WEBVTT - Money Has No Meaning: OpenAI, AIWB, WWJD 0:00:03.160 --> 0:00:08.720 Bloomberg Audio Studios, Podcasts, Radio News. 0:00:10.680 --> 0:00:13.640 Hello and welcome the Money Stuff Podcast. You're a weekly 0:00:13.680 --> 0:00:17.680 podcast where we talk about stuff related to money. I'm 0:00:17.680 --> 0:00:19.759 Matt Levian and I write the Money Stuff column for 0:00:19.840 --> 0:00:20.560 Bloomberg Good Pay. 0:00:20.960 --> 0:00:23.640 And I'm Katie Greifeld, a reporter for Bloomberg News and 0:00:23.680 --> 0:00:25.280 an anchor for Bloomberg Television. 0:00:26.200 --> 0:00:27.840 What are you talking about today, Katie? 0:00:28.160 --> 0:00:31.160 We're going to talk about open Aire, all the drama. 0:00:31.400 --> 0:00:35.000 We're going to talk about chatbots and stock picking, and 0:00:35.040 --> 0:00:37.480 then we're going to talk about bibles. 0:00:37.920 --> 0:00:39.200 I'm gonna bell out right now. 0:00:44.240 --> 0:00:48.720 Sam Altman open Ai. You know how open ai is 0:00:48.840 --> 0:00:49.600 a nonprofit? 0:00:49.960 --> 0:00:51.760 I don't well explain it to me. 0:00:52.800 --> 0:00:56.600 No, we don't need to, because it's not anymore. It 0:00:56.640 --> 0:00:59.560 won't be. According to people familiar with the matter, of course, 0:00:59.600 --> 0:01:02.600 they're just discussing how to restructure to become a for 0:01:02.840 --> 0:01:08.400 profit business officially and giving Sam Altman a seven percent 0:01:08.480 --> 0:01:09.160 equity stake. 0:01:09.440 --> 0:01:10.080 It's so good. 0:01:10.360 --> 0:01:10.600 Yeah. 0:01:10.840 --> 0:01:15.000 I read about this that if you just like didn't 0:01:15.280 --> 0:01:18.520 know that it was a nonprofit and you just ignored 0:01:18.720 --> 0:01:20.840 all of that stuff, you would sort of have a 0:01:20.840 --> 0:01:23.479 better understanding of open Ai than if you did think 0:01:23.520 --> 0:01:28.320 about it's nonprofit governance structure, because you know, they had 0:01:28.360 --> 0:01:31.480 a nonprofit board whose job was to develop AI for 0:01:31.520 --> 0:01:35.640 the benefit of humanity, and the nonprofit board decided that 0:01:35.720 --> 0:01:39.480 the CEO of Sam Altman was apparently developing it not 0:01:39.600 --> 0:01:41.560 for the benefit of humanity, and so they fired him, 0:01:41.600 --> 0:01:44.240 and then everyone was like, never mind, there's no nonprofit board. 0:01:44.480 --> 0:01:46.240 Well that's the thing. I feel like a lot of 0:01:46.280 --> 0:01:49.600 people didn't know that open ai was a nonprofit until 0:01:49.640 --> 0:01:51.800 they fired and then rehired Sam Ali. 0:01:51.880 --> 0:01:52.280 Yeah. 0:01:52.480 --> 0:01:54.920 Yeah, there was like a five day period where you 0:01:55.000 --> 0:01:58.160 could be like, wow, open ai is really a weird 0:01:58.240 --> 0:02:01.600 company and not just the hot tech startup. And then 0:02:01.640 --> 0:02:04.000 they're like, nah, we're just kidding. We're a context out 0:02:04.520 --> 0:02:07.080 And so now they're going to be a for profit 0:02:07.160 --> 0:02:07.520 for real. 0:02:07.720 --> 0:02:10.720 Yeah, taking the last step they're considering, according to people 0:02:10.720 --> 0:02:14.480 familiar with the matter, becoming a public benefit corporation, tasked 0:02:14.480 --> 0:02:17.600 with turning a profit but also helping society. But Matt, 0:02:17.680 --> 0:02:22.040 are those two things at odds? Can you help society 0:02:22.080 --> 0:02:24.040 and still make money? I don't know. 0:02:24.320 --> 0:02:26.920 This podcast will try to find out. Look, you know 0:02:27.000 --> 0:02:28.880 you can be a public benefit corporation. You know you're 0:02:28.880 --> 0:02:32.080 trying to make a profit. But like you, the distinction 0:02:32.160 --> 0:02:35.120 machine a public benefit corporation and a regular corporation is 0:02:35.200 --> 0:02:37.680 pretty vague, right, because you can be just a regular 0:02:37.760 --> 0:02:41.400 old for profit corporation and like also try to be good, 0:02:42.040 --> 0:02:44.800 and the constraints on you as a public benefit corporation 0:02:44.840 --> 0:02:47.959 are not that constraining, So you can be a public 0:02:47.960 --> 0:02:51.280 benefit corporation and be kind of bad. But you know, 0:02:54.120 --> 0:02:56.520 if you're going from being a nonprofit company to being 0:02:56.560 --> 0:02:58.760 a for profit company, it probably does make sense to 0:02:59.000 --> 0:03:01.760 stop along the way being a public benefit company because 0:03:01.800 --> 0:03:05.200 it like sounds better, sounds like less of a stark transition. 0:03:05.639 --> 0:03:06.760 It still has benefit. 0:03:07.000 --> 0:03:10.000 Yeah, it's still for the benefit of the public. You ask, 0:03:10.960 --> 0:03:14.880 somewhat sarcastically, can you do good while also making money? 0:03:15.280 --> 0:03:19.200 And I think it's really interesting that, like the Silicon 0:03:19.320 --> 0:03:24.079 Valley ethos is so strongly of the view that you 0:03:24.280 --> 0:03:27.440 always read VCS saying Elon Musk has done more for 0:03:27.560 --> 0:03:31.079 humanity in its for profit companies than you know, any philanthropist. 0:03:31.520 --> 0:03:34.000 And you know, I think there's some truth to the 0:03:34.040 --> 0:03:39.360 idea that like for profit development of technology can be 0:03:39.440 --> 0:03:42.680 really good for the world and also make the people 0:03:42.720 --> 0:03:46.960 who do it really rich, and it was always at odd. 0:03:47.000 --> 0:03:51.279 It's with that ethos that open ai was a nonprofit 0:03:51.920 --> 0:03:54.000 and always very awkward because you know it was like 0:03:54.080 --> 0:03:57.880 founded by Silicon Valley, you know, sounded by literally Elon 0:03:57.960 --> 0:04:00.120 Musk and like some PC types, you know, s Ham 0:04:00.160 --> 0:04:02.640 Altman is you know kind of a he's like a 0:04:02.760 --> 0:04:06.360 you know, bicombinator, like the VC guy, and you know 0:04:06.440 --> 0:04:10.080 it's staffed by tech people, and all of those people, 0:04:10.120 --> 0:04:13.040 I think, to some degree in their previous lives brought 0:04:13.120 --> 0:04:16.240 into like the Silicon Valley ethos of like the way 0:04:16.320 --> 0:04:23.920 to improve society is to to scalable low marginal cost 0:04:24.400 --> 0:04:27.680 technology products that make a lot of money for their developers. 0:04:27.880 --> 0:04:30.600 And there's this little blip of like at open Ai, 0:04:30.680 --> 0:04:31.720 I was like, no, we're going to do it a 0:04:31.760 --> 0:04:35.039 different way, and then they changed their months. So clearly 0:04:35.640 --> 0:04:38.360 a lot of people at open ai think that they 0:04:38.360 --> 0:04:41.080 can do good for humanity while also making a lot 0:04:41.120 --> 0:04:46.240 of money. So the mission of open ai being this 0:04:46.400 --> 0:04:50.120 technology in particular cannot be commercialized, This technology in particular 0:04:50.400 --> 0:04:53.520 has to be done by a nonprofit, which was apparently 0:04:53.560 --> 0:04:56.080 the theory for a while. It's always a little weird, right, 0:04:56.120 --> 0:04:58.400 because like the product they've rolled out is sort of 0:04:58.440 --> 0:05:01.359 this like consumer chat pottery. It's on a continuum with 0:05:01.400 --> 0:05:04.240 a lot of, like you know, for profit technology products. 0:05:04.360 --> 0:05:08.240 I think there's like the overlay of open ai believing 0:05:08.279 --> 0:05:14.000 that AI could enslave humanity or red work obsolete, or 0:05:14.040 --> 0:05:16.919 like usher in a new age of abundance where money 0:05:16.960 --> 0:05:19.880 has no meaning. Yeah, if you really believe that your 0:05:19.920 --> 0:05:23.520 product is going to create an age where money has 0:05:23.600 --> 0:05:26.200 no meaning, then I suppose it makes sense to operate 0:05:26.240 --> 0:05:32.040 it as a nonprofit. Yeah, but you know that market. 0:05:32.080 --> 0:05:35.280 Well, the point before right where I mean Altman saying 0:05:35.320 --> 0:05:38.240 publicly like I'm so scared of what we could be 0:05:38.360 --> 0:05:41.600 creating is a great pitch for why you should give 0:05:41.800 --> 0:05:43.680 me a lot of money. Like what we're working on 0:05:43.880 --> 0:05:47.240 is so incredible that I am scared, right. 0:05:47.320 --> 0:05:50.320 It makes it seem very important, and so it helps 0:05:50.360 --> 0:05:52.760 them raise a lot of money. But eventually, you know, 0:05:53.240 --> 0:05:54.960 those people who give them all the money want their 0:05:54.960 --> 0:05:55.960 money back. Yeah. 0:05:55.960 --> 0:05:58.680 I don't think like all the investors pouring so much 0:05:58.720 --> 0:06:02.120 money into open ai, which is raising right now with 0:06:02.200 --> 0:06:05.240 a one hundred and fifty billion dollars valuation, or necessarily 0:06:05.240 --> 0:06:09.760 investing to help humanity, they're investing with the goal of 0:06:10.120 --> 0:06:11.920 making a lot of money, I would imagine. 0:06:12.000 --> 0:06:14.000 Yeah. And by the way, like even before this like 0:06:14.080 --> 0:06:17.600 supposed change, open AI has this for profit subsidiary that 0:06:18.560 --> 0:06:21.280 was already able to give to promise investors a big 0:06:21.320 --> 0:06:23.039 return on their money. It was like a cap to 0:06:23.080 --> 0:06:25.480 profit subsidiary, so you could only make one hundred times your. 0:06:25.360 --> 0:06:27.800 Money up change. 0:06:28.000 --> 0:06:31.560 So what is changing is I think two things. One 0:06:31.680 --> 0:06:36.080 is like, like right now the sort of notional corporate 0:06:36.080 --> 0:06:39.040 structure is that on top is a nonprofit and then 0:06:39.040 --> 0:06:41.520 there's like the cap profit subsidiary under it, and you 0:06:41.560 --> 0:06:44.360 can invest in the cap profit subsidiary and get profits. 0:06:44.600 --> 0:06:48.480 But ultimately everything answers to the nonprofit board. The nonprofit 0:06:48.520 --> 0:06:51.719 board has no fiduciary duties to investors. Their duty is 0:06:51.720 --> 0:06:54.680 to humanity to develop AI in the best way. And 0:06:55.160 --> 0:06:59.080 you know what happened is the old nonprofit board thought 0:06:59.080 --> 0:07:02.479 that Sam all And wasn't doing that in some unspecified way, 0:07:02.839 --> 0:07:04.280 and so they fired him, and then they all got 0:07:04.279 --> 0:07:07.480 fired themselves. And so now there's a new nonprofit board 0:07:07.560 --> 0:07:10.000 that is like a little bit more profit oriented. Yeah, 0:07:10.040 --> 0:07:14.160 but it's still technically does an answer to the investors. 0:07:14.160 --> 0:07:16.880 And I think the restructuring would mean that the board 0:07:16.960 --> 0:07:20.440 is now more directly responsible to investors. And then the 0:07:20.480 --> 0:07:24.040 other thing that's happening in this change besiudes, like the 0:07:24.160 --> 0:07:28.440 for profit being the top company is sam Aldman is 0:07:28.480 --> 0:07:30.120 supposedly getting seven percent of the company. 0:07:30.160 --> 0:07:33.040 It's spend a lot of money. Theoretically. 0:07:33.680 --> 0:07:39.560 I was thinking, if like Harvard University became a for 0:07:39.720 --> 0:07:44.360 profit company tomorrow, would they give the president ten billion dollars. 0:07:45.120 --> 0:07:50.400 It's like a really interesting you know, like typically founders 0:07:50.440 --> 0:07:53.400 of hot tech companies get a lot of equity. Yeah, 0:07:53.440 --> 0:07:55.560 because they retain a lot of equity, right, Like they 0:07:55.640 --> 0:07:57.360 start with one hundred percent of the company and then 0:07:57.400 --> 0:07:59.520 they raise money by selling portion of the company. But 0:07:59.560 --> 0:08:01.040 like they started with a undred percent and they end 0:08:01.080 --> 0:08:03.280 up with you know, seven or ten or twenty or 0:08:03.320 --> 0:08:06.440 fifty percent. Sam Alban started with zero percent because it 0:08:06.480 --> 0:08:10.119 was a nonprofit, and now they're retroactively being like, wow, okay, fine, 0:08:10.120 --> 0:08:11.920 it was a regular startup, so you can't have zero 0:08:12.000 --> 0:08:13.960 percent equity, so we're going to give you seven percent. 0:08:14.320 --> 0:08:17.400 And so because he built this company from zero to 0:08:17.800 --> 0:08:20.520 let's say, one hundred and fifty billion dollar valuation, even 0:08:20.560 --> 0:08:25.080 though he did that without any equity stake. They're retroactively saying, well, 0:08:25.200 --> 0:08:26.960 you did that, so you're a founder, so you get 0:08:26.960 --> 0:08:28.760 ten billion of that value you created. 0:08:29.120 --> 0:08:32.440 I like equation that went into How did they land 0:08:32.440 --> 0:08:33.280 at seven percent? 0:08:34.160 --> 0:08:34.280 Like? 0:08:34.320 --> 0:08:34.600 Did they? 0:08:34.760 --> 0:08:37.200 I'm very curious about the capital structure of this company generally, right, 0:08:37.240 --> 0:08:39.800 because like this company is like you know, if you 0:08:39.920 --> 0:08:41.880 if you just think of the for profit subsidiary, right, 0:08:41.920 --> 0:08:45.520 like Microsoft owns quite a chunk of it, because they've 0:08:45.520 --> 0:08:48.319 put a lot of money in at like lower evaluations 0:08:48.600 --> 0:08:52.840 than the current one. And you know, there's some VC investors. 0:08:52.880 --> 0:08:56.360 And then there are a lot of employees, not Sam alten, 0:08:56.400 --> 0:08:59.000 but a lot of employees have gotten you know, chunky 0:08:59.280 --> 0:09:01.839 stock grants of the years. They're not called stock grunts. 0:09:01.840 --> 0:09:04.199 They're called something like performance something units, but they're. 0:09:04.080 --> 0:09:06.800 Stock grunts, right, and benefit grants. 0:09:06.920 --> 0:09:10.840 Yeah, there's stocks. And then the rest of the company 0:09:10.880 --> 0:09:12.560 is owned by humanity, right, the rest of the company's 0:09:12.559 --> 0:09:17.160 owned by the nonprofit. Right. No, and so the nonprofit 0:09:17.240 --> 0:09:21.079 now is basically allocating money away from humanity to say, 0:09:21.200 --> 0:09:24.199 Sam Altman, And how did they come up with that number? 0:09:25.160 --> 0:09:28.640 I don't know. I think that like it has to 0:09:28.679 --> 0:09:30.920 be a lot because first of all, he's a billionaire 0:09:30.960 --> 0:09:33.360 front it's all their adventures. But it has to be 0:09:33.440 --> 0:09:36.480 a lot because if you're just recasting open Ai as 0:09:36.520 --> 0:09:42.040 a regular startup, then the founder CEO has to own 0:09:42.080 --> 0:09:44.880 a lot of it because he has to be incentivized 0:09:44.880 --> 0:09:46.199 in the right way, right. I Mean, one thing that's 0:09:46.200 --> 0:09:47.880 happened in the open a story in the last year 0:09:47.960 --> 0:09:50.640 or two is that you're constantly reading about Sam Altman 0:09:50.720 --> 0:09:52.760 doing like other weird stuff. You know, he's like getting 0:09:52.800 --> 0:09:54.920 involved in like data center deals and like power deals 0:09:54.920 --> 0:09:58.920 and like other startups that use open ai technology, and 0:09:59.080 --> 0:10:01.920 they all feel like confident of interest. And one explanation 0:10:02.000 --> 0:10:05.360 of that is, Okay young zero percent of open Aif 0:10:05.360 --> 0:10:07.520 she like can like make a little money on the 0:10:07.600 --> 0:10:10.080 side by doing some sort of nuclear power deal, then 0:10:10.080 --> 0:10:12.920 he'll do that. And so if you're an investor looking 0:10:12.920 --> 0:10:14.520 to put money into open Ai at one hundred and 0:10:14.520 --> 0:10:17.920 fifty billion dollar valuation, one thing you want is for 0:10:18.000 --> 0:10:21.280 the CEO to be mostly working for open Ai. Yeah, 0:10:21.280 --> 0:10:23.120 and you can imagine a lot of ways to try 0:10:23.160 --> 0:10:25.559 to do that. But like the standard silicon of Valueay 0:10:25.600 --> 0:10:26.959 is like he has most of his net worth in 0:10:27.040 --> 0:10:29.600 open ai stock, so you have to give him enough 0:10:29.600 --> 0:10:31.800 stock that most of his network is in open A stock. 0:10:31.840 --> 0:10:33.439 And that's like, you know, come doing it. 0:10:33.640 --> 0:10:37.520 She with Elon Musk parallels because you want the founder 0:10:37.679 --> 0:10:41.400 CEO to be engaged and passionate and working on the company. 0:10:41.720 --> 0:10:44.120 Elon Musk when it comes to Tesla, I mean that 0:10:44.240 --> 0:10:47.240 was basically his argument that I want to focus on 0:10:47.320 --> 0:10:49.120 the thing where I have the most at stake here, 0:10:49.920 --> 0:10:51.680 and that's why I should own more of Tesla. 0:10:51.800 --> 0:10:55.360 Yeah. Absolutely, I don't know who is asking for what here. 0:10:55.520 --> 0:10:57.840 Like you could imagine I'm not saying this is true. 0:10:57.880 --> 0:10:59.720 I don't think this is true. But you could imagine 0:10:59.720 --> 0:11:01.200 Sam and be like, no, I'm good, I have two 0:11:01.200 --> 0:11:03.720 billion dollars. I don't need more money from this company. 0:11:03.960 --> 0:11:05.880 I'm in it for the benefit of humanity. You could 0:11:05.880 --> 0:11:09.679 imagine the board of the investors saying, no, you need 0:11:09.720 --> 0:11:11.840 to have a ligned incentives. You need to own seven 0:11:11.880 --> 0:11:14.440 percent of the company. Here, take ten billion dollars so 0:11:14.480 --> 0:11:16.920 that we know you're working for us. You can imagine that. 0:11:16.920 --> 0:11:19.520 Some time twists my arm, all right. 0:11:19.520 --> 0:11:21.800 Right, right, all right, And like, by the way, Elon 0:11:21.920 --> 0:11:24.120 Musk will ask for the money, but he will also say, 0:11:24.160 --> 0:11:26.000 I don't want it for the money. I just want it, 0:11:26.040 --> 0:11:27.520 you know, to make sure that I'm building it in 0:11:27.520 --> 0:11:31.520 the right place. So yeah, it's totally like Elon. It's 0:11:31.559 --> 0:11:35.480 a little different because Elon has a lot of very 0:11:35.520 --> 0:11:38.640 different ideas, but like have a lot of conflicts. But 0:11:38.760 --> 0:11:41.160 you understand why they're different. Companies like Sam all been 0:11:41.200 --> 0:11:44.440 when he's like raising money for ventures that build on 0:11:44.480 --> 0:11:46.280 top of open Ai, it's like, well, let's do that 0:11:46.320 --> 0:11:48.760 at open Ai. It's a little bit more like it 0:11:48.800 --> 0:11:51.240 looked like he was using his position at open ai 0:11:51.360 --> 0:11:53.000 to make money on the side because he couldn't make 0:11:53.000 --> 0:11:54.640 money at open ai. And now it's like, wow, I 0:11:54.640 --> 0:11:56.200 can make a lot of money open Ai, so do 0:11:56.280 --> 0:11:59.560 that instead. But yes, the perils that Elon are interesting, 0:11:59.640 --> 0:12:02.559 right because I said, you know, the notion that Elon 0:12:02.679 --> 0:12:04.560 Musk has done a lot for humanity by running for 0:12:04.640 --> 0:12:08.240 profit companies, It's like validates dispute to being a for 0:12:08.320 --> 0:12:11.199 profit company. On the other hand, Elon Musk is also 0:12:11.320 --> 0:12:14.040 suing open ai for becoming a for profit company. 0:12:14.080 --> 0:12:14.760 That's so true. 0:12:15.000 --> 0:12:18.280 This is it has one circumstance. He does not believe 0:12:18.320 --> 0:12:20.880 you should be benefiting humanity and also making a profit. 0:12:21.040 --> 0:12:22.960 It feels like a while since we've talked about Elon. 0:12:23.360 --> 0:12:26.240 I wonder like if the resolution to all of this 0:12:26.840 --> 0:12:29.160 is open a eye is now a for profit company. 0:12:29.200 --> 0:12:31.840 Microsoft will own x percent of it, Sam Alton and 0:12:32.520 --> 0:12:36.480 it humanity will ow you know, some single percentage. 0:12:35.960 --> 0:12:37.320 Of Well give you a benefit. 0:12:37.400 --> 0:12:39.440 We just give you on you know, an equity stake 0:12:39.520 --> 0:12:43.320 for his his contributions. Why not get on side then 0:12:43.360 --> 0:12:45.240 you get you on a line. Yeah, that's good. Although 0:12:45.280 --> 0:12:46.760 he's got his own AI company, so I guess he 0:12:46.800 --> 0:12:47.400 cant do that. 0:12:47.520 --> 0:13:03.040 True? Was that XII chat thoughts? Shall we? Shall we 0:13:03.080 --> 0:13:04.520 talk about Bridget. 0:13:04.600 --> 0:13:06.240 Another AI transition. 0:13:05.960 --> 0:13:07.679 Bridget Bridget bridge Wise. 0:13:08.000 --> 0:13:10.760 So Bridgewise is this Israeli tech company, fintech company that 0:13:11.160 --> 0:13:14.760 is launching a chatbot that will be your broker. 0:13:15.520 --> 0:13:18.199 Yeah. 0:13:19.080 --> 0:13:22.120 So it's this fascinating story because you read about it 0:13:22.120 --> 0:13:23.839 and you're like, Okay, what does this chatbot do, like 0:13:23.880 --> 0:13:26.240 gives your stock recommendations? Yeah? So, I mean I think 0:13:26.240 --> 0:13:27.640 it's like you type of O. Can you tell me 0:13:27.679 --> 0:13:30.480 like the top five stocks in like the industrial sector 0:13:30.520 --> 0:13:32.360 that I should buy today? And it's like, yeah, well, 0:13:32.400 --> 0:13:35.160 looking at the researcher, but like you can imagine just 0:13:35.360 --> 0:13:37.960 being like, what stocks should I buy? And then Bridgett 0:13:38.000 --> 0:13:40.240 says you should buy these five stocks, and you go 0:13:40.320 --> 0:13:43.200 buy those five stocks, and then the next day you 0:13:43.240 --> 0:13:45.440 come back and you're like, bridget they went down. What's 0:13:45.480 --> 0:13:45.880 going on? 0:13:46.400 --> 0:13:48.760 Yeah, Like, what information do you? What is the prompt? 0:13:48.840 --> 0:13:52.000 Prompts matter so much with these chatbots. Like if I 0:13:52.000 --> 0:13:55.400 said my name is Katie, I'm thirty one, I love horses, 0:13:56.000 --> 0:13:57.760 what would you tell me to buy? Versus if I 0:13:57.760 --> 0:14:01.959 said I'm Katie, you're up in New Jersey. That's it. 0:14:02.480 --> 0:14:05.880 But those are bad prompts. The prompt that you want 0:14:05.920 --> 0:14:07.319 is what stocks will go up the most? 0:14:07.679 --> 0:14:10.360 Well, yeah, because you know you don't care. 0:14:11.120 --> 0:14:13.160 But you're on the right track, because like this thing 0:14:13.200 --> 0:14:16.040 is clearly a substitute for a broker, right. This is 0:14:16.080 --> 0:14:18.720 like you call your broker, and your broker gets to 0:14:18.760 --> 0:14:21.280 know you and learns about your risk tolerances and your 0:14:21.320 --> 0:14:23.920 preferences and the themes that you're interested in. And then 0:14:23.920 --> 0:14:25.920 your broker is like, well, given your risk tolerance and 0:14:25.960 --> 0:14:27.840 you're interested in the theme of horses, you should buy 0:14:27.840 --> 0:14:31.640 the horse, etf or whatever. Right, But that's true because 0:14:32.120 --> 0:14:35.280 brokerage is a client service business. Where the product is 0:14:35.320 --> 0:14:38.160 making the client happy with the service. But you can 0:14:38.240 --> 0:14:41.400 imagine a different model where the product is like buying 0:14:41.440 --> 0:14:44.760 stocks that go up. Yeah, right, I think that we 0:14:44.840 --> 0:14:47.240 all understand that when you call your broker and you 0:14:47.280 --> 0:14:49.280 say what stock should I either be like, what are 0:14:49.280 --> 0:14:51.200 your risk tolerances? You should be in this CTF because 0:14:51.200 --> 0:14:53.400 there's a good balance of what like they'll say something 0:14:53.480 --> 0:14:56.000 other than like I know which stocks will go up 0:14:56.120 --> 0:14:58.600 and these are them and you should buy them, because 0:14:58.640 --> 0:15:01.960 like if they knewhich stocks would go up, yeah, they 0:15:01.960 --> 0:15:03.120 would be doing something else. 0:15:04.560 --> 0:15:08.240 Why can't I just ask chat ChiPT what stocks to buy? 0:15:08.760 --> 0:15:10.760 You can apparently sometimes tell you I can't give you 0:15:10.800 --> 0:15:11.600 stock recommendations. 0:15:11.840 --> 0:15:13.640 Oh, you can get around it. 0:15:13.680 --> 0:15:18.160 So chat GPT is like a general purpose chatbot, and 0:15:18.200 --> 0:15:20.480 it can like read the internet and sort of say 0:15:20.520 --> 0:15:22.760 what people will say, Like you can make like the 0:15:22.800 --> 0:15:24.840 best guess that like what people would tell you if 0:15:24.880 --> 0:15:27.920 you ask what stocks will go up? Bridget is trained 0:15:27.960 --> 0:15:30.680 a little bit more on like financial information, and so 0:15:30.760 --> 0:15:33.400 it might have a better idea what stocks will go up. Yeah, 0:15:33.520 --> 0:15:36.240 maybe like a less good general purpose chatbot, but a 0:15:36.280 --> 0:15:39.560 good picker of stocks. But I think the point is 0:15:39.760 --> 0:15:41.680 like it's so interesting reading, Like you know, there's there's 0:15:41.680 --> 0:15:43.360 a Bloomberg News story about it, and like you read 0:15:43.360 --> 0:15:45.080 the story and it's like it's been tested a lot 0:15:45.200 --> 0:15:46.880 to make sure it gives accurate recommendations. 0:15:46.880 --> 0:15:49.280 Well, so it's been tested. 0:15:49.440 --> 0:15:51.520 I don't think it means it's been back testing. I 0:15:51.520 --> 0:15:54.000 think it means it's like if it says to you, oh, 0:15:54.280 --> 0:15:57.760 JP Morgan Research that this stock is like x price target. 0:15:58.280 --> 0:16:01.400 Those quotes are accurate, right, it's quoting information to you, 0:16:01.760 --> 0:16:04.720 but it's not. I don't think back tested depict stacks 0:16:04.720 --> 0:16:06.160 that will go up because if it were back tested 0:16:06.200 --> 0:16:09.200 depict stacks like you go up and it was successful, 0:16:10.240 --> 0:16:11.840 it would not be a chatbot. 0:16:11.400 --> 0:16:13.800 For a brokerage form, that's true, it. 0:16:13.800 --> 0:16:14.520 Would be Bridge. 0:16:14.600 --> 0:16:18.840 It would be Renaissance, yeah, or you know, it could 0:16:18.880 --> 0:16:21.440 just be the Bridget hedge fund. Yeah, and they can 0:16:21.520 --> 0:16:23.680 make a lot more money. They should back test it, though, 0:16:23.720 --> 0:16:27.600 I want to back test Bridget. I must chat GBT 0:16:27.760 --> 0:16:28.200 to back test. 0:16:28.280 --> 0:16:31.440 It's like in beta, but like we'll certainly be testing Bridget. 0:16:31.640 --> 0:16:35.560 Yeah, this did remind me of a thing that you 0:16:35.560 --> 0:16:38.920 wrote about already, But I was on vacation last week, 0:16:38.920 --> 0:16:40.360 so we didn't talk about it on the pod, but 0:16:40.800 --> 0:16:45.000 there is that chatbot ETF that just launched that's supposed 0:16:45.040 --> 0:16:51.160 to mimic buffets Stanley Drackenmiller, David Tepper. And that's interesting. 0:16:51.320 --> 0:16:53.440 I mean it's so it's basically. 0:16:53.200 --> 0:16:55.400 Yeah, by the way, I haven't seen their back tests. 0:16:55.640 --> 0:16:57.520 I would love to see their back tests as well. 0:16:57.560 --> 0:17:02.240 But basically they have this investment committee that is built 0:17:02.240 --> 0:17:06.560 on training the chatbot around investment ideas. It basically wants 0:17:06.600 --> 0:17:08.560 them to mimic these guys, and he wants them to 0:17:08.600 --> 0:17:13.480 learn their personalities and then emulate it. And that's crazy. 0:17:13.640 --> 0:17:15.280 But it's kind of similar to what we're talking about 0:17:15.280 --> 0:17:18.640 with Bridget. I mean, she's just like general financial information, 0:17:18.680 --> 0:17:21.000 whereas this is like, this is what David Tepper would doing. 0:17:20.960 --> 0:17:23.040 We should you know, like when you like want to 0:17:23.040 --> 0:17:26.000 like change your airline flight, you like go to a chatbot, 0:17:26.000 --> 0:17:29.520 and like the chatbot is like trying to replace the 0:17:29.600 --> 0:17:35.560 human customer service but like someone must do it, and 0:17:35.640 --> 0:17:37.639 like that's supposed to be been making the human customer 0:17:37.680 --> 0:17:40.360 service agent, right, and like this I think is sort 0:17:40.400 --> 0:17:42.480 of the same thing, right, It's like making the human 0:17:42.680 --> 0:17:45.320 broker who like has the relationship with you and can 0:17:45.400 --> 0:17:48.960 like give you stock tips, but like not with like 0:17:49.000 --> 0:17:51.320 one hundred percent reliability that it will pick stocks that 0:17:51.400 --> 0:17:53.480 go up, just like you know, serving the purpose of 0:17:53.560 --> 0:17:58.520 a broker. The chat bot the like LLM ETF is 0:17:58.520 --> 0:18:00.359 a little different, Like they're trying to pick stacks that 0:18:00.440 --> 0:18:02.480 go up, Like you're not chatting with the chatbot, right 0:18:02.560 --> 0:18:04.919 Like the person who runs that ETF is chatting with 0:18:04.960 --> 0:18:07.600 the chatbot and asking what stocks will go up and 0:18:07.760 --> 0:18:10.359 hoping that the chatbot gives them the right answer. 0:18:10.560 --> 0:18:11.280 That's so funny. 0:18:11.440 --> 0:18:11.800 I don't know. 0:18:11.880 --> 0:18:15.200 I'm just the product with bridget is the relationship with 0:18:15.240 --> 0:18:17.600 the customer, right, Like you want the customer to feel better, 0:18:17.840 --> 0:18:20.120 to feel like they're getting the information they call for, 0:18:20.800 --> 0:18:23.639 and that includes like stock recommendations. The product for the 0:18:23.640 --> 0:18:25.879 ETF is just the returns stain of the ATF, right 0:18:25.920 --> 0:18:28.879 Like the chatbot is all behind the scenes, and that 0:18:29.080 --> 0:18:31.960 is putting a lot more emphasis on the ability of 0:18:31.960 --> 0:18:33.600 the chatbot to actually pick stocks. 0:18:33.760 --> 0:18:37.199 Also, the goal of the ETF isn't necessarily pickstocks that 0:18:37.240 --> 0:18:39.800 will go up, because there's no guarantee that, Like you're right, 0:18:40.160 --> 0:18:44.600 Warren Buffett or Stanley Drugon Miller will actually correctly identify 0:18:44.640 --> 0:18:45.080 those stocks. 0:18:45.119 --> 0:18:48.840 Okay, but two things. One, I think the ETF would 0:18:48.920 --> 0:18:52.199 much rather the chatbot pick stocks that go up that 0:18:52.240 --> 0:18:54.800 Warren Buffett would not have picked, than that it accurately 0:18:54.840 --> 0:18:57.160 reflects Warren Buffett and pick stocks that go down. Right, 0:18:57.400 --> 0:18:59.679 Like they don't actually care about Warren. 0:18:59.440 --> 0:19:01.720 Buff You don't care about their tracking error. 0:19:01.960 --> 0:19:04.600 The tracking air is entirely theoretical, right, because you can't 0:19:04.640 --> 0:19:07.000 like find the port. I mean you can literally like 0:19:07.040 --> 0:19:09.240 a Workshire Hathway. But this is like the stacks that 0:19:09.240 --> 0:19:12.679 Warren Buffett would pick if you ask them, there's no 0:19:12.800 --> 0:19:14.080 like visible tracking error. 0:19:14.160 --> 0:19:14.880 Yeah. 0:19:14.920 --> 0:19:16.800 But then the other thing is you're right that also 0:19:16.840 --> 0:19:18.639 their goal is not really to pick stacks like you up. 0:19:18.680 --> 0:19:21.159 Their goal is really to attract investors, right, And like 0:19:21.200 --> 0:19:23.400 it's a good stick, right, it's a good pitch, Like well, 0:19:23.600 --> 0:19:25.800