WEBVTT - The AI Con 0:00:05.200 --> 0:00:09.280 Ah, welcome back to it could happen here a podcast. 0:00:10.760 --> 0:00:15.159 It's a podcast. I'm Robert evans Uh and with me 0:00:15.280 --> 0:00:19.560 today is Garrison Davis and James Stout. 0:00:20.079 --> 0:00:24.840 Hello, a Canadian, a Britishman and a text and walk 0:00:24.880 --> 0:00:25.840 into a podcast. 0:00:26.040 --> 0:00:30.800 Yeah, welcome to a podcast. Only two of them can 0:00:30.880 --> 0:00:31.560 drink in a bar. 0:00:32.960 --> 0:00:34.360 That's not true in Canada. 0:00:34.400 --> 0:00:37.919 We can all drink in a bar now, Garrison, A 0:00:37.920 --> 0:00:40.520 moment ago, you were holding your hand above a lit 0:00:40.560 --> 0:00:43.440 candle in a way that reminded me of G. Gordon Liddy, 0:00:43.920 --> 0:00:47.640 the Nazi who masterminded the Watergate breaking, and in order 0:00:47.680 --> 0:00:49.920 to convince people that he was a hard man, would 0:00:49.960 --> 0:00:52.760 regularly burn the palm with his hand on a candle 0:00:52.760 --> 0:01:03.640 while staring at them and gender great g Gordon Letty, 0:01:03.800 --> 0:01:05.800 you don't know enough about We'll talk about g. Gordon Letty, 0:01:05.840 --> 0:01:11.640 but today we're talking about something else, problematic artificial intelligence, 0:01:12.640 --> 0:01:15.679 which is not a thing that exists anywhere. It is 0:01:15.760 --> 0:01:20.360 instead a terrible, terrible error going back to like the 0:01:20.400 --> 0:01:25.080 sixties in case of terminology, when we talk about all 0:01:25.080 --> 0:01:28.280 of the things that people are like, you know, flipping 0:01:28.319 --> 0:01:31.880 out as ais chat GPT and stable diffusion and fucking 0:01:32.840 --> 0:01:38.360 all these other sort of like different programs. They're not intelligences. 0:01:38.440 --> 0:01:41.320 They're you know, the chat GPT is like a large 0:01:41.400 --> 0:01:45.800 language model. They're all essentially like bots that you train 0:01:47.240 --> 0:01:50.880 to understand kind of like what the likeliest thing that, 0:01:51.120 --> 0:01:54.840 what the likeliest appropriate response is to like a given prompt. 0:01:56.120 --> 0:01:59.440 That's kind of like the broadest way to explain it. 0:01:59.440 --> 0:02:01.880 It's comp located and they're you know, very useful. But 0:02:01.880 --> 0:02:05.160 obviously if you've been paying attention to the world right now, 0:02:05.520 --> 0:02:08.120 there's just a whole bunch of bullshit about them, and 0:02:08.320 --> 0:02:12.000 I think to kind of make sense of why we're 0:02:12.120 --> 0:02:14.680 seeing some of the shit around AI that we're seeing. 0:02:14.720 --> 0:02:18.200 And for a little bit of specificity, there have been 0:02:18.280 --> 0:02:21.359 like this kind of endless series of articles around this 0:02:21.560 --> 0:02:24.280 open letter signed by a bunch of luminaries in the 0:02:24.280 --> 0:02:26.959 AI field talking about how, you know, there need to 0:02:27.000 --> 0:02:29.200 be laws put in place to stop it from ending 0:02:29.240 --> 0:02:32.880 the world. You know, you've seen articles about like oh 0:02:33.600 --> 0:02:36.480 x percentage of AI researchers think that it could it 0:02:36.520 --> 0:02:40.280 could destroy the planet, destroy the human race kind of. 0:02:40.320 --> 0:02:44.480 Most recently, the biggest article, the biggest like viral hype article, 0:02:44.680 --> 0:02:48.760 was that the Pentagon had supposedly been testing an AI 0:02:49.520 --> 0:02:53.959 like missile system that blew up its operator in a 0:02:54.040 --> 0:02:57.240 simulation because the operator was trying to stop it from 0:02:57.320 --> 0:03:00.600 from firing or whatever. It was bullshit, like what was that? 0:03:00.680 --> 0:03:03.240 What actually had? Like Vice ran with the article. It 0:03:03.320 --> 0:03:06.160 was very breakfas Advice would do this flipping out about 0:03:06.160 --> 0:03:09.639 how horrifying you know, our AI weapons future is and like, yeah, 0:03:09.680 --> 0:03:12.560 we shouldn't give AI the ability to like kill people, 0:03:12.560 --> 0:03:14.880 but that's not at all what happened. Basically, a bunch 0:03:14.919 --> 0:03:17.760 of army nerds or air force nerds were sitting around 0:03:17.760 --> 0:03:20.120 a table doing the D and D version of like 0:03:20.480 --> 0:03:23.240 military planning, where you say, what if we did this, 0:03:23.520 --> 0:03:26.560 what kinds of things could happen if we did this system? 0:03:26.880 --> 0:03:29.320 And another guy around the table said, oh, well, if 0:03:29.320 --> 0:03:32.200 we build the system this way, it might conceivably attack 0:03:32.280 --> 0:03:35.440 its operator, you know, in order to optimize for this 0:03:35.560 --> 0:03:40.120 kind of result, which is like not scary, Like it's 0:03:40.280 --> 0:03:42.960 it's just people talking through pot like a flow chart 0:03:43.000 --> 0:03:45.160 of possibilities around a fucking table. You don't need to 0:03:45.160 --> 0:03:47.600 worry about that. There's so many other things to worry about. 0:03:47.960 --> 0:03:50.640 New York City is blanketed in a layer of smogs 0:03:50.640 --> 0:03:52.640 so thick you could cut it with a butter knife. Like, 0:03:52.920 --> 0:03:57.920 don't don't flip out about AI weapons just yet, folks. 0:03:58.160 --> 0:04:00.360 But I wanted to kind of talk about why this 0:04:00.400 --> 0:04:02.160 shit is happening, And a lot of it comes down 0:04:02.200 --> 0:04:05.640 to the fact that when we're talking about the aspects 0:04:05.640 --> 0:04:08.120 of like the tech industry that have an impact on 0:04:08.640 --> 0:04:12.400 outside of the tech industry. Right, there's basically three jobs 0:04:12.400 --> 0:04:16.280 in big tech. One job is creating iterative improvements on 0:04:16.360 --> 0:04:18.560 existing products. These would be the teams of folks who 0:04:18.600 --> 0:04:21.560 are responsible for designing a new iPhone every year, right, 0:04:21.800 --> 0:04:24.680 every couple of years. Lenovo puts out a new series 0:04:24.720 --> 0:04:27.159 of think pads and idea pads every couple of years. 0:04:27.240 --> 0:04:30.520 You know, you get a new MacBook every couple of years. 0:04:30.640 --> 0:04:33.919 Razor puts out a new blade. This is you know, 0:04:33.960 --> 0:04:37.120 these are the folks who kind of move along technology 0:04:37.160 --> 0:04:42.080 at a relatively like steady pace for consumer devices. And 0:04:42.120 --> 0:04:45.760 then you have the people who are responsible for kind 0:04:45.800 --> 0:04:48.719 of what you might call the moonshot products. This is 0:04:48.760 --> 0:04:53.280 a mix of the next big thing and doomed failures, 0:04:53.600 --> 0:04:55.520 and it's often pretty hard to tell you know what's 0:04:55.560 --> 0:04:57.600 going to be what ahead of time. A very good 0:04:57.600 --> 0:05:00.360 example would be back in the nineties, Apple a bunch 0:05:00.360 --> 0:05:03.080 of resources into launching an early tablet computer called the 0:05:03.120 --> 0:05:06.279 Newton that was a fabulous disaster. And then in the 0:05:06.279 --> 0:05:08.440 mid oughts they put a bunch of resources into launching 0:05:08.480 --> 0:05:12.560 the iPad, which was a huge success. And when you 0:05:12.640 --> 0:05:15.800 kind of think about like the folks doing this, like 0:05:15.880 --> 0:05:20.279 working on the moonshot products. The most recent example would 0:05:20.320 --> 0:05:22.159 be whatever team at Apple, the team at Apple that 0:05:22.279 --> 0:05:25.479 was behind putting together these new Apple goggles, which I 0:05:25.520 --> 0:05:28.400 don't think are going to be a wildly successful product 0:05:28.480 --> 0:05:30.080 in the way that they need it to be, like 0:05:30.120 --> 0:05:33.240 a smartphone scale success. But this is an example of 0:05:33.279 --> 0:05:35.240 like a thing that didn't exist and a bunch of 0:05:35.240 --> 0:05:37.960 people had to invent new technologies or new ways to 0:05:37.960 --> 0:05:41.599 combine technologies in order to make it exist. The third 0:05:41.760 --> 0:05:45.640 kind of job that the tech industry has, broadly speaking, 0:05:46.120 --> 0:05:49.919 are conmen, right, And the state that we are in 0:05:49.920 --> 0:05:52.560 in the industry right now is that every major tech 0:05:52.560 --> 0:05:55.400 company is run by some form of con man. Right. 0:05:56.279 --> 0:06:00.320 Tim Cook is, you know, kind of the least cones 0:06:00.480 --> 0:06:03.960 of the conment among them. But like Mark Zuckerberg obviously 0:06:04.560 --> 0:06:07.680 is a fucking flim flam artist, you know, and you 0:06:07.680 --> 0:06:10.279 can see this with the huge amount of money, like 0:06:10.320 --> 0:06:12.840 it's something like eleven billion dollars at least that Facebook 0:06:12.839 --> 0:06:16.800 pumped into this bullshit metaverse scheme that like Apple barely 0:06:16.880 --> 0:06:21.800 even talked about during their event unveiling like a headset 0:06:21.839 --> 0:06:25.440 that has VR potential in it. I'm getting away from 0:06:25.440 --> 0:06:27.240 myself here. Kind of the point that I'm making is 0:06:27.240 --> 0:06:30.640 that you can often have very real products. There's actual 0:06:30.680 --> 0:06:36.640 technology going into the Apple glasses marketed by conman flim 0:06:36.640 --> 0:06:38.920 flam artist. This is not always like a bad thing, Right. 0:06:38.960 --> 0:06:41.599 Steve Jobs was a con man, and it worked out 0:06:41.640 --> 0:06:44.760 pretty well for him because it just so happened that 0:06:44.800 --> 0:06:46.960 the tech not He had a decent enough idea of 0:06:47.000 --> 0:06:49.880 what the tech was capable of that it was able 0:06:49.920 --> 0:06:53.159 to kind of meet the promises he was making in 0:06:53.200 --> 0:06:55.920 more or less real time. An example of what happens, 0:06:56.000 --> 0:06:58.400 you know, pretty spectacularly when that's not the case is 0:06:58.440 --> 0:07:01.159 what we saw with Theronose and Elizabeth Holmes who started 0:07:01.160 --> 0:07:04.600 prison last week. Right, You've got these promises being made 0:07:04.600 --> 0:07:06.760 by the con man and the people who are responsible 0:07:06.760 --> 0:07:10.760 for the moonshots can't make it work. I'm bringing this 0:07:10.840 --> 0:07:13.560 up right now because there's a lot of folks, I 0:07:13.640 --> 0:07:17.720 think who believe that the the nate like the actual 0:07:17.760 --> 0:07:21.200 potential of AI has been proven in a spectacular way 0:07:21.680 --> 0:07:24.600 because the tools that have been released are able to 0:07:24.640 --> 0:07:28.360 do cool things, and I think those people are missing 0:07:28.400 --> 0:07:32.400 some key aspect, like some key things that like might 0:07:32.600 --> 0:07:36.160 cause one to think more critically about the actual potential 0:07:36.200 --> 0:07:38.160 the industry has and also might cause one to think 0:07:38.160 --> 0:07:42.640 more critically about how earth shattering it's all going to be. 0:07:42.680 --> 0:07:45.240 It's being taken kind of as red right now by 0:07:45.280 --> 0:07:50.080 a lot of particularly journalists and media analysts outside of 0:07:50.400 --> 0:07:53.080 the tech and or like outside of you know, the 0:07:53.160 --> 0:07:56.200 dogged tech press, that like, well, this is going to 0:07:56.280 --> 0:07:58.880 upend huge numbers of industries and put massive numbers of 0:07:58.880 --> 0:08:01.560 people out of work. And you know, that may seem 0:08:01.720 --> 0:08:03.760 if you sat down in front of this chatbot and 0:08:03.840 --> 0:08:06.960 had like a mind blowing experience, that may seem credible. 0:08:07.960 --> 0:08:11.280 There's not the evidence behind that yet. If you actually 0:08:11.360 --> 0:08:13.920 look at the numbers behind some of these different companies 0:08:13.960 --> 0:08:17.480 and like how their usership has grown and how it's 0:08:17.480 --> 0:08:19.600 fallen off, one of the things you've seen is that 0:08:19.680 --> 0:08:22.760 a lot of these tools had this kind of massive 0:08:23.760 --> 0:08:26.120 surge peak in terms of the number of people adopting 0:08:26.160 --> 0:08:28.080 them and in terms of their profitability. You saw this 0:08:28.160 --> 0:08:31.080 with like Stable Diffusion, right, and then this kind of 0:08:32.080 --> 0:08:36.720 fairly rapid fall afterwards, not because people are like giving 0:08:36.720 --> 0:08:39.679 it up forever or whatever, but because, like, once you 0:08:39.880 --> 0:08:42.840 fucked around with it and generated some images or generated 0:08:42.840 --> 0:08:45.679 some stories, there's not a huge amount to do unless 0:08:45.760 --> 0:08:48.280 you're someone who's specifically going to be using this for 0:08:48.360 --> 0:08:50.600 your job. And most of the people that wanted to 0:08:50.600 --> 0:08:53.520 fuck around with a lot of these apps didn't have 0:08:53.679 --> 0:08:57.040 long term use cases for them. This is why while 0:08:57.080 --> 0:09:00.680 you've got like, for example, Stability, which is the company 0:09:00.800 --> 0:09:04.400 or at least the main company behind Stable Diffusion, has 0:09:04.440 --> 0:09:07.480 been valued at like four billion dollars I think last 0:09:07.520 --> 0:09:10.720 it was checked, but their annualized revenue is only about 0:09:10.760 --> 0:09:15.840 ten million dollars. So that's a pretty significant gap. And 0:09:15.880 --> 0:09:19.480 it's a pretty significant gap because the actual money in 0:09:19.559 --> 0:09:23.200 AI so far isn't with the service providers really like 0:09:23.520 --> 0:09:25.040 you've got some that have made in like the one 0:09:25.080 --> 0:09:28.280 hundred million dollar range, although it's not entirely clear what 0:09:28.320 --> 0:09:31.240 their margins are or what the kind of the long 0:09:31.320 --> 0:09:34.559 term reliability of that profit is But the vast majority 0:09:34.600 --> 0:09:37.240 of money in AI, like almost all of it has 0:09:37.240 --> 0:09:40.160 been made by companies like Nvidia, and Vidia jumped up 0:09:40.160 --> 0:09:42.560 to become like a trillion dollar company as a result 0:09:42.600 --> 0:09:49.240 of this, because the hardware needs of these products are 0:09:49.280 --> 0:09:54.240 so intense, and obviously that shows there's money here for somebody. 0:09:54.240 --> 0:09:56.360 But the fact that like a shitload of people got 0:09:56.400 --> 0:10:00.520 curious about these apps and use them in in quick 0:10:00.559 --> 0:10:03.840 succession and then kind of dropped off is an evidence that, 0:10:04.000 --> 0:10:07.880 like we're seeing entire industries replaced as much of it 0:10:07.920 --> 0:10:10.480 as it is evidence that like a lot of people 0:10:10.520 --> 0:10:15.000 thought this was interesting briefly, and so I think kind 0:10:15.000 --> 0:10:17.360 of when you look at the data, one of the 0:10:17.360 --> 0:10:19.960 things that suggests is that we're heading towards a point 0:10:20.200 --> 0:10:22.880 in AI, and I think we're probably going to hit 0:10:22.920 --> 0:10:24.840 it within the next six months to a year that 0:10:24.960 --> 0:10:27.520 is broadly referred to as like the trial of disappointment, 0:10:28.360 --> 0:10:31.600 And this is what happens when kind of the promises 0:10:31.720 --> 0:10:33.880 of a new technology that are being made by the 0:10:33.920 --> 0:10:36.240 hypemen or con men as I tend to call them, 0:10:36.640 --> 0:10:40.400 meet with like the actual reality of its execution, which 0:10:40.400 --> 0:10:42.880 in some areas is going to be significant. There are places, 0:10:42.920 --> 0:10:45.719 I think medical research maybe one of them. We'll talk 0:10:45.720 --> 0:10:47.600 about that in a bit, where a lot of the 0:10:47.600 --> 0:10:52.000 promises people are making about AI will be fairly quickly realized, 0:10:52.040 --> 0:10:54.040 and then there are areas where it won't be. I 0:10:54.080 --> 0:10:57.640 think content generation is one of those things. But yeah, 0:10:57.880 --> 0:11:00.760 so that's kind of like what I'm seeing when I'm 0:11:00.760 --> 0:11:04.120 looking at the broad strokes of where this technology is 0:11:04.160 --> 0:11:06.319 here and kind of the gap between how people are 0:11:06.320 --> 0:11:08.880 talking about it and what we're actually seeing in terms 0:11:08.880 --> 0:11:21.160 of monetization. I want to talk a little bit now 0:11:21.240 --> 0:11:24.040 about kind of one of the guys I would call 0:11:24.120 --> 0:11:26.120 him kind of a con man who's been a big 0:11:26.200 --> 0:11:29.640 driver of the current AI push. He's a dude named 0:11:29.640 --> 0:11:34.160 Amad Mustock, and he's the founder of Stable Diffusion right, 0:11:34.200 --> 0:11:37.000 which is a text to image generator that was kind 0:11:37.040 --> 0:11:39.920 of like before chat GPT hit. This was like the 0:11:39.920 --> 0:11:44.800 first really really big mainstream AI thing. Chat GPT was 0:11:45.280 --> 0:11:49.440 a lot larger, but Stable Deviusion came first, and you know, 0:11:49.640 --> 0:11:51.839 was critical behind, among other things, a lot of the 0:11:52.160 --> 0:11:58.320 silliest NFT bullshit. And he's a really interesting dude, Like 0:11:58.360 --> 0:12:00.920 if you look at kind of his own claims his background. 0:12:01.640 --> 0:12:04.440 He says that he's got an Oxford master's degree, that 0:12:04.520 --> 0:12:07.960 he was like the behind an award winning hedge fund, 0:12:08.480 --> 0:12:10.800 that he like worked for the United Nations and a 0:12:10.840 --> 0:12:14.839 really important capacity, and also that he obviously founded this 0:12:14.840 --> 0:12:17.920 this a I bought. None of that's true. He has 0:12:17.960 --> 0:12:21.200 a bachelor's degree from Oxford, not a master's degree. 0:12:21.480 --> 0:12:24.280 He did well, that's what he's playing off. A thing 0:12:24.360 --> 0:12:27.640 that happens where like you can you can get if 0:12:27.640 --> 0:12:29.800 you have a BA ox and you can you can 0:12:29.840 --> 0:12:31.520 get it to be an MA. Doesn't mean you did 0:12:31.559 --> 0:12:36.840 a master's. It's just a wealthy people flex. Yeah, it's 0:12:37.080 --> 0:12:38.760 not a master's degree. You shouldn't qute it that. If 0:12:38.760 --> 0:12:40.040 you're quoting it now, you're taking the pace. 0:12:40.559 --> 0:12:42.720 Yeah. Yeah, he's taken the piss knowing no one's gonna 0:12:42.720 --> 0:12:45.239 call him on it, or at least knowing that people wouldn't, 0:12:45.320 --> 0:12:48.040 like at large, like loudly enough for it to matter 0:12:48.120 --> 0:12:52.199 for him. He hasn't worked with the UN in quite 0:12:52.200 --> 0:12:55.320 some time, and never did in a major capacity. He 0:12:55.400 --> 0:12:58.040 did run a hedge fund that was successful in its 0:12:58.080 --> 0:12:59.920 first year, but then got shut down in its sex 0:13:00.280 --> 0:13:04.080 year because he lost everybody's money. So like this is 0:13:04.160 --> 0:13:05.720 this is. But you see with this guy, if you 0:13:05.720 --> 0:13:08.199 go through his like history, he's like he's like chasing 0:13:08.240 --> 0:13:11.760 hedge funds in the early aughts. He first gets in 0:13:11.880 --> 0:13:14.840 with stable diffusion after COVID, and he's kind of like 0:13:14.880 --> 0:13:17.160 billing it as this is gonna help with like research 0:13:17.280 --> 0:13:20.720 into trying to like you know, fight the COVID nineteen pandemic, 0:13:21.880 --> 0:13:24.120 and then he kind of pivots to like, oh, this 0:13:24.240 --> 0:13:26.400 is a great way to like make NFTs and shit, 0:13:26.520 --> 0:13:28.360 you know when that hit, Like he's he's just sort 0:13:28.400 --> 0:13:31.880 of like chasing where the money is. Yeah, any way 0:13:31.920 --> 0:13:34.640 he kind of can. And he's not, by the way, 0:13:34.679 --> 0:13:36.280 he's not the guy who wrote any of the source 0:13:36.320 --> 0:13:38.360 code for this. That was done by like a group 0:13:38.440 --> 0:13:42.840 of researchers, and he you know, he essentially like acquired it, 0:13:43.320 --> 0:13:45.600 which is usually what happens here. Now, none of this 0:13:45.640 --> 0:13:48.240 has stopped him from getting one hundred million dollars or 0:13:48.320 --> 0:13:53.840 so in investments from various venture partners, and hasn't stopped 0:13:53.840 --> 0:13:56.439 his company from getting this massive violation. It hasn't stopped 0:13:56.440 --> 0:14:00.920 the White House from inviting him to talk as part 0:14:00.960 --> 0:14:04.640 of like a federal AI safety initiative, But it is 0:14:04.679 --> 0:14:06.640 one of those like when I kind of look into 0:14:06.679 --> 0:14:09.520 this guy and kind of the gap between his claims 0:14:09.559 --> 0:14:12.800 and what's actually happened and the claims that are being 0:14:12.840 --> 0:14:14.960 made about the value of his company and what it's 0:14:15.000 --> 0:14:17.760 actually like proved to be worth so far. I think 0:14:17.760 --> 0:14:20.560 a lot about Sam Bankman Freed because a lot of 0:14:20.600 --> 0:14:23.680 like the early writing around this guy was similar, and 0:14:23.720 --> 0:14:25.840 a lot of the kind of shit that he's claiming 0:14:26.080 --> 0:14:29.560 is similar. And yeah, I'm not sure if this is 0:14:29.600 --> 0:14:31.800 a case where because Bankman Freed is one of these 0:14:31.920 --> 0:14:36.240 people who, like Elizabeth Holmes, I think, backed the wrong 0:14:36.360 --> 0:14:40.920 technology because it's fine in Silicon Valley. It's fine, generally 0:14:40.920 --> 0:14:43.440 speaking in capitalism to lie about what a product can 0:14:43.480 --> 0:14:46.600 do if you can, you know, fake it till you 0:14:46.640 --> 0:14:49.400 make it. And maybe AI is there. He may have 0:14:49.440 --> 0:14:52.120 this guy may have made a good bet as to 0:14:52.160 --> 0:14:55.600 the future, but that's kind of far from certain yet. 0:14:55.640 --> 0:14:59.280 And it's it's just really clear how much of this 0:14:59.400 --> 0:15:02.640 industry is being built on, or is being built by. 0:15:02.680 --> 0:15:05.040 How much of the people running sort of these AI 0:15:05.160 --> 0:15:08.760 companies are dudes who managed one way or another, either 0:15:08.800 --> 0:15:12.760 through access to VC funding or kind of like you know, 0:15:12.880 --> 0:15:14.600 just being in the right place at the right time 0:15:15.240 --> 0:15:18.560 to jump in on the bandwagon in the hopes that 0:15:18.560 --> 0:15:20.640 they'll be able to cash out very, very quickly. I 0:15:20.680 --> 0:15:23.720 found a good quote from a Forbes article talking about 0:15:23.760 --> 0:15:28.400 like a big part of why guys like Mustock are 0:15:28.480 --> 0:15:31.720 so interested in AI right now from a financial perspective, 0:15:32.640 --> 0:15:34.640 And this is true, not just this was true about 0:15:34.680 --> 0:15:38.840 like crypto before, but AI because there's more to the technology. 0:15:39.960 --> 0:15:44.560 This is kind of even more so valid quote. Venture 0:15:44.600 --> 0:15:47.960 capitalists historically spend months performing due diligence, a process that 0:15:48.000 --> 0:15:50.760 involves analyzing the market, vetting the founder, and speaking to 0:15:50.760 --> 0:15:53.640 customers to check for red flags before investing in a startup, 0:15:54.000 --> 0:15:56.600 but start to finish. Mstock told Forbes he needed just 0:15:56.640 --> 0:15:59.040 six days to secure one hundred million dollars from leading 0:15:59.080 --> 0:16:02.480 investment firms Coach You and light Speed once Stable Diffusion 0:16:02.520 --> 0:16:05.120 went viral. The extent of due diligence at the firms 0:16:05.160 --> 0:16:08.080 performed is unclear given the speed of the investment. The 0:16:08.080 --> 0:16:10.520 investment thesis we had is that we don't know exactly 0:16:10.560 --> 0:16:12.480 what all the use cases will be, but we know 0:16:12.560 --> 0:16:15.080 that this technology is truly transformative and has reached a 0:16:15.080 --> 0:16:17.880 tipping point in terms of what it can do. Garav Gupta, 0:16:17.920 --> 0:16:20.280 the light Speed partner who led the investment, told Forbes 0:16:20.320 --> 0:16:23.040 in a January interview. So again they're being like, yeah, 0:16:23.040 --> 0:16:25.720 we're pumping tens of million dollars of dollars into this. 0:16:25.840 --> 0:16:28.680 We don't know how it'll make money. It just seems 0:16:28.720 --> 0:16:31.880 so impressive that it has to be profitable. Now that 0:16:32.080 --> 0:16:37.280 line is particularly funny, maybe the wrong word when compared 0:16:37.360 --> 0:16:40.760 alongside this paragraph from later in the article. In an 0:16:40.760 --> 0:16:44.560 open letter last September, Democratic Representative Anna Essue urged action 0:16:44.640 --> 0:16:47.360 in Washington against the open source nature of stable Diffusion. 0:16:47.600 --> 0:16:49.720 The model, she wrote, had been used to generate images 0:16:49.760 --> 0:16:52.800 of violently beaten Asian women and pornography, some of which 0:16:52.800 --> 0:16:56.440 portrays real people. Bashara said new versions of stable Diffusion 0:16:56.440 --> 0:17:00.440 filtered data for potentially unsafe content, helping to prove users 0:17:00.480 --> 0:17:04.119 from generating harmful images in the first place. So it's like, 0:17:06.359 --> 0:17:11.000 part of what's happening here is you've got this thing 0:17:11.119 --> 0:17:14.560 that seems really impressive, and that is to some extent 0:17:14.560 --> 0:17:17.399 because it's able to like remix stuff that exists in 0:17:17.440 --> 0:17:21.120 a way that you haven't done automatically before. But all 0:17:21.119 --> 0:17:23.560 of these kind of valuations are based number one, and 0:17:23.680 --> 0:17:28.000 ignoring the problems with monetizing this stuff, including like the 0:17:28.080 --> 0:17:30.800 still very much unsorted nature of how copyright's going to 0:17:30.800 --> 0:17:34.040 affect this, and also like the question of is this 0:17:34.160 --> 0:17:39.399 really worth that much money? Like is this actually is 0:17:39.440 --> 0:17:43.639 being able to generate kind of weird slightly off putting 0:17:43.680 --> 0:17:50.119 AI images a huge business, like how much of because 0:17:50.200 --> 0:17:51.800 like from where I'm seeing it, one of two things 0:17:51.840 --> 0:17:55.439 is possible. Number one, this replaces all art everywhere and 0:17:55.520 --> 0:17:57.960 so there's a shitload of money in it. Or number two, 0:17:58.480 --> 0:18:01.600 this remains a way that like low quality websites and 0:18:02.000 --> 0:18:05.760 like Amazon drop ship scammers who are like putting up 0:18:05.800 --> 0:18:09.600 fake books on Kindle and whatnot to trick people using 0:18:09.640 --> 0:18:12.639 keywords like that this is just like a way to 0:18:12.680 --> 0:18:15.560 fill that shit out. Like I don't see a whole 0:18:15.560 --> 0:18:17.680 lot of room in the middle there. You know, maybe 0:18:17.680 --> 0:18:21.240 I'm being like overly pessimistic there, but that's that's that's 0:18:21.280 --> 0:18:22.080 where I'm sitting. 0:18:22.600 --> 0:18:25.639 I mean, some of the models we've seen used is 0:18:25.800 --> 0:18:30.439 selling like subscription packs for like access to these tools 0:18:30.480 --> 0:18:33.560 and access to use them for like commercial reasons. The 0:18:33.640 --> 0:18:36.680 other thing we can see is just like corporations selling 0:18:36.680 --> 0:18:40.560 to other corporations like basically having Disney and Warner Brothers 0:18:40.600 --> 0:18:42.840 be able to use this to generate concept art and 0:18:42.960 --> 0:18:45.280 now they don't need to pay concept artists and instead 0:18:45.359 --> 0:18:49.399 they just have like pretty pretty uh pretty like nicely 0:18:49.480 --> 0:18:53.480 curated tools for them to generate this type of Yeah A, 0:18:53.640 --> 0:18:56.360 I imagining those are kind of two of the biggest 0:18:56.480 --> 0:18:59.440 use cases that at least I'm seeing right now from 0:19:00.200 --> 0:19:04.080 more on like the creative filmmaking art side of things. 0:19:04.800 --> 0:19:06.320 Because I mean, I don't think it's going to replace 0:19:06.359 --> 0:19:10.959 all all art. I think nobody nobody is actually uh 0:19:11.400 --> 0:19:13.680 is actually thinking it's just going to replace all all art, 0:19:13.760 --> 0:19:16.159 just like photography did not replace all art. It just 0:19:16.200 --> 0:19:19.520 it changes the paradigm. And because this this tool does 0:19:19.560 --> 0:19:24.119 seem like specifically useful for the for the way that 0:19:24.160 --> 0:19:27.560 we're seeing like corporations make the same a movie every 0:19:27.600 --> 0:19:29.880 five years, like it's all it's it's it's it's all 0:19:29.880 --> 0:19:32.480 built on all of the same stuff. And I think 0:19:32.480 --> 0:19:34.160 that that's how a lot of a lot of it's 0:19:34.160 --> 0:19:35.840 gonna get used. It's gonna be a lot of weird 0:19:35.840 --> 0:19:40.080 scam artists, people just messing around for fun, and then 0:19:40.320 --> 0:19:43.640 people not paying like illustrators as much. 0:19:43.600 --> 0:19:47.040 Like yeah, and I think that's kind of like I 0:19:47.080 --> 0:19:49.760 see this being adopted widely, but that's not the same 0:19:50.040 --> 0:19:54.400 as it like being a huge success. Like right now 0:19:54.400 --> 0:19:57.119 I'm looking at an article that's estimating the current value 0:19:57.119 --> 0:19:59.000 of AI in the US is at one hundred billion 0:19:59.040 --> 0:20:01.800 dollars and that buy twenty thirty, it'll be worth two 0:20:01.800 --> 0:20:05.400 trillion US dollars. And it's like, I don't know, man, 0:20:05.520 --> 0:20:06.360 like is. 0:20:06.400 --> 0:20:08.920 I mean the AI is more than just like mid 0:20:08.960 --> 0:20:12.720 Journey image creation right there, is just like open AI 0:20:12.840 --> 0:20:16.000 and chat GPT, and like AI is in everything we 0:20:16.160 --> 0:20:18.520 use now, like yeah, like AI is in your smartphone. 0:20:18.520 --> 0:20:20.880 AI is going to be in your refrigerator soon. It's 0:20:20.880 --> 0:20:23.600 like it's not just image generation by any means. 0:20:23.760 --> 0:20:25.600 That kind of gets to what I'm saying, because that's 0:20:25.760 --> 0:20:29.119 that's when you look at AI as a tool. Is 0:20:29.160 --> 0:20:30.800 more of like a paint brush than a painter. Is 0:20:30.840 --> 0:20:33.199 a tool that will like augment or be used in 0:20:33.240 --> 0:20:34.959 because I think a lot of a number of times 0:20:35.320 --> 0:20:36.800 it may be used in a way that makes the 0:20:36.840 --> 0:20:41.000 product worse and a lot of existing technologies, Well that's 0:20:41.000 --> 0:20:44.520 really different from kind of number one, the doom and gloom, 0:20:44.600 --> 0:20:47.440 like this is an intelligence on its own that could 0:20:47.480 --> 0:20:51.040 like overtake humanity. I think the worry is more like 0:20:51.800 --> 0:20:54.480 this could make get adopted on such a large scale 0:20:54.520 --> 0:20:56.239 that it like makes a lot of shit worse. Like 0:20:56.400 --> 0:20:59.040 my biggest fear with AI is that it kind of 0:20:59.119 --> 0:21:02.119 hyper charges the SEO industry and the way that that 0:21:02.240 --> 0:21:04.919 has worked to destroy search and destroy so much of 0:21:04.960 --> 0:21:05.719 Internet content. 0:21:06.040 --> 0:21:08.560 Yeah, I think that is very possible. Like if I 0:21:08.600 --> 0:21:11.560 look at chat GPT, like, I don't think that's going 0:21:11.600 --> 0:21:14.639 to be writing features for Rolling Stone anytime soon, but 0:21:15.080 --> 0:21:19.720 what it can probably do because SEO max copy is derivative, right, 0:21:19.760 --> 0:21:23.080 like like it's predictable, it's derivative, it's based on other stuff. 0:21:22.800 --> 0:21:23.480 It's supposed to be. 0:21:23.840 --> 0:21:26.600 Yeah, and so it can do that SEO max copy 0:21:26.760 --> 0:21:30.360 and some of that ad copy like very well, and yeah, 0:21:30.560 --> 0:21:33.959 either really fuck up searches, which is quite possible, and 0:21:34.040 --> 0:21:37.800 also make the lowest kind of acceptable tier of that 0:21:37.920 --> 0:21:42.439 kind of copy what it can generate. And because you 0:21:42.440 --> 0:21:43.919 can just shove that copy in front of people with 0:21:44.040 --> 0:21:45.960 SEO max and then have shitty AD copy written by 0:21:46.000 --> 0:21:50.919