WEBVTT - How Substack Creators Are Covering This Strange Markets Era 0:00:02.720 --> 0:00:15.840 Bloomberg Audio Studios, Podcasts, Radio News. 0:00:18.480 --> 0:00:21.320 Hello and welcome to another episode of the Ad Thoughts Podcast. 0:00:21.440 --> 0:00:22.720 I'm Tracy Alloway. 0:00:22.600 --> 0:00:23.800 And I'm Joe Wisenthal. 0:00:24.079 --> 0:00:26.680 So, Joe, we continue to publish some of the conversations 0:00:26.680 --> 0:00:29.720 from our recent live show in New York, and I 0:00:29.720 --> 0:00:31.920 hope people didn't mind. They didn't seem to, they seem 0:00:31.960 --> 0:00:33.760 to enjoy it. But we did do a little bit 0:00:33.760 --> 0:00:36.280 of media industry naval gazing. 0:00:36.080 --> 0:00:37.960 Totally, so as we've been talking about it. If you 0:00:38.040 --> 0:00:40.720 listened to past episodes from the show, you know, the 0:00:40.720 --> 0:00:43.760 theme obviously had this big like sort of future of markets, 0:00:43.880 --> 0:00:47.640 future of trading theme. But you know, I would say, 0:00:47.840 --> 0:00:50.879 and maybe this is like I'm very biased here, but 0:00:51.080 --> 0:00:56.160 I would say that information dissemination collection And I guess 0:00:56.200 --> 0:00:59.000 you say, journalism is an important part of markets, right, 0:00:59.000 --> 0:00:59.920 how people get informed? 0:01:00.120 --> 0:01:02.040 Say journalism? Like that journalism? 0:01:02.200 --> 0:01:04.520 No, But you know, like the thing like in the 0:01:04.520 --> 0:01:06.800 way if you think about like the floor and the 0:01:06.880 --> 0:01:10.040 faun of the market ecology, then the people who like 0:01:10.280 --> 0:01:13.919 report the news, digest the news, explain the news, highlight 0:01:14.040 --> 0:01:17.080 what news is relevant, what is not relevant are important 0:01:17.120 --> 0:01:19.760 actors in that system and in a period of so 0:01:19.880 --> 0:01:23.280 much change and uncertainty. The question of like A, how 0:01:23.319 --> 0:01:26.800 you decide what's important and b how you convey that 0:01:26.880 --> 0:01:29.440 to an audience. These are really tough questions. 0:01:29.560 --> 0:01:32.280 Yeah, so, I think there's two major challenges here. One 0:01:32.280 --> 0:01:34.360 of them is what you just said, which is like, 0:01:34.480 --> 0:01:38.160 how do you explain these huge technological shifts to people 0:01:38.240 --> 0:01:40.840 who may not you know, they might be outside of 0:01:40.880 --> 0:01:43.480 the tech industry. How do you explain like an incremental 0:01:43.560 --> 0:01:47.200 improvement in a model to the average lay person. And 0:01:47.280 --> 0:01:51.200 then secondly the other big challenges. The media itself is 0:01:51.240 --> 0:01:54.240 caught in the crosshairs of the debate over whether AI 0:01:54.360 --> 0:01:55.840 is going to take all the jobs. 0:01:55.920 --> 0:01:56.120 Right. 0:01:56.200 --> 0:01:58.440 That's it, right, So any one of us who's in 0:01:58.520 --> 0:02:02.120 the business of trying to consume, figure out what's important 0:02:02.160 --> 0:02:05.120 and relay it is also thinking about well AI do 0:02:05.160 --> 0:02:08.120 a better job than we do, and precisely that included 0:02:08.160 --> 0:02:10.519 writing newsletters, are producing podcasts. 0:02:10.560 --> 0:02:11.000 That's right. 0:02:11.040 --> 0:02:13.760 So we had a trio of perfect guests for this 0:02:13.840 --> 0:02:17.240 particular discussion. All of them have been on the show before. 0:02:17.720 --> 0:02:19.600 We had James van Gelan. He is, of course the 0:02:19.680 --> 0:02:22.880 founder of Satrini Research and the author of the viral 0:02:23.040 --> 0:02:27.560 AI Jobs Doom scenario as well as Jasmine's son. She 0:02:27.680 --> 0:02:31.399 writes a substack that's all about AI and the culture 0:02:31.440 --> 0:02:34.840 of Silicon Valley and tech as well as samro He 0:02:34.960 --> 0:02:37.959 is the author of one of my favorite newsletters with 0:02:38.080 --> 0:02:41.360 the best name, the t kre So take a listen. 0:02:42.200 --> 0:02:43.000 Where to even begin? 0:02:43.280 --> 0:02:43.840 I'll start with. 0:02:43.840 --> 0:02:46.040 James because he's next to me and I just want 0:02:46.040 --> 0:02:47.800 to know what his life is like. Right now, do 0:02:47.880 --> 0:02:50.320 people like recognize you on the street and scream at 0:02:50.320 --> 0:02:51.680 you about AI job losses? 0:02:52.560 --> 0:02:56.639 I've done two credible death threats, but not on the street. Okay, 0:02:57.000 --> 0:03:00.239 I guess yeah, now yeah, I mean you know better 0:03:00.240 --> 0:03:04.359 when it's no person right, But now I've I luckily, 0:03:04.440 --> 0:03:07.760 I have been very much an internet personality and not someone 0:03:07.800 --> 0:03:10.480 that does a lot of taped interviews. 0:03:09.880 --> 0:03:10.280 Like this. 0:03:14.639 --> 0:03:17.880 Percent of you. No, But in case if you randomly don't, 0:03:17.919 --> 0:03:20.440 I assume you do. When was it that February or 0:03:20.480 --> 0:03:24.520 something you wrote a post about a theoretical possibility of 0:03:24.639 --> 0:03:27.040 you know, we're all going to lose our jobs but anyway, 0:03:27.680 --> 0:03:31.160 so I could see why people were upset about that. Jasmine, 0:03:31.200 --> 0:03:33.880 you write about AI and you sort of like try 0:03:33.880 --> 0:03:36.640 to when I read you're writing it's like bridging a 0:03:36.720 --> 0:03:39.760 gap right to some extent, because there are people, most 0:03:39.760 --> 0:03:42.120 of them in San Francisco, often on a lot of 0:03:42.160 --> 0:03:45.440 drugs or maybe they're on an occult or something like that, 0:03:45.680 --> 0:03:47.720 and there's like these are like the people who are 0:03:47.760 --> 0:03:50.080 like going to influence the rest of our lives. How 0:03:50.080 --> 0:03:52.560 do you think about the question of, like what is 0:03:52.640 --> 0:03:56.000 actually important to communicate to a broader public. 0:03:56.200 --> 0:03:57.880 Yeah, I mean it's really interesting because I think the 0:03:57.960 --> 0:04:01.240 vast majority of aimedia coming out of Silicon Valley is 0:04:01.280 --> 0:04:03.760 buy AI people for other AI people, and it serves 0:04:03.760 --> 0:04:06.000 that purpose really well, right, Like you got really in 0:04:06.080 --> 0:04:08.200 depth podcasts, you got the Door Hush podcast, you got 0:04:08.200 --> 0:04:10.880 this whole ecosystem of sub sex. One thing that I 0:04:10.960 --> 0:04:13.160 really noticed in that I think most people who listen 0:04:13.240 --> 0:04:15.760 to the comms coming out of the industry leaders will 0:04:15.760 --> 0:04:18.560 notice is they're saying all these crazy things without like 0:04:18.800 --> 0:04:21.640 four second thinking that like any normal person might hear 0:04:21.680 --> 0:04:23.320 them say it, and they're like, oh, man, why don't 0:04:23.320 --> 0:04:25.279 these people like AI very much? Just like I don't know, 0:04:25.320 --> 0:04:27.240 You've been telling them you're going to take all their 0:04:27.320 --> 0:04:29.919 jobs and kill everybody for the last several years, and 0:04:30.040 --> 0:04:32.359 like you're on the record saying this stuff, and so 0:04:32.880 --> 0:04:35.640 I think it's really interesting where it's not like crypto 0:04:35.680 --> 0:04:38.039 where only a small fraction of sort of the broad 0:04:38.080 --> 0:04:39.800 public ever deeply engaged with it. 0:04:39.880 --> 0:04:42.200 AI is something that, whether people want to or not, 0:04:42.320 --> 0:04:43.279 is impacting their lives. 0:04:43.279 --> 0:04:45.599 It's on their social feeds, their kids are using it, 0:04:45.600 --> 0:04:47.880 it's in their workplace, and so people have all these 0:04:47.960 --> 0:04:52.080 questions about the intersection of AI and politics, AI and affordability, 0:04:52.120 --> 0:04:55.560 AI and parenting and education, and the majority of I 0:04:55.640 --> 0:04:58.200 think AI media historically has not really focused on that. 0:04:58.240 --> 0:05:00.560 It's been more of the business and technology community. And 0:05:00.600 --> 0:05:03.039 so I'm sort of trying to say, like, okay, but 0:05:03.080 --> 0:05:03.640 what about the. 0:05:03.560 --> 0:05:04.120 Rest of us? 0:05:04.640 --> 0:05:07.440 Sam, you write a newsletter and the tag line is basically, 0:05:07.640 --> 0:05:11.000 you know, stock markets usually go up over time. You've 0:05:11.240 --> 0:05:13.719 you've been very right for the past couple of years 0:05:13.760 --> 0:05:17.280 because they've certainly gone up, But like, are things starting 0:05:17.279 --> 0:05:20.159 to get uncomfortable for you when people are talking about 0:05:20.200 --> 0:05:21.880 AI bubbles and overvaluations? 0:05:21.880 --> 0:05:22.000 Oh? 0:05:22.040 --> 0:05:25.560 Yeah, things are always uncomfortable for me. I think that's 0:05:25.560 --> 0:05:28.039 sort of like one of the most important things about 0:05:28.120 --> 0:05:30.960 understanding this idea of stocks UJ go up is the 0:05:31.040 --> 0:05:35.240 key word there is usually right, Like every way you 0:05:35.320 --> 0:05:38.560 cut the data historically, even good times and bad times 0:05:38.560 --> 0:05:40.919 and bull markets and bear markets, you always have like 0:05:41.000 --> 0:05:44.440 these periods of volatility where you know, stocks do go down, 0:05:44.839 --> 0:05:47.000 like that's the that's the catch share with stocks uj 0:05:47.160 --> 0:05:50.440 go up is that they go down a lot often. So, yeah, 0:05:50.480 --> 0:05:54.200 of course I'm nervous, especially when you're at at peaks. 0:05:54.440 --> 0:05:56.200 I mean, of course, the data will also tell you 0:05:56.279 --> 0:05:59.400 that the you know, twelve month returns after all time 0:05:59.440 --> 0:06:02.080 highsac tend to be higher than when you're at low's. 0:06:02.400 --> 0:06:04.560 But yeah, of course I'm nervous. I mean, I think 0:06:04.560 --> 0:06:07.479 that's like very healthy for anybody in the investor class. 0:06:08.120 --> 0:06:11.080 Is even no matter how optimistic you might be over 0:06:11.120 --> 0:06:13.039 the near term or long term, you have to be 0:06:13.080 --> 0:06:15.600 prepared for those big drawdowns because they do happen, and 0:06:15.600 --> 0:06:16.200 they happen. 0:06:16.000 --> 0:06:20.120 Frequently, Jasmine. You know, so James wrote his piece about 0:06:20.240 --> 0:06:22.880 you know, the potential at least and it wasn't a forecast, 0:06:22.880 --> 0:06:24.520 it was a prediction, but a scenario. 0:06:24.600 --> 0:06:28.440 We've also written others right, right, but no one. 0:06:28.400 --> 0:06:31.159 Likes the idea of like mass job laws. But I've 0:06:31.160 --> 0:06:33.320 been reading some other stuff and there's a lot of 0:06:33.360 --> 0:06:36.400 people who don't talk about AI job laws. They talk 0:06:36.440 --> 0:06:40.520 about AI will literally kill every single person in the 0:06:40.560 --> 0:06:44.120 world if the labs don't do a good job of 0:06:44.160 --> 0:06:48.640 aligning the models. How seriously do you think the public 0:06:48.680 --> 0:06:54.159 should take that specific element that if the models are misdesigned, 0:06:54.560 --> 0:06:56.760 they will destroy humanity as we know it. 0:06:57.680 --> 0:07:03.599 Ah Man contents, Oh god, they call this a pe 0:07:03.760 --> 0:07:06.800 doom out in San Francisco. I won't give you a probability, 0:07:06.839 --> 0:07:09.479 I suppose, but without getting to the question of like, 0:07:09.480 --> 0:07:12.160 will literally every single person die? I do think there 0:07:12.240 --> 0:07:15.880 is a very good reason to be concerned about misalignment 0:07:15.960 --> 0:07:18.800 and ROGUAEI, because when you think about it, there's sometimes 0:07:18.840 --> 0:07:21.480 people say that safety and the safety of a model 0:07:21.600 --> 0:07:25.440 is maybe intention with these goals like acceleration, and is 0:07:25.480 --> 0:07:26.520 the model really useful? 0:07:26.760 --> 0:07:27.840 And I think the thing that's. 0:07:27.720 --> 0:07:30.120 Really interesting is in the history of AI research is 0:07:30.160 --> 0:07:32.840 often the people who make the biggest technical advances and 0:07:32.880 --> 0:07:36.200 these foundational models who end up the misconcerned about misalignment. 0:07:36.240 --> 0:07:40.400 And that's because actually the product functionality. 0:07:39.800 --> 0:07:42.080 Of the models, their utility is only as good as 0:07:42.160 --> 0:07:44.160 how controllable they are and how much. 0:07:44.080 --> 0:07:45.200 They do what you expect. 0:07:45.240 --> 0:07:47.480 So when you say, hey, like AI agent like go 0:07:47.560 --> 0:07:49.520 off and like make me a million dollars or whatever, 0:07:49.880 --> 0:07:53.280 like whether that model goes and does so in a 0:07:53.360 --> 0:07:56.760 law abiding and safe and non violent way. The safety 0:07:56.760 --> 0:07:59.040 of the model, the alignment of the model is very 0:07:59.080 --> 0:08:02.320 correlated with whether it is economically useful as well. 0:08:02.480 --> 0:08:03.400 So I think that it's. 0:08:03.240 --> 0:08:06.480 Important that we not hold them in total tension. And 0:08:06.560 --> 0:08:08.480 I think there are really good reasons to be concerned 0:08:08.480 --> 0:08:10.080 with whether these agents are doing the things that we 0:08:10.120 --> 0:08:10.760 expect them to do. 0:08:11.280 --> 0:08:13.520 Just going back to AI job losses, and I guess 0:08:13.560 --> 0:08:16.440 any of you can answer this question, and please like 0:08:16.480 --> 0:08:19.240 feel free to chime in on every question that we ask. 0:08:19.360 --> 0:08:21.040 But you know, you also hear a lot of the 0:08:21.040 --> 0:08:26.520 big tech CEOs talk about being genuinely concerned about the 0:08:26.560 --> 0:08:30.760 future of society once AI is developed and starts like 0:08:30.960 --> 0:08:35.959 taking people's jobs. Some people interpret that as AI CEO 0:08:36.040 --> 0:08:38.800 is basically talking their own book and being you know, 0:08:38.920 --> 0:08:42.840 like hyping up the capabilities of AI. And then you 0:08:42.840 --> 0:08:45.760 have some people who argue that like maybe they're genuinely concerned, 0:08:45.800 --> 0:08:49.520 not least because they're going to get like molotov cocktails 0:08:49.559 --> 0:08:51.440 thrown at their houses and things like that. 0:08:52.120 --> 0:08:53.679 Where do you fall in that spectrum. 0:08:53.760 --> 0:08:55.840 Is this like a real concern when you talk to 0:08:56.080 --> 0:08:57.600 senior people in Silicon Valley? 0:08:57.720 --> 0:09:00.280 I do think that most of the people that I've 0:09:00.280 --> 0:09:03.520 spoken to who are worried about it are genuinely worried 0:09:03.520 --> 0:09:08.240 about it. Something that I'd add to it is pretty 0:09:08.280 --> 0:09:11.160 much unanimously throughout history, when we've had any sort of 0:09:11.160 --> 0:09:14.640 technological leap forward, it's been a positive thing. And I 0:09:14.760 --> 0:09:19.160 think that AI will mirror that. It's just the question is, 0:09:19.679 --> 0:09:22.840 we've never had a technological kind of advancement this quickly. 0:09:23.400 --> 0:09:27.120 We you know, we go from having the Industrial Revolution 0:09:27.200 --> 0:09:31.240 and the mechanization of agriculture and ninety five percent of 0:09:31.280 --> 0:09:33.960 people used to work in agriculture, and then now I 0:09:33.960 --> 0:09:36.440 think it's five percent too. And I'm very happy to 0:09:36.480 --> 0:09:41.079 work in an air conditioned office rather than in a field. 0:09:41.440 --> 0:09:44.760 A little bit. Yeah, so maybe not. But the thing 0:09:44.880 --> 0:09:48.880 is that will happen. I think believe very strongly it 0:09:48.920 --> 0:09:51.599 will happen with AI. It's just what happens in the interim. 0:09:51.760 --> 0:09:54.400 What happens if the models get so good and and 0:09:54.840 --> 0:09:57.800 can and people kind of adopt them with the same 0:09:58.480 --> 0:10:02.240 approach and use them correct over five years instead of 0:10:02.280 --> 0:10:04.520 one hundred. So I think speaking to some of the 0:10:04.559 --> 0:10:07.600 senior people, they would echo the fact that over the 0:10:07.600 --> 0:10:10.840 next thirty years, AI is going to be an immense 0:10:10.880 --> 0:10:13.880 force for good and going to make people's lives more comfortable. 0:10:14.200 --> 0:10:17.160 They're just worried about the pace of the trend. 0:10:17.320 --> 0:10:21.240 Yeah, I think the transition is like super important because 0:10:22.040 --> 0:10:25.080 as much as everyone likes this idea, like even on 0:10:25.120 --> 0:10:28.600 the corporate side or the shareholder side, like the promise 0:10:28.679 --> 0:10:32.840 of all this productivity that's unlocked by AI, and you know, 0:10:32.880 --> 0:10:35.199 maybe some people are thinking quietly in terms of like, oh, 0:10:35.240 --> 0:10:37.480 this is going to replace all of our workforce. But 0:10:37.760 --> 0:10:40.960 you know, the economy sort of stops working when no 0:10:41.000 --> 0:10:43.280 one has jobs and no has income. Right, It's like, 0:10:43.320 --> 0:10:45.480 it's great that you have this huge profit margin, but 0:10:45.559 --> 0:10:47.920 you have no revenue because everyone's unemployed and they have 0:10:48.000 --> 0:10:50.600 any money. And you know, those people aren't shopping, those 0:10:50.640 --> 0:10:53.120 people aren't buying you know, washing machines and the washing 0:10:53.120 --> 0:10:56.320 machine manufacturer kit, you know, by you know, mental parts. 0:10:56.320 --> 0:10:58.200 And suddenly we're digging up less copper, and the whole 0:10:58.240 --> 0:11:01.320 economy shrinks because no one has jobs. Then the whole 0:11:01.360 --> 0:11:05.080 point of this exercise becomes sort of meaningless, So which 0:11:05.080 --> 0:11:08.080 is why I think you're increasingly hearing from folks on 0:11:08.160 --> 0:11:10.800 this side, you know, talking about things like you know, 0:11:11.200 --> 0:11:14.040 basic income and guarantee jobs and all these things. So 0:11:14.040 --> 0:11:17.520 it's I think that's where you know, the conversation sort 0:11:17.520 --> 0:11:20.600 of goes, is, all right, let's make this assumption that 0:11:20.640 --> 0:11:23.319 we have all this productivity because of job loss. Well, 0:11:23.880 --> 0:11:24.600 that doesn't work. 0:11:24.720 --> 0:11:24.920 You know. 0:11:24.960 --> 0:11:27.000 We people need to have money to go out and 0:11:27.000 --> 0:11:30.320 spend and buy these products that are being produced so efficiently. 0:11:30.800 --> 0:11:32.800 And I mean it's worth noting that, like for example, 0:11:32.840 --> 0:11:35.760 like people politicians will point to say, retraining is the default. 0:11:35.760 --> 0:11:38.720 We have never had a successful large scale reskilling program 0:11:38.760 --> 0:11:41.360 in the history of the United States. During say, you know, 0:11:41.400 --> 0:11:44.400 the de industrialization. You lost all these factory jobs. It 0:11:44.440 --> 0:11:46.959 wasn't that many jobs, and we created way more net jobs, 0:11:47.000 --> 0:11:49.160 say in Silicon Valley, but the steel workers did not 0:11:49.240 --> 0:11:51.640 learn to code, right, And so people are right to 0:11:51.640 --> 0:11:54.800 be concerned, both morally and about the political backlash that 0:11:54.840 --> 0:12:15.040 could occur even with relatively small numbers of net job loss. 0:12:15.080 --> 0:12:17.920 Sam like, I love your newsletter, I read it and 0:12:18.120 --> 0:12:23.000 paid subscribe. Where all that I think I could train 0:12:23.120 --> 0:12:26.080 a model to write a newsletter that say stocks go up? Yeah, 0:12:26.120 --> 0:12:28.320 I think I could, like I believe or not? 0:12:28.520 --> 0:12:33.120 My uh my, my sister. My sister recently got into 0:12:33.120 --> 0:12:36.680 claude coding and she did this project with herself and 0:12:36.679 --> 0:12:38.600 and thought that, you know, I find this really interesting. 0:12:38.640 --> 0:12:41.640 And basically she created this app that was like a 0:12:41.720 --> 0:12:46.199 sam robot that was informed by you know, stuff that 0:12:46.240 --> 0:12:48.760 I've published and like by by bilin all this free 0:12:48.760 --> 0:12:51.280 stuff that's out there, and it's like, I was kind 0:12:51.280 --> 0:12:54.400 of offended first of all, but yeah, it's a problem. 0:12:54.559 --> 0:12:57.480 Do you worry about it? Is it absolute proprietor of 0:12:57.520 --> 0:12:59.960 a yeah, the company, you know, it's one. 0:13:00.160 --> 0:13:02.520 It's like there was this time where you could go 0:13:02.559 --> 0:13:06.240 out and say, well, you know, something that you can't 0:13:06.440 --> 0:13:10.199 really replace is you know, the individual's voice, the personality 0:13:10.280 --> 0:13:12.800 or whatever. And it's like I, you know, I run 0:13:12.840 --> 0:13:14.439 some of this stuff too, and run these queries and 0:13:14.480 --> 0:13:16.680 say like, well, how would Samua write this? And it's 0:13:16.760 --> 0:13:20.600 like it sounds like me? And sometimes it actually you know, 0:13:20.760 --> 0:13:23.200 uses the language and comes up with the words faster 0:13:23.600 --> 0:13:26.280 than how I would. So it's like, you know, what 0:13:26.360 --> 0:13:28.640 I do has to be about something more than just 0:13:29.000 --> 0:13:33.200 having the samual voice right, so to answer the question 0:13:33.320 --> 0:13:35.200 like the problem and the challenge and I you know, 0:13:36.000 --> 0:13:38.240 James and Jasmine and we talked about this a little 0:13:38.240 --> 0:13:40.720 bit in the back is like, you know, where do 0:13:40.800 --> 0:13:43.000 you add that value? And it's like it's it can't 0:13:43.000 --> 0:13:45.560 be something that's like just sort of replicable or or 0:13:45.600 --> 0:13:49.200 something that you know, someone who might be interested in 0:13:49.240 --> 0:13:51.920 what you're reading or writing about can put into a 0:13:52.000 --> 0:13:55.640 query and get back like, you know, the toughest part 0:13:55.800 --> 0:13:57.840 about my job, and I think a lot of our 0:13:57.960 --> 0:14:01.480 jobs is being able to come up with those ideas 0:14:01.600 --> 0:14:05.120 and those angles that no one's asking for. Yeah, and 0:14:05.400 --> 0:14:07.560 you know, again like I write about the stock market 0:14:07.640 --> 0:14:09.920 usually going up, which, by the way, even before AI, 0:14:10.040 --> 0:14:11.960 it's kind of a ridiculous thing to be writing about 0:14:12.160 --> 0:14:15.200 because everyone who has money in their fore one ks 0:14:15.240 --> 0:14:17.640 and iras, I've already been told this. They already know this. 0:14:18.000 --> 0:14:21.400 So it's like why do I exist? And you know, 0:14:21.440 --> 0:14:24.000 it's a relatively small market of people who are reading 0:14:24.000 --> 0:14:26.920 what I'm writing about, But yeah, I guess there's something 0:14:26.960 --> 0:14:29.400 about And not to toot my own hornets, this is 0:14:29.480 --> 0:14:32.440 reader feedback that I'm getting, but I've been told that 0:14:32.560 --> 0:14:36.200 there's something about whether it's the cadence or the timing 0:14:36.600 --> 0:14:38.760 of things I do write about, or the way I 0:14:38.760 --> 0:14:41.800 frame it that feels a little bit more human than 0:14:42.600 --> 0:14:45.160 the research or the analysis that you know, the chief 0:14:45.160 --> 0:14:47.800 investment strategists or whatever is sending out to their financial 0:14:47.800 --> 0:14:49.960 advisor who's regurgitating it back to them. 0:14:50.320 --> 0:14:52.560 James, one of the things you've been doing that I 0:14:52.600 --> 0:14:56.160 don't think a bot can replicate just yet is sending 0:14:56.240 --> 0:14:59.080 actual human beings or a human being to the strait 0:14:59.080 --> 0:15:03.080 of hormone. Is that like the edge in media? Is 0:15:03.120 --> 0:15:06.640 it like first person experiments and data gathering? 0:15:06.800 --> 0:15:09.200 Yeah, I mean you just have to look at this 0:15:09.280 --> 0:15:12.080 from I'm pretty confident that investment research won't exist in 0:15:12.120 --> 0:15:14.440 the same way that it does now. The Internet kind 0:15:14.480 --> 0:15:18.160 of made it democratized, so to speak. The distribution. It 0:15:18.240 --> 0:15:21.160 was like the only the banks had distribution. Now we 0:15:21.240 --> 0:15:24.000 have distribution and we can utilize that. But if we're 0:15:24.000 --> 0:15:26.160 just utilizing it to kind of do the same thing 0:15:26.480 --> 0:15:29.680 that the banks are doing, well, there is so much 0:15:29.680 --> 0:15:32.600 investment research that these models have been trained on, and 0:15:32.720 --> 0:15:34.600 as they get better, they will pretty much be able 0:15:34.640 --> 0:15:37.640 to do the same thing. So it becomes very much 0:15:37.640 --> 0:15:40.800 about doing something different and I think that there's like 0:15:40.840 --> 0:15:43.440 two ways you can do that. One you can be 0:15:43.560 --> 0:15:47.480 right all the time, so hopefully you know fingers crossed well, 0:15:47.760 --> 0:15:50.960 you know, I think that that one aspect is like 0:15:50.960 --> 0:15:54.080 like when I came on Ale's was the only media 0:15:54.120 --> 0:15:56.080 that I did after that piece, which I was very 0:15:56.120 --> 0:15:58.640 happy to thank you. I said, one of your questions 0:15:58.680 --> 0:16:01.240 show was like, why do you think this so viral? 0:16:01.760 --> 0:16:04.720 And I think it's like, this is not like tooting 0:16:04.720 --> 0:16:06.320 my own horn. It's just sometimes you go on a 0:16:06.320 --> 0:16:08.560 hot streak. And when we started that was very much 0:16:08.600 --> 0:16:11.280 in like a year of just like one hundred percent 0:16:11.360 --> 0:16:13.480 correct calls, and you know, obviously I've had a bunch 0:16:13.520 --> 0:16:15.480 of wrong ones since then. But it was like a 0:16:15.520 --> 0:16:17.560 bunch of people were reading something and they would read 0:16:17.600 --> 0:16:19.680 it and they would say, well, if I hadn't read 0:16:19.720 --> 0:16:21.480 this piece, I wouldn't have bought in video in twenty 0:16:21.480 --> 0:16:24.120 twenty three, or wouldn't have bought Eskehinich something. So I 0:16:24.480 --> 0:16:27.960 generated value from it. And then if you can add 0:16:27.960 --> 0:16:31.040 into that something that the model can't do with a 0:16:31.120 --> 0:16:33.640 query they can't be out in the real world, then 0:16:33.760 --> 0:16:35.840 it's going on in the real world. If there's if 0:16:35.920 --> 0:16:37.680 you know in five years, the robots come and then 0:16:37.720 --> 0:16:39.120 the models can go out in the real world. We'll 0:16:39.120 --> 0:16:40.600 have to find something else that they can't do. 0:16:40.840 --> 0:16:43.520 Judgment, How do you think about this question of like 0:16:43.920 --> 0:16:47.120 what the human writer can still bring to the table. 0:16:47.240 --> 0:16:49.280 The one time you were on Autolaus before was nothing 0:16:49.280 --> 0:16:52.880 about you were talking about Chinese peptize. 0:16:53.280 --> 0:16:56.400 But you did go to a peptid rate now, yeah. 0:16:56.360 --> 0:16:58.320 So like that is sort of your equivalent of sending 0:16:58.360 --> 0:16:59.920 an analyst to the Strait of Homers. 0:17:00.800 --> 0:17:02.760 We were talking about this, I would be the San 0:17:02.800 --> 0:17:06.120 Francisco spy that Jay sets up. I guess I'm I've 0:17:06.160 --> 0:17:08.840 doxed myself already, but I can get away. 0:17:09.000 --> 0:17:10.600 But is that like, do you feel like that's a 0:17:10.640 --> 0:17:14.040 big like yes, margat that like being in it in 0:17:14.160 --> 0:17:16.399 some way physically in it is that it is like 0:17:16.520 --> 0:17:19.439 what you can do at least currently that the AI can. 0:17:19.640 --> 0:17:22.560 Basically, I mean, I'm very bullish on secrets and I'm 0:17:22.640 --> 0:17:24.560 very bullish on gossip, right and and. 0:17:25.920 --> 0:17:28.000 I mean that's kind of what journalism is in the end. 0:17:28.320 --> 0:17:30.400 It's you have what is not in the trading data, 0:17:30.480 --> 0:17:32.280 Like data is like one of the most valuable things. 0:17:32.280 --> 0:17:35.360 It's like chips, data algorithms, right, and the data's out 0:17:35.359 --> 0:17:37.520 there in the world, and they've scraped the entire Internet. 0:17:37.560 --> 0:17:39.200 They've torn the covers off all these books. 0:17:39.359 --> 0:17:41.600 Their markre is like paying hundreds of dollars for these 0:17:41.640 --> 0:17:43.760 experts to like feed all their knowledge into the models. 0:17:43.880 --> 0:17:46.520 But the thing is, the world's constantly changing. It's very dynamic. 0:17:46.680 --> 0:17:49.560 They're parts of the world, parties u straits that are 0:17:49.640 --> 0:17:51.840 unexplored by humans. And if you go and you are 0:17:51.840 --> 0:17:53.720 the first person to see that, or you're the first 0:17:53.720 --> 0:17:55.720 person to be there at a critical moment in time 0:17:55.920 --> 0:17:58.440 as news is happening, as news is breaking, that kind 0:17:58.440 --> 0:18:01.720 of actual reporting is like more valuable than ever. One 0:18:01.760 --> 0:18:03.919 way that I think about reporting is that's you have 0:18:04.040 --> 0:18:06.480 this private knowledge that's maybe known in whisper networks, in 0:18:06.520 --> 0:18:09.119 the you know, in some party scene or whatever, or 0:18:09.160 --> 0:18:11.760 you have tacit knowledge that nobody's really written down yet. 0:18:11.840 --> 0:18:14.280 And as a reporter, my job is to take the taseit, 0:18:14.400 --> 0:18:16.280 or to take the private and to turn it into 0:18:16.280 --> 0:18:19.520 public knowledge. Like that particular task I think is extremely 0:18:19.600 --> 0:18:20.040 robust in. 0:18:20.040 --> 0:18:20.640 The era of AI. 0:18:20.960 --> 0:18:23.720 Okay, so speaking of secrets and gossip and going out 0:18:23.760 --> 0:18:26.359 into the real world, you recently came back from a 0:18:26.400 --> 0:18:29.000 trip to China, right where you were sort of comparing 0:18:29.119 --> 0:18:32.520 I guess how China is thinking and undertaking AI versus 0:18:32.560 --> 0:18:34.400 the US. What's your big takeaway? 0:18:34.480 --> 0:18:37.000 Oh yeah, I mean China is one of those places 0:18:37.080 --> 0:18:40.159 where it's hard to get a sense of without going 0:18:40.200 --> 0:18:42.120 in person, because it's so restructed. 0:18:42.160 --> 0:18:44.840 What's on the public Internet, and we have these off internets. 0:18:45.680 --> 0:18:47.639 One interesting thing is the way that I think about 0:18:48.200 --> 0:18:50.800 USAI research is it sort of had three eras of 0:18:50.840 --> 0:18:53.520 American AI. You had the academic era, maybe you'd call 0:18:53.560 --> 0:18:55.879 the boom kicking off with image net or alphag. You 0:18:55.960 --> 0:18:58.560 had the commercial era, let's call it kicked off with CHATGBT, 0:18:59.000 --> 0:19:02.040 you know, the maybe geopolitical era that's sort of Pentagon 0:19:02.040 --> 0:19:05.120 in Mythos drama. And China is weirdly still in the 0:19:05.160 --> 0:19:08.720 academic era. More so, it's very collaborative, very pro open source. 0:19:09.200 --> 0:19:12.760 The labs and the researchers themselves seem unconcerned pretty much 0:19:12.760 --> 0:19:13.320 with the sort of. 0:19:13.240 --> 0:19:16.840 Big philosophical and geopolitical questions, and that's there's sort of. 0:19:16.800 --> 0:19:19.960