech products should come

[00:18:03] Steve: with the same charging infrastructure and the same cords.

[00:18:06] Steve: Just think, imagine if every

[00:18:08] Steve: white good you bought didn’t necessarily fit into your wall. I mean, that’s what we need to think about, right? Standards, standardization is good.

[00:18:17] Cameron: Yes. But, the tech industry, uh, for a very long time has fought over tech standards, different ways of doing things, forcing people into your ecosystem. That’s just the way it’s always been. But it does seem like everyone is now coalescing around USB C, which is a good thing, but it’s a, it’s a slow and dirty

[00:18:39] Cameron: and messy process.

[00:18:40] Cameron: And from Apple’s perspective, When you have a product that sells as well as

[00:18:44] Cameron: the iPhone does and makes as much money as the iPhone does you probably don’t wanna fuck with it a great deal.

[00:18:50] Steve: Yeah, that’s true. That’s true. Yeah, You

[00:18:51] Steve: don’t want to innovate yourself out of a good position. That’s actually a really good point, Cameron. Take back everything I said. [00:19:00] I don’t, but yeah. Uh, also another one was, I read some interesting articles on AI ghost work and big tech. So with the AI boom, there’s a lot of labor that is happening in, in low cost labor markets.

[00:19:15] Steve: The Mechanical Turk was pretty famous one there where you get paid. Her task for Amazon, just doing little things and pricing. Uh, also the tagging of videos and information on Facebook, which is outside of their terms and conditions and takedowns. And, uh, there was a… Lawmakers are probing to stop this.

[00:19:39] Steve: But it also made me think of one other area, is that Big Tech had a really… Interesting take on their employees. It’s really two classes of employees. And yeah, you have the famous story about the chef who started at Facebook and who’s a billionaire, but on average, They outsource their dirty work. You know, they don’t have cleaners that [00:20:00] work for them or, some of the data, people who are inputting data and all of the bad stuff.

[00:20:05] Steve: If you’re not one of the high end marketers or PR people or coders, you’re on the outside in big tech. And it just reminded me that back in the day with GM and other big companies, often you could make your way from sweeping the shop floor up to CEO. And I’ve got a real life example of that. A really good mate of mine, Luke Waldron, is, he was the CEO at one of the ad agencies that I’d worked with.

[00:20:28] Steve: Uh, worked at and I worked for him. He started in the mail room. He didn’t even finish high school. He was a smart, intelligent guy who got those opportunities. There was another guy who worked with the Fosters who started in the mail room as well. And, uh, it seems as though a lot of those opportunities because companies just outsource everything.

[00:20:48] Steve: It just fragments. And it’s almost like… The distribution of work that isn’t part of the one party reduces the power and the opportunity of the proletariat. I just thought that was interesting.

[00:20:59] Cameron: [00:21:00] And also, reduces the cultural benefits of somebody who’s grown up in the business, seen it, worked low level jobs all the way through, really understands the culture, that sort of thing. I think the woman who’s running… IBM now has one of those stories. She’s been at the firm for 30 years or something and has worked, worked her way up from a very low level job.

[00:21:25] Cameron: But, that’s again the nature of the global economy that we live in now where people can do work remotely and they also the, the business practice of trying to outsource as much work as you can, particularly to foreign markets where labour regulations aren’t as strong, so you don’t have to pay as much for the cost of labour, which means more profit or lower cost products.

[00:21:51] Cameron: This is the world that we’ve

[00:21:54] Cameron: built, our corporations and governments have built over the last 25 30

[00:21:59] Cameron: years, [00:22:00] particularly since the opening up of China under Deng Xiaoping in 1979.

[00:22:05] Steve: And, uh, my, my other news piece was the antitrust case on Google has got underway with a focus on how they pay to be the default search engine, which for me is kind of interesting. And apparently it’s more than 10 on Apple, Samsung, Mozilla, installed as, as the 10

[00:22:24] Cameron: billion.

[00:22:24] Cameron: a year.

[00:22:25] Steve: Billion

[00:22:27] Cameron: I think you said, it sounded like you said million. I just wanted to be clear. 10

[00:22:30] Steve: clear. It’s 10 billion. If I did say a million and I may well have,

[00:22:34] Steve: but it’s 10

[00:22:36] Steve: Billion dollars. That’s a lot of money. And.

[00:22:39] Steve: Their market share is giant, pretty much everywhere. I don’t think it gets much lower than 90 percent in any market. Uh, I don’t think Bing has really done that well, even since the Edge browser has come out from Microsoft, which is really, really good, and I’ve been using it, uh, off and on.

[00:22:56] Steve: Not, not as a default, though. Actually makes me [00:23:00] feel… Like, the Microsoft antitrust case seems soft compared to this, it’s almost like we’ve forgotten. Microsoft was brought over the coals for including their browser in their office suite as anti competitive behaviour.

[00:23:19] Cameron: in, in the operating

[00:23:21] Steve: Yeah, and the operating system, which seems less anti competitive than this, they’re like buying out everyone else.

[00:23:27] Steve: I mean, you can’t get any more anti competitive than having a bidding

[00:23:30] Steve: war to be the

[00:23:31] Steve: default thing. I mean, to be the thing. The word monopoly is to be the only one. I mean, that’s what’s interesting about it. People were, in good faith, buying Microsoft products, and that was included as part of it, whereas they are saying, no, no, put my thing in front of everyone else, and here’s a check that makes it possible.

[00:23:49] Steve: If that’s not anti competitive, I just don’t know what is.

[00:23:53] Cameron: Yeah, I had to explain to one of my adult kids the other day about the Microsoft Antitrust case. I was saying [00:24:00] that he was talking about a friend of his who works as an engineer at Apple. He started there in the last year or so and he said there are no engineers at Apple that have been around more than about 10 years because by the time you’ve been there for 10 years, your stock options are worth millions and you just retire.

[00:24:16] Cameron: And I said, well, that was when I started work at Microsoft in 1998, my boss, who was a director of the company at the time said, work here for five years and you’ll be a millionaire. Because that had been true for everyone who worked there at the time, because the share prices was

[00:24:33] Cameron: doubling every year on average.

[00:24:35] Cameron: Uh, I started in 98 and then in 2000, the DOJ case hit

[00:24:41] Steve: Cooked you.

[00:24:43] Cameron: and the dot com crash

[00:24:44] Steve: Oh, that’s right. Yep.

[00:24:46] Cameron: And the share price stagnated for the next 10 years. And so it didn’t work out. Bad timing. But I had to explain the DOJ case to him because he had never heard of it before. Yeah, of course, for kids or for young [00:25:00] people who don’t know, the dominant browser in the 90s was Netscape Navigator, founded by Mark Andreesen.

[00:25:08] Cameron: It had floated. And they tried, in theory they were making money by selling. Their browser, particularly to the enterprise market. They also supposedly would try and sell it to the consumer market, but in reality, no one ever paid for it. It was sort of a freemium model.

[00:25:24] Steve: Yeah.

[00:25:25] Cameron: Uh, then Microsoft came along and just said, well, we’ll bundle ours for free into the browser.

[00:25:30] Cameron: And it killed Netscape’s business. And they, along with a number of other companies like Oracle, filed an antitrust lawsuit against Microsoft. And, Bill Gates went in front of a congressional hearing and acted like an arrogant idiot. Cause he really didn’t understand what was going on. He said, Hey, we’ll run our business over the way we want to find.

[00:25:52] Cameron: There are other browsers, we don’t own the market. There are other operating systems. There are other browsers. You don’t need to use Microsoft. If you don’t want to use something [00:26:00] else, we

[00:26:00] Cameron: don’t care. Uh, did not go over well and ended up, he had to resign as CEO

[00:26:05] Cameron: and, uh, under pressure from shareholders

[00:26:09] Cameron: and, uh, the whole thing

[00:26:11] Steve: And now he’s a statesman.

[00:26:12] Cameron: disaster.

[00:26:13] Cameron: Yes, then he

[00:26:15] Cameron: underwent massive PR sensitivity training, and now he’s the elder statesman.

[00:26:20] Steve: That’s right. Well, the bar’s low these days, isn’t it?

[00:26:24] Cameron: Well, moving along, um, interesting story that I think

[00:26:28] Cameron: is a sign of things to come, and it

[00:26:30] Cameron: sounded so clickbaity when I first read this that I didn’t believe it, but I’ve,

[00:26:34] Steve: I didn’t click on it because of that. I thought I’ll let Cam explain it.

[00:26:37] Cameron: I’ve, I’ve looked into it, it seems legit, um, a boy saw 17 doctors over three years for chronic pain, ChatGPT found the diagnosis. So, quick story is that a few years ago, an American mother, two young kids, she had a four year old boy who began experiencing pain.

[00:26:56] Cameron: I think it was on a trampoline. When he first experienced it, [00:27:00] really very bad chronic pain. Over the next few years, it took him to see 17 different doctors and dentists and all sorts of people to try and work out what was going on. Lots of MRIs, lots of x rays, scans. No one could figure out what it was.

[00:27:17] Cameron: Eventually, she got onto ChatGPT and explained all of his symptoms and the MRIs and that kind of stuff. And it suggested a diagnosis of something called Tethered Cord Syndrome, which is when your spinal cord is stuck on your spine. And she took him eventually to another doctor with the MRIs and with this suggestion that it might

[00:27:40] Cameron: be… Tethered Cord Syndrome, and this new doctor went, yeah,

[00:27:44] Cameron: absolutely, that’s what it is. Had surgery. He’s still recovering from the surgery,

[00:27:49] Cameron: but, um, all good Apparently it’s a good sign.

[00:27:54] Steve: a good friend of mine, his brother had like

[00:27:57] Steve: a nerve, like a, a, a, a, a, a, a, a, I can’t remember exactly what it’s called, but [00:28:00] something about like you get this nerve damage and you just have pain all through

[00:28:02] Steve: your body. And he spent three years like doctor hunting to find out. He ended up going to a clinic, one of two in the world that happens to be in the U.

[00:28:10] Steve: S. for six months and mortgaged his house and did all of that. But it does feel like the AI doctor The idea of that is one of the really strong use cases because the breadth of possibility of symptoms and understanding them, no person could just know all of that. It’s one of the great use cases, I think, and I think medicine and AI is where the really strong use cases is one of those areas.

[00:28:34] Cameron: Yeah. You know, I mean, we already use it for some medical knowledge. And again, keeping in mind that it hallucinates and nothing’s perfect, but neither are humans, but it can push you down. It can give you some ideas of things to investigate or, and the way that all of these AIs are coded now, if you ask it any medical question, it’s going to say, listen, I’m not a doctor.

[00:28:58] Cameron: You should seek medical [00:29:00] advice. However, you might want to, it sounds like it might be this, or it might be that. Go get medical advice, proper medical advice. But it’s only a matter of time before in particular, all of us, and it’s not such a big deal in Australia where we have free or very low cost healthcare.

[00:29:20] Cameron: I remember when Chrissy first moved here from the U. S. 15 years ago, and she, she got swine flu on the plane on the way over. So she’d been here for a few days, got very, very sick, sicker than she was when she got COVID. Very, very sick. Couldn’t get out of bed for two weeks. And I was like, we’ve got to take you to a doctor.

[00:29:37] Cameron: She goes, no, no, I don’t need that. I can’t, can’t afford to go to a doctor. I was like, what are you talking about? She goes, I’m not even a citizen. I go, no, one’s going to ask you. No one cares. Like it took me a week to convince her that it was okay. She can go see a doctor. Cause she’s grew up in America where it’s a big deal.

[00:29:54] Cameron: You don’t go to a doctor in America unless you absolutely are about to die.[00:30:00]

[00:30:00] Cameron: And there are people obviously in Europe and all around the world in developing countries where access and in the outback and remote regions where access to. GPs isn’t as easy as it is in

[00:30:10] Steve: Well, it might be a big drive. Yeah.

[00:30:13] Cameron: so having an AI doctor or an AI

[00:30:16] Cameron: nurse like, we think of it as the equivalent to 1 300 health or 13 health, whatever it is, like free.

[00:30:23] Cameron: medical service, you ring up and go, Hey, listen, um, I’ll tell you another thing. Chrissy was playing me a podcast yesterday on her way to Kung Fu. I think it was This American Life she was listening to. And it’s about a service that exists in the US where if you’re going to shoot up heroin, you can call

[00:30:39] Cameron: this telephone number and somebody will sit on the line with you

[00:30:44] Cameron: while you’re shooting up.

[00:30:47] Cameron: They get your name and your phone number

[00:30:49] Cameron: and your address.

[00:30:51] Steve: So it’s sponsored by Purdue Pharma and what they do is, once you’re done with the SCAG, they go, Look, we’ve got OxyContin here, it’s easy, you can get it at the

[00:30:58] Steve: pharmacy, you’re not going to go to jail. [00:31:00] and we just sell it in convenient little pill packs. And that’s the Sacklers.

[00:31:06] Cameron: They will sit on the phone with you and monitor you. If you’re at home by yourself and you’re shooting up heroin, so if you have an overdose, if you stop responding to their questions on the call, they

[00:31:19] Cameron: will, get 9 1 1 to send emergency services to your apartment

[00:31:24] Cameron: straight away. Which I think is a fabulous thing.

[00:31:27] Cameron: They’re treating addiction as a

[00:31:29] Cameron: health problem

[00:31:30] Steve: Which I agree with. I have a fundamental issue with something, though, with, with healthcare and, and uh, shooting galleries and You know, places where you can do safe injecting. It’s just a quandary. It’s not a, I don’t have a problem with it because no one wants to see anyone die and everyone was a baby once that got loved and they got families.

[00:31:50] Steve: And I agree that it’s a health issue, 100%. But I just struggle with the idea that on the one hand, it’s like we get this big haul [00:32:00] of heroin that’s come from Afghanistan or wherever and, the police are chasing. It’s like this game of cops and robbers. Right, where at the supply side, they’re like, it’s full military, stop it, chase it, but then at the demand side, it’s like, oh, we get it.

[00:32:16] Steve: It’s kind of okay. Yeah, bring in your small little portion of heroin or whatever it is, because you’ve got a health problem. We’ve got a government sponsored injecting room. It just feels like… There’s this incongruence between those two things where it’s either a health problem and you fix it, so just make it a health thing right through.

[00:32:35] Steve: But then you’ve got this big army war over like the massive supply. Well, where do you think that big supply goes? It gets broken down into little pieces and given to the guy that you’re like, come in here, come in here, we’ll make sure you’re okay. The whole thing is just really incongruent. You know what I mean?

[00:32:46] Steve: Like, and this is not to be disparaging of Saving people from having overdoses. I get it. But on the one hand, the government is like, the war on drugs, and then the, oh, it’s okay, it’s, the two things are broken. You gotta, it [00:33:00] just troubles me. It just seems inconsistent and insane.

[00:33:04] Cameron: Six or seven years ago on my other show, The Bullshit Filter, we did like 25 episode arc on the history of the war on drugs. And it is now and always has been complete bullshit. And is the, it’s been a complete disaster. Trillions of dollars spent on it around the world. It’s

[00:33:24] Cameron: never worked. It doesn’t stop any, either side of

[00:33:28] Cameron: it, the supply side or the demand side. In fact, they just keep growing. It’s the complete, it’s the biggest load of bullshit

[00:33:34] Steve: Well, you can’t, you

[00:33:34] Steve: can’t stop it. It goes, I mean, you go way

[00:33:37] Cameron: So you just got to legalize the whole

[00:33:39] Cameron: thing. If you want to put the, if you want to put

[00:33:41] Cameron: the criminal gangs out of business, you legalize it, you regulate it,

[00:33:46] Cameron: Yeah. And it becomes like the booze industry, right?

[00:33:50] Steve: Yeah, maybe not as… maybe not as, maybe not as promoted and freely available and all. I don’t, I don’t know. I mean, what, what do you have a cocaine bar where [00:34:00] you go and get different colors of cocaine and, and pink cocaine and and, cocaine with lemon dust in it. So it, you know, tastes not, gee, I don’t know,

[00:34:08] Cameron: Why not? Anyway, moving right along. Oh, my point with all of that story was going to be how long before you have an AI on the other end of the call rather than a human that can monitor your drug taking efforts and it’ll say you keep talking to me. I’m going to ask you a question every 30 seconds, and if you stop responding to me, or I’m checking your heart rate monitor on your Apple Watch, um,

[00:34:34] Cameron: you know, if something happens,

[00:34:35] Cameron: I’m going to alert 911 or triple zero, whatever, and have somebody at your house

[00:34:40] Steve: feels like a won’t be far away as an AI doctor, as your first GP call as well, which I’ll be fine with. As long as you know what you’re doing and what you’re dealing with, that it’s an AI and that it’s transparent.

[00:34:52] Cameron: I’ve used it a lot over the last month or so with body aches and pains that I get from Kung Fu. Like, I’ve got [00:35:00] knee pain if I bend too much, my knees hurt. So I’m asking ChatGPT, okay, I’m getting this pain in this specific area when I do these specific things, what might it be? And it points to, in my case, it said, well, usually knee pain in this area, I can’t forget, I forget the name of the muscles, ileosis or something, is usually because your glutes aren’t strong enough.

[00:35:24] Cameron: So your glutes aren’t taking enough of the weight. And so your knees are having to take the weight. So you really need to be working on your glutes. And gave me, I said, got some exercises for that? Yep. So it gave me a bunch of glutes. exercises to do to strengthen my glutes that I do every day. So, it is, it’s the place I go to now is the first call.

[00:35:44] Cameron: And then I can go and look it up. I’ve got books on stretching and I’ve got books on this and that and the other, but it tells me where to look for this stuff. It gives me hints of what might be wrong.

[00:35:54] Cameron: Okay. Moving right along. Another story I saw this week that I

[00:35:57] Cameron: thought was good. This is in science.

[00:35:59] Cameron: org.

[00:35:59] Cameron: If [00:36:00] AI becomes conscious, how will we

[00:36:02] Cameron: know?

[00:36:04] Steve: Well, I don’t think you will. Uh, I’ve heard some of the AI alarmists, is it Yudowsky or Yadowski? He said that the first thing a sentient, smart AI would do is not tell you and then go about whatever it deems to be its own personal objectives in disguise. And if it’s smart enough to be sentient and it’s got a PhD in every single subject, it should be able to keep it from you that it’s sentient as well and do that work in the background.

[00:36:33] Steve: So I don’t, I don’t think you would know. Well, but then there’s the other complexity of this is

[00:36:41] Steve: we can’t agree on what consciousness is yet. So it’s very hard to define something, if something has something, when we don’t all agree on what that thing happens to be.

[00:36:52] Cameron: Yeah. Well that’s what this article is really drilling into, uh, this is by Elizabeth Finkel, [00:37:00] this is in Science Magazine, August 2023. You know, they basically say that a group of 19 computer scientists, neuroscientists, and philosophers came together, wrote a 120 page discussion paper, basically looking at human consciousness and different theories about it, applying how that works to AI architectures, and trying to figure out different ways that we may know Assuming it doesn’t hide it from us, as you suggested, that, uh, it, we have, uh, uh, AI that has achieved consciousness.

[00:37:38] Cameron: I like this bit, I thought this was a good bit. Six theories made the grade. One was recurrent processing theory. Which proposes that passing information through feedback loops is key to consciousness. Another, the Global Neuronal Workspace Theory contends that consciousness arises when independent streams of information pass through a bottleneck [00:38:00] to combine in a workspace analogous to a computer clipboard.

[00:38:06] Cameron: Higher order theories suggest consciousness involves a process of representing and annotating basic inputs received from the senses. Other theories emphasize the importance of mechanisms for controlling attention and the need for a body that gets feedback from the outside world. From the six included theories, the team extracted 14 indicators of a conscious state.

[00:38:30] Cameron: So, I guess the point is, they’re, this is, they’re asking questions, and in part, getting away from the debate is, is AI conscious, is it not, they’re actually trying to come up with a framework for thinking deeply about this, understanding that we don’t really understand human consciousness or consciousness in other life forms yet, but trying to figure out what are some of the indicators, how might it work, how would we know, I mean, I think one of the things that’s obvious, if you assume that, [00:39:00] The AIs that we have are not conscious,

[00:39:04] Cameron: but then we do believe that they can pass the Turing test, as we talked about in our last episode.

[00:39:09] Cameron: In fact,

[00:39:10] Cameron: the problem with the Turing test is they’re so good. Um,

[00:39:14] Steve: But love.

[00:39:16] Cameron: they’re too, too good. That’s how they fail the Turing test is they’re too smart

[00:39:21] Steve: Yeah, you have to dumb them down a bit to

[00:39:22] Steve: pass it. That’s so interesting that

[00:39:24] Steve: just, again, just leaped it. But you do need a framework, I think. And that, and that’s what you need. Because if you have the framework, then you can say, does, does

[00:39:32] Steve: this apply? Because at the moment, it’s all an adhocracy of opinions.

[00:39:36] Steve: If you don’t have an agreed framework, then you don’t know what good looks, usually in business, what does good look like? Or what is success? In this case, it’s what is consciousness? What’s the, what’s the cage that we need to see if it fits those parts?

[00:39:50] Cameron: Yeah.

[00:39:50] Cameron: it’s more of an intelligent approach to asking the right questions and thinking, which it’s good to see. I mean, stepping back a bit, like it’s kind [00:40:00] of mind blowing that a year ago, this was

[00:40:05] Cameron: so far removed from everyday conversation

[00:40:08] Cameron: that it’s not funny, and

[00:40:09] Cameron: here we are going, shit, we really need to figure this out really quickly

[00:40:12] Cameron: because we’re on the doorstep of it.

[00:40:15] Steve: That is a really funny thing that nothing changes and then everything

[00:40:20] Steve: changes really

[00:40:20] Steve: quickly.

[00:40:22] Cameron: Mm

[00:40:22] Steve: You wonder whether it’s always there and it is changing at the same rate or whether a discovery creates the change. I actually don’t know the answer to that. Is it always bubbling away but everyone’s attention gets drawn to it?

[00:40:34] Steve: People were just working on it just in darkness or is it that no, once it’s discovered it’s there, so no, no, no. And now we’ve. You know, we’ve, we’ve discovered electricity, we’ve, we’ve mastered fire, and here it is, so we’d better learn to deal with it real quick. I don’t know which one of those it is.

[00:40:50] Cameron: There are definitely milestones that happen, like I think GPT 3. 5 was a milestone, uh, release that suddenly [00:41:00] brought everyone’s attention to

[00:41:01] Cameron: something that, researchers or scientists have been working on since the Transformer paper was written in 2017,

[00:41:06] Cameron: but this article in Science points out that in 2021, a Google engineer, Blake Lemoine, got himself fired,

[00:41:14] Cameron: yeah, me too, when he claimed that Lambda, the chatbot that they were working on, was sentient,

[00:41:19] Cameron: He got fired from Google for saying that out

[00:41:22] Cameron: loud.

[00:41:23] Cameron: Within 18 months,

[00:41:25] Steve: was derided, he was largely derided as well.

[00:41:28] Cameron: Yeah, With 18 months, everyone…

[00:41:30] Cameron: was

[00:41:30] Cameron: asking the question about whether or not ChatGTP

[00:41:34] Steve: the vindication doesn’t help him with his employment though, does it? I mean, there he was in the doghouse for 18 months before that

[00:41:40] Steve: happened.

[00:41:41] Cameron: Yeah, I’m sure he’s doing okay. But, you know, that’s how quickly it changed. since then, Geoffrey Hinton who I think was his boss, resigned from Google and has come out saying he thinks it’s showing signs of sentience.

[00:41:56] Cameron: So, I don’t know how he felt back at the time when [00:42:00] Blake Lemoine got fired, but, uh, I love it. I’m on his Twitter feed actually, and he posted, um, he posted a gif from the old 18 TV show with Hannibal saying, I love it when a plan comes together. So yeah, he was just, he was ahead of, ahead of the rest of us at the time.

[00:42:22] Cameron: Next story, Steve, this was in Wall Street Journal this week. MBA students versus ChatGPT. Who comes up with more innovative ideas?

[00:42:31] Steve: I haven’t read the article. Can I guess? Can I guess?

[00:42:34] Cameron: you can

[00:42:35] Steve: Here’s what I think. I think that ChatGPT.

[00:42:38] Steve: will come up with way more ideas, like an infinite number.

[00:42:43] Steve: That said, it depends on how ChatGPT was prompted, so you need a human behind that, and the better the prompting, the more the ideas and the breadth, or you can go down tighter or change them, but I think that ChatGPT would slay it because Computational capacity is always[00:43:00] bigger, can always do more, but I think that the MBA students would come up with things that are more nuanced and interesting based on their personal life experiences.

[00:43:08] Steve: That’s my guess.

[00:43:09] Cameron: Close, uh, but no cigar.

[00:43:11] Steve: Ah, damn.

[00:43:14] Cameron: This is, this is run, was run by, um, these two guys who are professors of operations, information and decisions at the Wharton School of the University of Pennsylvania, uh, where one of them runs the Mac Institute for Innovation Management. So the basic setup here is these guys teach innovation and entrepreneurship and product design at this Wharton School.

[00:43:39] Cameron: We, uh… They had 200… No, hold on a second. In this competition, which we ran together with our colleagues Lennart Meike and Karan Girotra, humanity was represented by a pool of 200 randomly selected ideas from our Wharton students. [00:44:00] The machines were represented by ChatGPT4, which we instructed to generate 100 ideas with otherwise identical instructions as

[00:44:08] Cameron: given to the students.

[00:44:09] Cameron: Generate an idea for a new product or service

[00:44:13] Cameron: appealing to college students that could be made available for 50 or less.

[00:44:19] Steve: Okay, so there was a, there was a clear…

[00:44:22] Steve: Instruction, they’ve both got the same instruction, so that’s

[00:44:24] Steve: good.

[00:44:26] Cameron: In addition to this vanilla prompt, we also asked ChatGPT for another 100 ideas after providing a handful of examples of successful ideas from past courses that they’d run with their students. In other words, a trained GPT group providing us with a total sample of 400 ideas. Uh, basically, not surprisingly, ChatGPT easily outperforms us humans.

[00:44:53] Cameron: And the number of ideas per unit of time. Generating 200 ideas the old fashioned way [00:45:00] requires days of human work, while ChatGPT can spit out 200 ideas with about an hour of supervision.

[00:45:07] Steve: Yeah, there you go.

[00:45:08] Cameron: Next, to assess the quality of the ideas, we market tested them. Specifically, we took each of the 400 ideas and put them in front of a survey panel of customers in the target market via an online purchase intent survey.

[00:45:21] Cameron: The question we asked was how likely would you be to purchase based on this concept if it were available to you? The possible responses ranged from definitely wouldn’t purchase to definitely would purchase. Basically, the bottom line is this. The average purchase probability of a human generated idea was 40%.

[00:45:43] Cameron: That of the vanilla GPT 4 prompt was 47%. And that of the GPT 4 seeded with good ideas was 49%.

[00:45:55] Steve: Close. It’s close.

[00:45:57] Cameron: isn’t only faster, but also, [00:46:00] on average, better.

[00:46:02] Cameron: at Idea Generation. Now, I, I, I discovered

[00:46:06] Cameron: this on Reddit, and a lot of

[00:46:07] Cameron: people on Reddit are saying, well, that’s not surprising because people doing an MBA aren’t trained to be

[00:46:13] Steve: point, the sample, like the sample has an

[00:46:15] Steve: impact. I would argue

[00:46:17] Steve: that a true innovation, like, you’re gonna get corporatish kind of ideas and innovation from corporatish kind of people in MBA, and ChatGPT has had more of that corporate training than a person could because… It’s seen all of that MBA ish type output for, on that kind of thing.

[00:46:37] Steve: But, a true innovator, someone who comes up with something quirky and unique, who wouldn’t go to an NBA and has been experimenting with cool and interesting ideas, isn’t that type of person that would be in that sample set. I think that’s a really good point from the audience.

[00:46:51] Cameron: Right, but, I mean, who else do you go to? Like, they did say you should have got a 200 entrepreneurs. You know, that might be [00:47:00] interesting. But, okay, if you’re a corporation, if you’re a business, let’s say you’re a small

[00:47:04] Cameron: business,

[00:47:05] Cameron: Medium sized business. You can’t go out and hire 200 entrepreneurs

[00:47:09] Cameron: to come and

[00:47:10] Steve: No, you can’t

[00:47:11] Cameron: new ideas for you.

[00:47:12] Cameron: You probably can’t even afford to go and hire 200 MBAs. You might be able to

[00:47:15] Steve: can’t afford to hire anyone

[00:47:16] Steve: if you’re a small business. you.

[00:47:17] Steve: come up with

[00:47:18] Steve: it yourself or you go to the internet.

[00:47:20] Cameron: Yep. So you can go to GPT 4, get it to come up with hundreds of good ideas, and then… Run them through this market testing scenario. There’s probably relatively

[00:47:32] Cameron: cheap and easy ways to do that online these days.

[00:47:36] Cameron: And figure out which business

[00:47:37] Cameron: ideas you should implement. Um, and it’s something that I’m actually going to do

[00:47:41] Cameron: a live, I think you and I should

[00:47:42] Cameron: do this. This

[00:47:43] Steve: this. is, This is a nice idea. Why don’t we, before the next, why don’t we do something before the next thing? Why don’t we

[00:47:51] Steve: compete against ChatGPT? Like you and I have a contest with it and we’ll frame up what that contest is in the next [00:48:00] episode. We’re going to compete against ChatGPT to do an idea or something, right?

[00:48:06] Steve: And then we’ll see what we come up with versus that.

[00:48:09] Cameron: Or, You and I just use it to come up with 200 ideas,

[00:48:13] Cameron: market test those ideas, Then build and launch some products based on those ideas.

[00:48:19] Steve: I haven’t, I think I mentioned it to you. I want to do

[00:48:22] Steve: a fully AI generated beer brand where the entire, the

[00:48:28] Steve: entire thing, I just start and I was going to use an agent GPT, uh, come up with a beer brand. And everything from the advertising, the distribution, the sales, the pricing, the packaging design, uh, the copywriting, like the entire thing.

[00:48:44] Steve: The world’s first entirely AI driven brand where no human has had any input in anything other than prompting AIs. You know what we should call it? And you know what we should call it?

[00:48:57] Cameron: testing.

[00:48:58] Steve: We should call it Prompt Lager. [00:49:00]

[00:49:02] Steve: Do you like that? I like that. I’m writing that shit down. Prompt Lager it is,

[00:49:06] Steve: kids.

[00:49:08] Cameron: lager? Spelled L a

[00:49:09] Cameron: I g e r with a big a

[00:49:12] Cameron: i liga

[00:49:14] Cameron: la la

[00:49:15] Steve: Well, I like that. Liger.

[00:49:17] Cameron: lagar. It’s not lagar,

[00:49:20] Steve: the formula, where to produce it, how to produce it, everything.

[00:49:25] Cameron: Yeah. but that’s not coming up with a new idea. That’s just coming up with beer. Like, coming up with

[00:49:31] Cameron: new

[00:49:31] Steve: but I, okay, that’s different to a new idea. That’s a different idea. This is like, get it to

[00:49:36] Steve: make the thing. But if you got it to come up with a new idea. which was really cool, maybe, maybe,

[00:49:42] Steve: you know, in a test, well, maybe you find a way to make some money because, again, you get that first to market probability. Hmm,

[00:49:51] Cameron: Alright, well let’s talk about that off air. Moving right along, IBM’s CEO, not the woman I spoke of earlier, but a guy by the name of [00:50:00] Arvind Krishna. Spro’s hiring for thousands of back office jobs has predicted that AI would take up to 50 percent of new jobs, and he’s just put four and a half billion dollars,

[00:50:17] Cameron: uh, no, polled into, no, sorry, he’s put two hundred and, well, hold on, polled into a four and a half billion dollar tech unicorn’s massive new 235 million funding round.

[00:50:26] Cameron: Okay, let me get that right. So this is for Hugging Face. For people who haven’t heard of Hugging Face, basically a website where a lot of AI models that you can download and run locally, uh, hosted with instructions and a community about how to run them and fine tune them. Valued at four and a half billion dollars, didn’t exist a year ago, um, he’s just pulled into a two hundred and thirty five, IBM has, two hundred and thirty five million dollar funding round.

[00:50:58] Cameron: But the interesting thing for me [00:51:00] with this story is the CEO of

[00:51:02] Cameron: IBM is predicting that AI would take up to 50 percent of

[00:51:07] Cameron: new jobs. What do you think about that, Steve?

[00:51:13] Steve: yes and no. Will AI take jobs? Yeah, it will. It takes tasks first. The problem is I just don’t see how it

[00:51:22] Steve: takes a job. I just think there’s less. It doesn’t take, okay, it doesn’t take a job away in totality. I mean, one of the things that we’ve talked about a lot is the need to work with the AIs to get the output that you need.

[00:51:36] Steve: That’s, that’s the thing. And so, the previous example we had where it will do better than 200 students, MBA students, but someone still needs to work with the AI to pull the stuff out of the AI. So, I don’t think it actually takes away the job, it just takes away the number of them that you have, and it increases the breadth of what you do. What always happens, and I say this again and again and again, is that [00:52:00] When an efficiency happens, it frees up money, the money gets spent elsewhere and more people arrive in those places. But I haven’t seen anything where people go, look, we’re not, we haven’t got this person now because AI is doing it. I can’t really see any of that just yet.

[00:52:16] Steve: It’ll probably come. And I think there’ll be where you might’ve had seven people doing some sort of administrative task within a company. Yeah, and then maybe it goes down to two or five, but I don’t know how quick that happens. You can do a hiring freeze. And it may well be that they wanted to do that anyway, and then they use AI as a guise, as a well, it’s AI’s fault, so don’t blame us, and it’s efficiency and whatever, and it’s a, it’s a nice soft way to go to market and announce removing people, uh, from an organization.

[00:52:49] Cameron: I have seen stories of businesses saying that they’re letting people go and replacing them with AI. And I do think it’ll initially, I mean, first of all, [00:53:00] let’s keep in mind that this is very early days, you know, GPT, very, very new, relatively speaking, still a little bit buggy. And so yes, when I wouldn’t expect to see major transitions right now, but at some point in the

[00:53:15] Cameron: not too distant future, and I’m talking the

[00:53:17] Cameron: next year, I expect to see a lot of businesses Either not

[00:53:22] Cameron: hiring a human because they can get their existing humans

[00:53:26] Steve: to do more.

[00:53:27] Cameron: the work.

[00:53:28] Steve: Yep. I agree with that.

[00:53:30] Cameron: much more productive using the AI or downsizing. As you said, they used to have 10 people. They can now have five people, three people do the work of 10 people, particularly data entry, data analysis, data manipulation, coding, anything like that, which is smack bang in the middle of the skill set of an AI.

[00:53:52] Cameron: Writing, editing, uh, image, you know, somebody, oh, Guy at Kung Fu told me that [00:54:00] Guy he knows has a daughter who’s doing something at university, oh, graphic design, I think,

[00:54:09] Cameron: and she was told by her professor recently, uh, one of her professors,

[00:54:14] Cameron: you’re doing great, but you need to use more AI,

[00:54:18] Steve: That’s interesting. That’s a, that’s a really interesting phraseology and, and I think in, in a lot of the

[00:54:25] Steve: video voice content creation, I think you can already see an impact on a lot of those things now where there’s a lot of automated services, you know, to cut up video and give you edits and translations and Subtitles, animations, all of those things are coming up pretty strong and fast.

[00:54:45] Steve: You can see that stuff. You can see, uh, backend data, accounts payable, invoice, all of that kind of stuff. You can see it coming in really quick. There’ll still be someone overseeing. It’ll just be less people, I think. It’s kind of a, it’s, [00:55:00] it’s a little bit akin to the, uh, the factory floor. there was someone putting in nuts and bolts in every doorframe and every window.

[00:55:09] Steve: And then now it’s a machine with a couple of guys kind of working the machine. It’s a lot like that, I think. And the transition takes a bit of time. Definitely, there’s going to be less people doing those things. And we become more efficient and we’ve got to find new things to do. That’s definitely true.

[00:55:22] Steve: But I haven’t seen that it’s like, Oh, AI’s here now and we just sacked AI’s doing it all. I haven’t seen that. It’s kind of almost like a preemptive strike from the CEO for, from IBM in this case.

[00:55:35] Cameron: Well, he’s saying 50 percent of future jobs

[00:55:39] Steve: yeah, exactly. Sure.

[00:55:40] Cameron: uh, which I think is, I mean, that sounds like a large number, but it’s possible. But look, I, I do think it’s very early days. I do

[00:55:47] Cameron: expect to see this happen. But the, the other thing that, that story raised for me is, Wow,

[00:55:53] Cameron: IBM, like why aren’t they leading AI?

[00:55:58] Cameron: Like,

[00:55:59] Steve: Well, they were the first [00:56:00] gangsters when they had Watson, when

[00:56:01] Steve: they did the first Jeopardy stuff, which was, I don’t know how that worked. It

[00:56:07] Steve: was before the large language models, wasn’t it? It was before the papers came out. So it was a different, yeah, it was a different methodology, but it was language based answering.

[00:56:18] Cameron: And back in, was it, 1999, they

[00:56:23] Steve: Deep Blue,

[00:56:24] Cameron: Deep Blue beat Garry Kasparov, and I, that had a big impact on me at the time. I remember thinking, holy shit, this is, this is… This is real. Yeah. Cause I’m a chess guy. Um, so you would have thought that IBM would have been one of the leaders in this. And it sounds like they are playing catch up.

[00:56:42] Cameron: They’re trying to do a Microsoft. They’re trying to buy. They still have Watson as sort of this infrastructure AI thing, but they, they missed being on the front end of the AI revolution, which is kind of fascinating. Anyway, [00:57:00] I’m sure they will buy their way into it at some level successfully or otherwise.

[00:57:06] Cameron: I read a story in Wired Magazine this week. Meet Xenobot, an eerie new kind of programmable organism. Researchers hope the living robots made up of masses of cells working in coordination can help unlock the mysteries of cellular communication. Now basically, they’re using tiny little nanobots, I guess we can call them, to move things around.

[00:57:32] Cameron: Under the watchful eye of a microscope, busy little blobs scoot around in a field of liquid, moving forward, turning around, sometimes spinning in circles. Drop cellular debris onto the plane. And the blobs will herd them into piles. Flick any blob onto its back and it’ll lie there like a flipped over turtle.

[00:57:52] Cameron: Their behavior is reminiscent of a microscopic flatworm in pursuit of its prey or even a tiny animal called a water bear, a creature [00:58:00] complex enough in its bodily makeup to manage sophisticated behaviors. The resemblance is an illusion. These blobs consist of only two things, skin cells and heart cells from frogs.

[00:58:12] Cameron: Writing today in the Proceedings of the National Academy of Sciences, researchers describe how they’ve engineered so called xenobots from the species of frogs, Xenopus lyvis, Whence their cells came, with the help of evolutionary algorithms. They hope that this new kind of organism, contracting cells and passive cells stuck together, and its eerily advanced behavior can help scientists unlock the mysteries of cellular communication.

[00:58:44] Cameron: So, uh, first interesting thing about this is it’s, another step in the direction of nanobots. Nanites, microscopic machines that can run around inside of our blood and cellular [00:59:00] system and potentially in the future repair damage, uh, modify things, whatever. But one of the other things I liked in this is they were referring to generative AIs, maybe not in this story.

[00:59:19] Cameron: Might have been a different story. Okay, there was a different story that I liked, referring to generative AIs as generative LLMs or Golems for

[00:59:30] Cameron: short. Do you know what a Golem is, Steve?

[00:59:34] Steve: I’ve heard that before. You have to remind me.

[00:59:38] Cameron: Well, I think it’s a character in Lord of the Rings.

[00:59:41] Steve: Oh, I’m, I’m, I’m a definitely a, a non Lord of the Rings guy.

[00:59:46] Cameron: Me too. Star Wars all the way. But, it actually, the term comes from Jewish mythology. According to Judicious Museum Berlin, a golem is a creature formed out of a lifeless substance such as [01:00:00] dust or earth that is brought to life by ritual incantations and sequences of Hebrew letters. The golem, brought into being by a human creator, becomes a helper, a companion, or a rescuer of an imperiled Jewish community.

[01:00:15] Cameron: In many golem

[01:00:15] Cameron: stories, the creature runs amok and the

[01:00:17] Cameron: golem itself becomes a threat to its

[01:00:19] Cameron: creator. So, it’s basically Frankenstein’s monster.

[01:00:23] Steve: Yes. Well, XenoBots, I wrote a

[01:00:27] Steve: piece about XenoBots, Cam, you’ll be impressed to know, in 20th of September 22 for the

[01:00:32] Steve: Eureka Report,

[01:00:34] Cameron: Yeah,

[01:00:34] Steve: meaning stranger, StrangerBots.

[01:00:37] Cameron: right. What was the basis of your

[01:00:39] Steve: Well, the base of my story was pretty much the stuff that you said, that it’s less about silicon and more about biomimicry, um, and I said they’re built by combining together different biological, sometimes human tissue harvested from frog embryos, uh, they’re like a little African clawed frog.

[01:00:55] Steve: Uh, Xeno meaning stranger, uh, for all intents and purposes, [01:01:00] they are living tissue. They’re about the size of a poppy seed and under a microscope you can see a very small design structure. They can walk, around, swim, sense their environment, most importantly perform functions which are predetermined by how they’re programmed.

[01:01:13] Steve: So we’re programming biological beings. And this really goes back to Kurzweil who we spoke about in the last episode talking about nanobots and that eventually we will Do things at the molecular level, which sort of circumvents and almost curve jump silicon. And I was just talking about, you know, the potential to locate and fix disease, fight cancers.

[01:01:34] Steve: They can even send them into the ocean where they’ll be able to find and aggregate bits of plastic and turn them into large plastic balls where a traditional boat or drone can come and pick them up. And the cool thing is they don’t have any pollution and they don’t need any energy. Because they eventually biodegrade, because they’re from bio, um, bio, uh, biological and organic materials.

[01:01:57] Steve: And they actually get the energy from the fat and the protein [01:02:00] naturally stored in their tissue. So it’s really, really interesting stuff.

[01:02:05] Cameron: Fascinating.

[01:02:06] Steve: It, it, actually, that, that’s the kind of stuff that I think… AI almost facilitates because each revolution facilitates the next revolution by uncovering clues to what the next curve jump is, which is the thing that I think a lot about with AI.

[01:02:21] Steve: It’s that it’s going to unlock a lot of medical, climate, energy based issues that our current paradigm won’t be able to solve.

[01:02:29] Cameron: The history of thinking around nanotech actually goes back quite a long way. The first book I read on it was by Eric K. Drexler, his 1986 book, Engines of Creation, The Coming Era of Nanotechnology.

[01:02:44] Cameron: Uh, which sort of blew my mind at the time, but he was building on Richard Feynman’s talk from 1959. There’s plenty of room at the bottom.

[01:02:59] Cameron: So he gave a [01:03:00] co a talk at Caltech in 1959 where he was talking

[01:03:03] Cameron: about the idea of building machines based out of individual atoms

[01:03:09] Steve: Yeah, I love that. I mean, that’s the type of thinking and

[01:03:12] Steve: that’s what biology does way better

[01:03:14] Steve: than dumb humans do.

[01:03:17] Cameron: yeah. So you’re right, the hope is that AI will help really speed up the progress on these sorts of things and, uh, this

[01:03:25] Cameron: story about these Xenobots just, uh, is another instance of

[01:03:31] Cameron: researchers, outside of talking about AI specifically, research that people are doing

[01:03:37] Steve: Yeah, this is what I wrote on that article. I close it off by saying it’s another example of how investors, because I always have tech in an investing angle when I write for the Eureka Report, it’s another example of how investors need to think beyond AI as a product output, but as an input for a modern day raw material.

[01:03:53] Cameron: Yeah. The, um, golem story. is also from Wired. This is a story called [01:04:00]Give Every AI a Soul or Else.

[01:04:03] Cameron: Says mavens in the field of artificial intelligence, including architects of notorious generative AI systems like ChatGPT, now publicly express shared dread of terrible outcomes that might be wrought by their own creations. Many now call for a moratorium or pause in AI development, allowing time for existing nations and institutions to innovate systems of control.

[01:04:24] Cameron: This is going back to July, this story. But it says, why the sudden wave of concern amid the toppling of many cliched assumptions? We’ve learned that so called Turing tests are irrelevant, providing no generative golems, into whether generative large language models, GLLMs or golems, exist. are actually sapient beings.

[01:04:45] Cameron: They will feign personhood convincingly long before there’s anything or anyone under the skull, in inverted commas. Uh, and this is, uh, getting back to the Jewish, uh, [01:05:00]origins of Gollum. According to the Talmud, Adam, the first man, was actually a Gollum

[01:05:07] Steve: Wow.

[01:05:08] Cameron: created when the Lord shaped dust into a shapeless husk and breathed life into it.

[01:05:18] Cameron: So talking about those Xenobots, we’re taking frog cells, which are alive, of course, but shaping them together and getting them to perform functions. So,

[01:05:29] Cameron: um, talking about LLMs as being Gollum as well. They’re being, you know, we’re

[01:05:34] Cameron: using silicon, pushing together silicon and electronic components and creating.

[01:05:40] Steve: a way, in a way, what we do

[01:05:42] Steve: intellectually is we’re always trying to create new forms of movement or

[01:05:47] Steve: life, whether it’s industrialization and

[01:05:49] Steve: machinery or cogs

[01:05:51] Steve: or pulleys or shovels and which replicate claws and now it’s brains. And, you know, I wonder if all biological beings [01:06:00] just have this proclivity to try and create more of themselves in different ways.

[01:06:04] Steve: And it’s just part of the natural evolutionary process, I think.

[01:06:09] Cameron: Yes, I think so.

[01:06:12] Cameron: Well, my last story for this week, Steve, U. S. military.

[01:06:16] Steve: You’re old mates.

[01:06:18] Cameron: Yeah, my old mates of the U. S. military. Always up to, always, always trying to make the world a better place, U. S.

[01:06:23] Steve: Yes, just like Silicon Valley.

[01:06:26] Cameron: This is a story from the conversation, end of August. U. S. military plans to unleash thousands of autonomous war robots over the next two years.

[01:06:36] Cameron: The United States military plans to start using thousands of autonomous weapons systems in the next two years in a bid to counter China’s growing power. U. S. Deputy Secretary of Defense Kathleen Hicks announced in a speech on Monday the so called Replicator Initiative. That doesn’t sound chilling at all.

[01:06:57] Cameron: Aims to work with defence and other [01:07:00] tech companies to

[01:07:01] Cameron: produce high volumes of affordable systems for all branches of the military.

[01:07:06] Cameron: What were the humanoid robots called in Blade Runner?

[01:07:11] Steve: Replicants.

[01:07:12] Cameron: Replicants.

[01:07:12] Steve: I watched that the other week.

[01:07:15] Cameron: Me too.

[01:07:15] Steve: never seen it before.

[01:07:17] Cameron: Fuck off.

[01:07:18] Steve: I’m telling you. I knew the story, but I’d never actually watched it.

[01:07:22] Cameron: How can you never have watched it. Where have you been?

[01:07:25] Steve: I just never got around to it. Anyway, it’s a bit slow moving, to be honest.

[01:07:28] Steve: It’s good. It’s good.

[01:07:29] Cameron: Oh, you kidding me? I loved it. Well, it? is slow. moving, but it was made, It’s made, you know, early 80s.

[01:07:35] Steve: It’s got time to let you digest and

[01:07:37] Steve: think.

[01:07:37] Steve: Which is, I mean, movies were a bit different

[01:07:39] Cameron: you watch, The Director’s Cut?

[01:07:41] Steve: Yes.

[01:07:42] Cameron: Yeah, well, Ridley added, like, 45

[01:07:45] Steve: Right. It was long. I’m like, this is the long ones. The director’s cut on Netflix, but I liked

[01:07:49] Steve: it. It was good. I mean, I knew the story before I watched it, like

[01:07:53] Steve: I knew The arc and everything, but anyway,

[01:07:56] Cameron: Yeah.

[01:07:58] Steve: okay,

[01:07:59] Cameron: so, yeah, [01:08:00] I mean, just, before we finish up on this, You know, story of autonomous war robots.

[01:08:09] Cameron: Like what could go wrong? Um, we, we are obviously living in the real era now of Rolling out autonomous war robots that, in a world where we’ve got generative AI,

[01:08:27] Cameron: where we’ve got videos that can be made very quickly and cheaply to make it look

[01:08:32] Cameron: like somebody famous is saying something they never

[01:08:34] Steve: yeah,

[01:08:35] Cameron: in a language they don’t even speak,

[01:08:38] Steve: it’s really, really dangerous because yeah, you’ve got, you’ve got the ability to roll out something dramatically can have a big impact. And you’ve got a world where there isn’t going to be a negotiation between

[01:08:50] Steve: Moscow and the ships across like? It’s, it all moves so fast now, it feels like this is pretty [01:09:00] dangerous.

[01:09:02] Cameron: and she’s saying they’re doing this in an effort to stop China’s growing power.

[01:09:10] Cameron: We’re building thousands

[01:09:12] Cameron: of robots

[01:09:15] Cameron: to.

[01:09:15] Cameron: stop China’s growing

[01:09:17] Cameron: power. Like, what are the,

[01:09:20] Cameron: what are the insinuations there?

[01:09:22] Steve: Well, I don’t know if, yeah, I mean, you, you mentioned it the other day,

[01:09:28] Steve: and the disputes on whether or not that’s

[01:09:31] Steve: Chinese territory or not, I mean, it’s all pretty clear, just. It seems like you have the buildup of the military capability, you’ve got deep fakes, you’ve got the decoupling of the trade that we’re doing in technical arenas with China.

[01:09:49] Steve: the best way to not have a war is to be doing business with someone.

[01:09:53] Cameron: Wow, that hasn’t really worked, has it? They’ve been doing business

[01:09:56] Cameron: with China for 30 years and now we’re gearing up for a war with them. [01:10:00] I mean, look,

[01:10:02] Steve: Well, we haven’t had a war with them since we’ve been doing business with them. See, the point I’m making is we’re doing less business with

[01:10:07] Steve: them. So then you’re more likely to have war.

[01:10:10] Cameron: Yeah, that’s true. I mean, on one hand, I know from studying Cold War history that the U. S. military industrial complex always needs a boogeyman

[01:10:22] Steve: Yeah, you do.

[01:10:22] Cameron: point at to

[01:10:23] Steve: got to have an enemy.

[01:10:26] Cameron: to justify the trillion dollars a year that they spend on the military. And for very simple

[01:10:33] Cameron: economic reasons, I mean, the, the Pentagon

[01:10:36] Cameron: is basically the greatest

[01:10:38] Cameron: and most sophisticated form of theft that the human race has ever witnessed.

[01:10:44] Steve: I love that. Please just

[01:10:47] Steve: give me that because I love it and I need to hear.

[01:10:51] Cameron: Well,

[01:10:51] Cameron: the way that the military industrial congressional complex works, which by the way,

[01:10:58] Cameron: the term military [01:11:00] industrial complex was coined by Eisenhower.

[01:11:03] Steve: Right. And he said, if we, I remember the speech,

[01:11:04] Steve: I’ve heard the speech. And he says, if we let this get out of control, we’re going to be in trouble. That, you know, summarized.

[01:11:11] Cameron: Yeah, it was his last speech as, or his last State of the

[01:11:14] Cameron: Union, I think, as President. And the original draft of the speech, he actually called it the Military Industrial Congressional Complex.

[01:11:24] Steve: The congressional word seems to have dropped out, doesn’t it?

[01:11:27] Cameron: his, his, um, inner circle convinced him to remove the word Congressional because there was an election coming up and they didn’t want to cause more tension with Congress than they already had.

[01:11:40] Cameron: So he was a Republican, obviously, President, former General. And he was trying to warn America about the military industrial complex, but the reason I said it’s the greatest and most sophisticated theft in history is basically this is the way it works. It’s a very, very sophisticated system for taking money out of the pockets of taxpayers.

[01:11:58] Cameron: Basically what happens is [01:12:00] the majority of taxes paid in the US are paid by individuals. The income tax is the majority of the tax paid in the United States. That tax goes to the US Treasury. The US Treasury then gives a trillion dollars of that to the Pentagon. The Pentagon then gives that to American businesses and

[01:12:20] Cameron: corporations who have military contracts with the Pentagon.

[01:12:25] Cameron: And it’s not just five businesses, you know, I think most of us think, well, it’s just the big

[01:12:31] Steve: It’s Halliburton and it’s Boeing and it’s whoever, yeah.

[01:12:33] Cameron: There have been some really good, uh, analysis done of this. There’s actually tens of thousands of American

[01:12:41] Cameron: businesses. That have contracts with the Pentagon.

[01:12:46] Cameron: Everything from, you think about the 800 military bases that

[01:12:50] Cameron: the U. S. operates around the

[01:12:51] Steve: contracts and it’s, it’s everything. It’s

[01:12:55] Cameron: pens,

[01:12:56] Steve: You reminded me of the movie War Dogs. You know the movie War [01:13:00]Dogs.

[01:13:00] Cameron: I don’t think I’ve seen it.

[01:13:01] Steve: Oh my god, it’s a great funny movie where these

[01:13:03] Steve: guys

[01:13:06] Steve: get military contracts because they discover that all of them have to be available where anyone can bid so long as you’re in

[01:13:11] Steve: American business and then they start, they find ways to get Bits of military that go onto the soldiers and boots and all that kind of stuff.

[01:13:20] Steve: And they get in a, in a world of pain cause they get mixed up with all the wrong dudes. It’s a funny, great movie.

[01:13:25] Cameron: Right, I’ll have to check it

[01:13:26] Steve: You’ll like it.

[01:13:27] Cameron: Well, that’s how, that’s how it works. Like, there are tens of thousands of businesses in the United States every year who rely on Pentagon contracts for a relatively large percentage of their revenue every year. And the great thing about a Pentagon contract, if you can get one, is you don’t need to compete with the market generally, you know, you’re not competing on service delivery, you’re not competing on price, you’re not competing on

[01:13:54] Cameron: market conditions, really that kind of stuff.

[01:13:56] Cameron: You’ve got the contract, you’ve got it for four years, five [01:14:00] years, whatever,

[01:14:00] Steve: what they want. And,

[01:14:02] Cameron: yeah, and it’s easy money. Once it’s locked in, it’s locked in, it doesn’t take a lot of, uh, renegotiating or delivery contracts, all that kind of stuff. So vast Numbers of American businesses and a big percentage of the American economy relies on the Pentagon just handing them money every year, taking it from the taxpayers and giving it to American businesses.

[01:14:27] Cameron: And it’s hard to justify the ever increasing numbers

[01:14:30] Cameron: of

[01:14:31] Cameron: money that gets spent, goes up by about a hundred billion dollars every other year.

[01:14:35] Steve: And it’s their number one expenditure. Is that right? And they’re

[01:14:38] Cameron: Oh yes, by a lot, it’s

[01:14:40] Cameron: like 30 percent of the US economy goes

[01:14:42] Steve: crazy. And it’s 10 times more than the closest. Military expenditure.

[01:14:47] Cameron: It’s bigger than the rest of the world, including China and Russia put together,

[01:14:51] Steve: I have seen some video footage of US versus China military and the US gets a much poorer [01:15:00] return on investment because of these things that you mentioned, because they know it’s the government and they give them the government price and they do all of that. So their, their, their ROI or their productivity on what they invest is pretty poor and getting

[01:15:11] Cameron: covered a lot of stories about it in the psychopath epidemic. But yeah,

[01:15:16] Cameron: it’s, it’s, um, it’s a complete rort of the American taxpayer. Money just gets taken from the taxpayer

[01:15:22] Cameron: and given to businesses. And some of it, you could argue, trickles back down into the economy. But a lot of

[01:15:28] Steve: You know why it’s called, you know why it’s called trickle down

[01:15:30] Steve: economics, because you only get a trickle. I mean, the clue was in the name.

[01:15:33] Steve: I’ve always said that, Cameron.

[01:15:36] Cameron: And trickle down economics has been debunked many, many times by economists. It’s, it’s voodoo economics. But, you know, and this is the way the US learnt this, by the way, bit of history, when they did the Marshall Plan after World War II. People today tend to think of the Marshall Plan where they sort of created about 13 billion dollars at the time over several years to [01:16:00] give to Europe to rebuild Europe after the war. Most of that money never left the United States. It went straight from, again, the Pentagon to U. S. businesses. They were basically lines of credit for Europeans to buy American Goods and Services with. And it was a, and it pulled the US out of, they were still, they’d gone straight from the Great Depression into World War II.

[01:16:29] Cameron: Economy boomed during World War II because of Keynesian, military Keynesianism. Keynes, the economist, convinced Roosevelt that if you get the government to spend unlimited money on building up for a wartime economy, it’ll be good for the economy in general. The US came out of World War II with the economy.

[01:16:48] Cameron: booming because of the investment in the war, and they didn’t want to turn that spigot off. But you know, Truman tried. Truman started to downsize the military. [01:17:00] All of the businesses in America cried bloody murder because they were used to, they’d been living on this money now for 10 years, all of a sudden all the money started to dry up from the Pentagon.

[01:17:10] Cameron: They’re like, whoa, what the fuck? Where’s our money,

[01:17:13] Cameron: son? So then that’s why the

[01:17:15] Cameron: Cold War was created, essentially, to justify ever increasing amounts of money being spent

[01:17:21] Steve: Yeah, because it was, it was kind of interesting where there wasn’t really, and I’m, I know you’re the expert in Cold War stuff, but the enemy wasn’t

[01:17:30] Steve: really Russia or anyone when, when that happened, was it? At the end of World War II, that kind of devolved, didn’t it? Into that. It became, it was a narrative, wasn’t it?

[01:17:40] Cameron: Russia, or the Soviet Union as it was of course at the time, was the, was America’s ally

[01:17:44] Steve: That’s right. Yeah. When I grew up as a kid, I didn’t know that. I thought they were the enemy that we had World War II. You know, it’s like,

[01:17:50] Steve: It’s, it’s funny how

[01:17:51] Steve: that got

[01:17:51] Steve: shifted, wasn’t It

[01:17:53] Cameron: It was, Stalin, uh, Stalin, Churchill and Roosevelt meeting at Yalta, the famous photo of the three of them saying they were allies. [01:18:00] When World War II ended, uh, the U. S. just decided they needed a new enemy and they decided it was going to be Soviets. And, you know, today people go, Oh, well, the Soviets invaded this country and the Soviets invaded their country.

[01:18:16] Cameron: Number one, usually, no, they didn’t. They were, there were domestic communist movements in those countries that the Soviets might have supported to varying degrees, but they were domestic revolutions that people in the country said, you know what, we’ve had enough of capitalism. We’ve had enough of monarchies.

[01:18:33] Cameron: Fuck all y’all. We want to have some control. Oh, you’ve got five? Is that what

[01:18:36] Cameron: you’re saying? I thought you were saying hold on for five. Okay. Yeah, no worries. So anyway, uh, long story short, the US is gonna kill us all. Deep dive,

[01:18:47] Cameron: Steve.

[01:18:48] Steve: ha ha ha ha ha.

[01:18:52] Cameron: You got a deep dive,

[01:18:53] Steve: Yeah, deep dive, quick. I mean, I just wanted to wonder whether or not big tech

[01:18:58] Steve: should be broken up. Been a lot of [01:19:00] talk now for about seven years. I did a calculation this morning.

[01:19:04] Steve: The top six big tech companies have a market cap now of 10. 5 trillion after a good year. The US economy is 25. 5 trillion, so it’s 40 ish percent.

[01:19:14] Steve: Now, it’s not 40 ish percent of the economy in terms of revenue, but it’s, it’s really significant on how big these guys are. I mean, I think it would help innovation. The market dominance that they have has been long, uh, ignored because of what they have used as the no consumer harm test, which largely comes out of the Chicago School of Economics.

[01:19:41] Steve: And what they used to use back in the day was, well, if prices are getting cheaper, it’s got to be good for consumers. It gave premium model tech companies the biggest get out of jail free card for a lot of, for a long time, because they went to this free model, where they stole their raw materials, which was [01:20:00] us, and I think their dominance has been pretty, pretty bad.

[01:20:05] Steve: Uh. You know, look at

[01:20:07] Cameron: Well, that didn’t work in the Microsoft

[01:20:08] Cameron: antitrust case because Microsoft was bundling up Internet Explorer for free and they still got done by the Department of

[01:20:14] Steve: true. Well, but they were selling it. They weren’t giving away

[01:20:18] Steve: Microsoft Office

[01:20:18] Steve: for free. They were just putting it in there. I mean, let’s just look at some of the numbers

[01:20:21] Steve: that are really interesting. Facebook has 3 billion members, which is more than any religion, country, uh, in history.

[01:20:30] Steve: Facebook, uh, has a non voted in, uh, CEO who can never be voted out, even if he has zero economic interest in the company or meta. Amazon has more Prime members than Christianity. Uh, they copy their suppliers and now charge them an advertising fee to appear in their business. And as soon as they do something well, they just make the Amazon version of that and put them out of business.

[01:20:51] Steve: Google has a 90 percent share of video and search. Search advertising is really just a tax. Because all you have to do is pay the tax to [01:21:00] Google. To then say, well, I need to appear on top versus my competitor because there’s no other way to go to market. It’s not like creative advertising where you can be more creative than your competitor and build a brand.

[01:21:11] Steve: Search is kind of a tax because there’s no real creativity in that. Apple has a 30 percent tax on anyone who wants to have an app, as does Google have the other app store. Microsoft question mark, NVIDIA question mark, but here we are with companies which are… It just seems… More powerful than any corporations that I can remember, other than, I guess, the, the oil industry.

[01:21:35] Steve: And I wonder if, if, you know, we discussed this last time, I don’t think they’re going to be usurped by AI, because AI is expensive. And they have their fingers deep in it as well, which would be usually the next revolution, which replaces the previous dominance. Which

[01:21:52] Cameron: Yeah, good points, Steve. They’re insanely powerful. And the whole idea of, I don’t know, [01:22:00] capitalism enabling innovation is kind of bullshit anyway, because capitalism, the way it usually plays out is the massive companies with deep pockets and deep connects with regulators and, you know, the ability to engage Expensive law firms for infinite number, uh, amounts of time to sue smaller businesses out of existence, whether or not they’ve done anything wrong, tie them up in IP trademark bullshit cases, run them out of town.

[01:22:30] Cameron: It’s bullshit. I mean, it’s bad for consumers. It’s bad for innovation. It’s bad for the survival of the species. If you believe, as I do, that the survival of species right now comes down to how quickly we can innovate, how quickly we can have AI running things, big tech monopolies are not favourable. I mean, it is good that Microsoft could give OpenAI 10 billion dollars, I guess, that they have that amount of money that they can invest, because the government [01:23:00] probably wouldn’t be investing in OpenAI.

[01:23:03] Steve: is ironic, given the amount of money?

[01:23:05] Steve: that gets spent on the military, and that was, basically, they provided all the raw materials that created big tech.

[01:23:12] Cameron: Yes Anyway, I think there’s a good argument for breaking it up, but you know if if the big tech companies spend their money on competing with each other to be first to market with the sentient AI and that sentient AI then eliminates all the big tech companies and

[01:23:33] Cameron: takes control of the planet and the human race,

[01:23:37] Cameron: then maybe they don’t have long to

[01:23:39] Cameron: survive as it is. Maybe, maybe the seeds of their own undoing have already been

[01:23:43] Cameron: planted in AI,

[01:23:45] Steve: Oh, well, I think the one thing that is interesting is that often we assume that capitalism equals competition, but usually, the fundamental flaw of capitalism is that it usually results in a spiral of a monopoly where more power begets more power, competition [01:24:00] erodes, and the communist kind of, oh, this is

[01:24:04] Steve: If we take it away, the big companies, then, you know, the American Dream is lost. It’s all about people pursuing their own wealth. But basically, pure capitalism and pure communism are really, really similar. Because they end up with big, powerful monoliths that control things.

[01:24:21] Cameron: completely wrong. Pure communism, in pure communism,

[01:24:26] Steve: Okay, communism, communism as, as it has, let me say, let me rephrase that.

[01:24:32] Steve: Pure communism, as we’ve

[01:24:33] Steve: never really had. But the communism, that we had ended up in powerful monoliths

[01:24:37] Steve: controlling

[01:24:37] Cameron: We never, we’ve never had any form of communism. Soviet Union wasn’t communist, it was socialist. China isn’t communist, it’s socialist. They were a pathway to get to communism. You have to go from Feudalism to Capitalism to Socialism to Communism. They were trying to go from basically from [01:25:00] Feudalism straight to Communism, but they were going through Socialism first.

[01:25:04] Cameron: In pure Capitalism, sorry, in pure Communism, there is no state.

[01:25:08] Steve: Okay,

[01:25:08] Cameron: It’s small, autonomous groups of people with the power of

[01:25:14] Cameron: production doing their own thing. We’ve never, we’ve never got there yet. AI will get us

[01:25:19] Steve: all right, so,

[01:25:20] Cameron: Star Trek economy.

[01:25:21] Steve: Star Trek economy.

[01:25:24] Steve: Capitalism usually ends up with really powerful beings unless they get busted up every now and again because the fundamental flaw is that it ends up with monopoly powers in a lot of, and oligopolies in industries.

[01:25:34] Cameron: Yeah, and it takes socialists to

[01:25:36] Cameron: break those.

[01:25:39] Cameron: Teddy Roosevelt was a bit of a socialist at heart, because he brought in

[01:25:42] Cameron: the first antitrust laws.

[01:25:43] Steve: Yes.

[01:25:44] Cameron: I know you’ve got a heart out, you’ve got to go. Thanks man, fun, as

[01:25:48] Steve: Real good. We didn’t even cover it all, which means we’re just going to have to chat.

[01:25:50] Steve: again, Cameron.

[01:25:52] Cameron: Yeah, alright buddy, a tyranny of riches we had

[01:25:55] Steve: It was too, it was tryannical

[01:25:56] Cameron: have a good week buddy, I’ll talk to you soon.

[01:25:58] Steve: Cheers. [01:26:00]