Loading summary
Transcript116 lines
- [00:00]
A
Everyone has 10 year AGI timelines right now. It started when Sam Altman put out that post. Like, superintelligence is just a few thousand days away. And at the time it was kind of odd because like when he wrote that post last year, everyone was like, AGI is one year away. AGI is two years away. It was like fast takeoff time. Like, everyone was very excited. And then he came out and was like, it's a few thousand days away. And so Sam came out and, you know, and he was kind of in his blog post of the superintelligence age, talking about, you know, maybe we're a decade away. Then Andrej Karpath, he goes on Dwarkesh just a couple months ago, a couple weeks ago, says AGI is a decade away. And then Dwarkesh posts this, this probability density of, of when AGI will be achieved. There's a chance that America does it, there's a chance that China does it. And the median, the 50, the 50th percentile, was exactly 2035.
- [00:48]
B
Then you know that, that, that chart, that like very schizo chart that says periods when to make money. Have you seen this?
- [00:56]
A
No. No.
- [00:57]
B
And it's like it was created like, I guess about 100 years ago. People reference it any time it like actually aligns to events because it's basically has years in which panics have occurred, years of good times, high prices, and time to sell stocks, and years of hard times, low prices, and a good time to buy stocks. And so it's basically like astrology for stock picking.
- [01:21]
A
And what is it saying right now?
- [01:22]
B
2035 is a year is they're predicting is when a panic will occur.
- [01:27]
A
Oh, interesting. Well, that certainly aligns with all these AGI timelines.
- [01:30]
B
There you go.
- [01:30]
A
And then I was looking at meter and this one we'll have to debate a lot more, but meter has been tracking AI's ability to complete long tasks and it's growing exponentially. It used to be like six seconds. Now it's like two hours. And you know, when you talk to anyone who's in the AI field, they'll tell you that the agents are getting more and more capable of handling longer, longer time horizon tasks. The question is, I feel like humans don't have a time horizon. I feel like humans, they're just born and the goal is like, survive, be fruitful, multiply. Right? And so I feel like if you're tracking the meter data, you need to get out to like 30 years, like a full career. Right? Like the prompt needs to Be like, go make money. And then it just goes and becomes a lawyer and you know, lives its full life and retires after, after a 35 year run. And of course when you track out the doublings one in 2030, the meter is projecting based on that log curve or that log, that log graph, that AI will be able to have a time horizon in the decades. My read on the meter data is that AGI 2035, again maybe the messiest, the least definitive. But what's interesting is that it just feels like 10 years is this consensus right now. And there's much less diversity of opinion. There aren't that many people saying two years anymore. There aren't that many people saying 50 years anymore. Everyone's kind of saying 10 years. And I just wonder like let's put aside like let's try and accurately predict when this thing happens and let's just analyze it from a psychological perspective and like what does it mean when the tech community all has a consensus of something that's a decade, Like a decade away could just be what people say when they don't know. Like if you ask me when flying cars are going to happen, I'm going to say a decade. If you're going to say when quantum computing, oh, that'll be a decade. Oh Mars, yeah, that's a decade.
- [03:31]
B
Most people under overestimate what they can accomplish in one year, underestimate what they do what they can do in 10 years. And in this situation we're just like, yeah, we're estimating that we can just achieve AGI in 10 years.
- [03:44]
A
When I personally feel like about AGI, like I'm convinced by these, like I feel like AGI is 10 years away and I don't feel like I'm like coping or doing some sort of like mental logic jumps or something. It's just like if you actually force me to put a prediction down, I probably would say about a decade. Then I go to, well like how should I actually be changing my behavior? Like if something big is coming in a decade, it feels like you should actually should not just be acting normally. It feels like there's some sort of preference falsification going on. Like how many, everyone's saying a decade. How many people are actually acting like it's a decade? Like what should you be doing in the intervening years if it's a decade away? Like are you just supposed to like build technologies that are fun little dopamine rewards? Are you trying to like accumulate as much capital as possible before who plans.
- [04:34]
B
Around 10 years at all right. People tend to go for things that they want today.
- [04:40]
A
Yeah.
- [04:41]
B
In some ways. Right. So somebody like, let's say they're in their 20s, they're like, I want to own a home by the time, by 2035, they want that thing today. And so they, they. But they might understand. Okay. It's going to take some time to get there.
- [04:54]
A
Yeah.
- [04:54]
B
But actually making real changes in your life for this, like, impending scenario, that's hard to predict entirely. It's very, very difficult. And I don't know that, you know, it was maybe more popular earlier this year to joke about, you know. Know what? You know, the golden retriever maxing.
- [05:15]
C
Right.
- [05:15]
B
Oh, yeah, we, we would talk about this. Right. But it feels like that dialogue has kind of changed.
- [05:21]
A
Tyler, what. What do you think about my take? What do you think? What do you agree with? What do you disagree with?
- [05:27]
C
Yeah, I, I think it's definitely interesting that people seem to kind of align around this 10 year thing. I think there's also some sort of bias. Right. If you think AI is coming in three years, you're probably not just going to be like, writing blogs. Like, you're, maybe you're going to start a macro hedge fund. Maybe you're going to go work at one of the labs to, like, really try to influence how it's going to happen. If it's going to happen super quick.
- [05:46]
A
Yeah.
- [05:46]
C
So I think there's some sense of that. And if you really don't believe AGI is coming at all, then, like, you're also probably not going to be writing blogs about AGI. You're just going to be like, yeah, doing your normal job or whatever. So I think there's some, like, confirmation bias there.
- [06:00]
B
Tyler is probably the most AGI pilled person on the team. And so it's interesting that you're. You kind of day to day act like everyone else on the team.
- [06:10]
A
Yeah. What's your biggest revealed preference?
- [06:14]
C
That's hard to say.
- [06:15]
A
Maybe just hanging out on the show.
- [06:17]
B
Yeah, just hanging out. Podcasting instead of going to university. He's just like, yeah, nothing matters.
- [06:21]
A
This clip of Alex Wang that's sort of going viral, people are dunking on it because he recommends that kids learn to vibe code. And I just disagree with the haters. Like, I, I think the haters are wrong on this one. And I was thinking about, like, well, I have a kid, like, would I teach him to vibe code? And I was playing with Legos last night, I was assembling Legos. And from what I've done, Vibe Coding and what I've done Legos. I'm like, this is going to be very fun and this is going to be an activity that we do together. And I'm super. I'm like, super in agreement that I think learning to Vibe code is good. What do you think, Tom?
- [06:54]
C
I think like, the point of view of like the people hating were that like, oh, your 13 year old should be like making Beats B2B Sass. Which is like, not. That's different than saying they should be vibe coding because Vibe coding is just like, it's like basically playing a video game where it's like making a Minecraft mod or something that's like, seems totally fine.
- [07:10]
B
Yeah. Some of the criticism, some of the criticism was like, you should just be doing really hard things. You have to ask is like, okay, should 13 year olds, like only be hand coding? Like if you want to get them into maybe making. Doing engineering work, is it. Should they be doing it with a pen and like, what. What is the alternative? When I was 13, I was working on little iPhone apps.
- [07:32]
A
Yeah.
- [07:33]
B
And knowing the vibe coding tools that are available today, I would have been able to make way more progress. I would have had a lot more fun.
- [07:41]
A
Yeah.
- [07:41]
B
It would have been like having an expert, like software engineer sitting next to me. Kind of like pair programming, right?
- [07:48]
A
Yeah.
- [07:48]
B
And so I'm on your side. I think. I think people generally just, just don't like Alex because he's been wildly successful.
- [07:56]
A
He's basically the youngest, most successful person in the world. With Alex Wang, it's like, yeah, he's rich. And I have no idea what the thing he. I don't interact with it at all.
- [08:06]
C
Fun fact. So in fifth grade in the yearbook, they asked like, what you want to be. I put investor.
- [08:11]
A
Investor in fifth grade.
- [08:13]
C
Yeah. I don't even really know what that meant.
- [08:16]
A
SoftBank's big profits jump this quarter came from OpenAI's increased valuation that SoftBank lifted higher by buying shares from employees and selling them at a higher price. And so there is a question about, is this too circular? It's unclear if SoftBank was really the price setter on this deal. But just Dario is certainly just Dario.
- [08:39]
B
Saying SoftBank books a capital gain on an investment it hasn't paid for and recorded in its assets that there's a scenario where you would do this if you were a company that invested in a fundamental.
- [08:50]
A
Yeah.
- [08:50]
B
And it's called maybe some capital, but not all of it yet. And so you can show that there's a gain. Yes. You haven't paid actually paid in the capital yet? Yes, but I think the way in which SoftBank is doing VC firms doesn't seem illegal, but I think it's non traditional. Sophie over on X says SoftBank is selling its Nvidia stake to fund companies whose main expense is buying from Nvidia. The more important thing here is that this is not the first time Masa has exited Nvidia. He exited in 2019 before the run up. He was then Nvidia's largest shareholder and he had a 5% stake in the company that he sold for 3.6 billion and it would now be worth over 200 billion.
- [09:39]
A
Are people mad about this because they think Nvidia is going to rip further?
- [09:43]
B
Or did selling your stake in Nvidia?
- [09:44]
A
Yeah, yeah.
- [09:45]
B
So basically SoftBank is booking profits.
- [09:49]
A
Sure.
- [09:49]
B
OpenAI profits. They are then selling Nvidia shares to fund the original investment of which they've already booked the profits on. And the number one driving force behind Nvidia's growth is OpenAI.
- [10:01]
A
Today we will get the SoftBank completely sold out of Nvidia. Fear from bears on mainstream media and Wall Street. Does anyone do objective research anymore? Softbank initially bought its Nvidia stake through the vision fund in 20, then exited completely in January 19th. They missed it. They lost patience. Not the first time. Yeah. When did they buy this latest slug? That's the question. SoftBank sold 32 million shares of Nvidia in October. Also sold part of its stake in T Mobile for 9 billion. It's not the first time they've cashed out of the chip maker. That's a funny way to put it. Despite the sale, SoftBank remains tied to Nvidia through its other ventures. Yeah, that makes sense. Warren Buffett says he's going quiet. The world's most famous investor warns against corporate greed as he prepares to hand over the reins of Berkshire Hathaway. The 95 year old will step back from day to day responsibilities at Berkshire at the end of this year when he retires from his role as chief executive. Craig Abel, of course, is coming in. Investors have long seen Buffett, often called the Oracle of Omaha, as a corporate folk hero, interspersing guidance on his portfolio company's performance with life and business advice. He noted, for instance, that requirements for executive compensation disclosures backfired as business chiefs engaged in a race to earn more than rivals. What often bothers very wealthy CEOs they are human after all, is that other CEOs are getting even richer. Buffett said envy and greed walk hand in hand. Buffett added that Berkshire should try to avoid future CEOs who are looking to retire at 65 or who want to become look at me rich or initiate a dynasty. Look at me rich.
- [11:38]
B
Buffett has never been photographed doing a money spread wearing Balenciaga Rick Owens.
- [11:44]
A
Yeah, so he took the pledge to give away, to give away his wealth. And so it's going to be a pretty big changing of the guard. There's that interesting stat that like, if you go back 30 years ago when Buffett was 65, he was, I don't think he was close to the richest man in the world. He, he compounded a ton in his third 30 year run, like from 65 to 95, he had a particularly good run that took him from pretty rich to one of the richest people in the world. It's very rare. We just don't see that many business executives or that many people broadly where that's like the highlight of the career is 65 to 95. Buffett has commissioned no plays, no poems, no symphonies, no operas, no ballets, funded no paintings or sculptures that will outlive him, endowed no theaters, choirs, orchestras, built no monuments, monasteries. Is this true at all? He's committed to giving away half of his wealth. So, like some of the money is going to wind up with the opera. I imagine just because it's going to go out and get diffused amongst all the different charity efforts, it's not going to all go into one thing. Howard agrees with him on this. He says he's right that greatness does not come about through accumulating great amounts of money, which is what he is known for. But beyond that, it's fortune cookie level advice. It's not wrong, but that's about it. His partner Charlie Munger's contribution to architecture was to fund factory like college dorms in which a majority of the apartments don't have windows. Was that ucsb? Where is that?
- [13:21]
B
Yeah. Have you been to that dorm you made the donation?
- [13:24]
A
Yeah, yeah, yeah.
- [13:25]
B
There's some reference to it on the campus, but they abandoned the plans to build a windowless dorm.
- [13:31]
A
So saying it's so Soviet. It's extremely Soviet. I don't know why they wanted.
- [13:39]
C
I thought there were fake windows and they were like TVs essentially, but they would show like outdoor scenes.
- [13:44]
A
It should be vertical TV windows that you can just put Sora on. Is it safe to take the other side of this and just say that like he was about to create the greatest lock in of all?
- [13:53]
B
I mean, what if he was Basically saying, like college students spend very little time in their actual dorms because they're out and about in the world. They're studying the library, they're at events.
- [14:02]
A
Totally.
- [14:03]
B
They're really just using the dorm to sleep. Why don't we create common areas that have windows, an outdoor area that then you go in your pod.
- [14:09]
A
This is. Eric Adams floated a similar idea to Munger calling for stripping legislation that, that promises each citizen, each city citizen window access this past March. You don't need no window where you're sleeping. It should be dark. He said the $2,000 dividend could come in lots of forms. Let's play the clip. The $2,000 dividend could come in lots of forms in lots of ways. George. You know, it could be just the tax decreases that we are seeing on the President's agenda. You know, no tax on tips, no tax on overtime, no tax on deductibility of auto loans. So, you know, those are substantial deductions that are being financed in the tax bill. Tyler, do you feel like this, how.
- [14:58]
B
How is that going to help the day trading community?
- [15:00]
C
Yeah, I was like doing rampant speculation with this $2,000. What am I going to do now?
- [15:06]
B
You're actually. You're basically fully. You've already committed the funds, right?
- [15:10]
C
Yeah, yeah, I've already placed a ton of parlays.
- [15:13]
A
Yes.
- [15:14]
C
So like what am I supposed to do?
- [15:16]
A
I can't pay my parlay bill with a tax decrease on auto loans.
- [15:22]
B
Yeah, you're not looking for passive income. You're looking for massive income, bro.
- [15:26]
C
I'm looking for massive income.
- [15:28]
A
They should put a button on the IRS website that allows you to either receive the $200 stimmy check or if you press the button, you have a 50, 50 chance of getting either $4,000 or zero. I really think a lot of people would press that button. Two of the world's biggest data center developers have projects in Nvidia's hometown that may sit empty. The local utility isn't ready to supply electricity. 48 megawatts. What are we doing? That's. That's child.
- [15:57]
B
The data center for ants. Calm down, people. Mas's son is a male Cathie Wood. Him selling Nvidia today means Nvidia has plenty of upside left.
- [16:07]
A
It is.
- [16:07]
B
It is. I mean, it just comes down to he will take. He's selling Nvidia, he's going to give that money to OpenAI and then OpenAI is going to give it to Nvidia.
- [16:15]
A
So sold.
- [16:17]
B
Yann Lecun. He's out, says see ya.
- [16:20]
A
He's out.
- [16:21]
B
He's out from Meta. He's going to launch his own startup. Tyler, what's your reaction?
- [16:28]
C
I mean, this is pretty big news. I think people have kind of expected this for a while.
- [16:31]
A
Yeah.
- [16:32]
C
Basically since Alexander Wang came in. Because Alexander Wang kind of in some sense like took over as like the chief AI person at Meta, which was.
- [16:41]
A
Not enough room in this town for two AI.
- [16:44]
C
Not enough room for chief scientists.
- [16:47]
A
Yeah, yeah, yeah. Not enough room in this town for two chief AI scientists.
- [16:50]
C
And so Jan is like very goaded AI researcher. He's been in the game for like super long today. He's like very much seen as being bearish on LLM.
- [17:00]
A
Sure.
- [17:00]
C
He doesn't think they can reason. He doesn't think they can like do novel tasks or whatever. So he's been like very outspoken about that. I think that's maybe part of the reason why he's, he just has kind of a different vision from, from Zuck and Alexander Wang. I forget exactly what it's called, but he has like another kind of path that he sees to AGI. He's, he's, he's still a professor at nyu, which I think is where he does most of his research these days. So I think he'll probably just stay in academia.
- [17:26]
B
Does he have any active classes going on this semester? Because if he does, we should send you onto the campus to study.
- [17:34]
A
Do we have an update from Brian Johnson? He came back from his trip. He was on psychedelics. I feel like I'm expecting him to fall in love with at least one fast food restaurant. I want him to flip around on at least one and say, okay, yeah, I'm an in and out guy now. He's been quiet. I do wonder if he'll update on anything or if he's just like so, so powered through that it just will not affect him at all. Because like he was kind of framing it as like, this is like a big deal. It seems like he got through it just super fine.
- [18:15]
B
Saw a very viral post of somebody saying, I don't think it's a good thing that billionaires can talk about taking Schedule 1 drugs publicly with no repercussions. It did seem that he was doing like this seems like such, still such a gray area where you can get it prescribed by a doctor, but it's still schedule one. He posted an update on, on his experience. He also talked about more about why he says there's potential. It's potentially a longevity therapy for psilocybin. Expands lifespan in mice. It can reduce inflation markers tied to aging, it can increase brain entropy, breaks rigid patterns and boost long term cognition and flexibility.
- [19:03]
A
And yeah, I don't know, I kind of agree with that, with that person who is saying like, they shouldn't be able to talk about it publicly. I feel like he should have a disclaimer or something because there's going to be, there's people, he's such a huge audience. There's going to be some people that just listen.
- [19:22]
B
In the early podcast era, you had Tim Ferriss, who to his credit was like doing work on himself and would share what he was doing, but he would talk about doing things like ibogaine, which is like a cra, you know, super powerful psychedelic. And he wasn't directly saying, hey, I think you should go take this. But when hundreds of thousands of people like follow Tim for health advice, it sort of is like an indirect. It's somewhat of an, it's somewhat of an endorsement, even if it's not directly so.
- [19:59]
A
No, no. I mean, I 100% think some people will see this and be like, oh, looks like he had a good experience. Let me jump straight to exactly his protocol, which is clearly not accessible for the average person. Like, he has been building up a tolerance to this particular chemical for probably a decade and so it's going to hit him very differently than it will someone who's not able to in the same place. We were joking yesterday about like, oh, if he wanted to like really challenge himself, he would have been at like a crowded concert or something. And there are going to be people that see this post and wind up actually doing that and winding wind up in a very, very rough spot.
- [20:38]
B
I did see some people responding and saying, I took a similar dose to this and basically ruined my year. I have joked in the past, it's like, are psychedelics going to fix your life or just doing the tasks that you've been avoiding? Is that going to make you feel relief? You know, is that going to make you feel better about your life?
- [20:55]
A
It's like that meme of like the anime ninja. If you're tired, do it tired. If you're sober, do it sober. Right? If you're sober, do it sober.
- [21:05]
B
2016. So almost 10 years ago says digital addiction. Digital addiction is going to be one of the great mental health crises of our time. And Blade says Unk had this thought and immediately was like, how do I profit from this? The greatest to ever do it. Tbh.
- [21:21]
A
Like, if you're, if you're trying to like Layer up. Like, the. The open AI is bad. Like, case. Like, you don't start with, like, the product's too addictive. Like, I don't think that's the claim. The claim is like, it uses too much water or like, it. It like one shots people or it does erotica or it's like AI slop. But very few people are saying, like, oh, yeah, people are on their phones too much and it's because of OpenAI. Now, people are on their phones too much, but it's because of Instagram and it's because of, like, social media, actually. Not the AI.
- [21:52]
B
Because of Sora.
- [21:53]
A
I think it's a little bit too early to say that.
- [21:56]
B
I don't think that.
- [21:57]
A
I don't think it's working.
- [21:58]
B
Certainly. Short. Short form video.
- [21:59]
A
Have you ever been in, like, the gym or in a coffee shop and seen someone scrolling Sora, like, over your shoulder? I haven't seen.
- [22:07]
B
I meant. I meant vertical video. Broadly.
- [22:09]
A
Totally. Totally. Which is, I guess. I guess, like, a good take here, which is like, how do I profit from this? I got to make an AI version of it. That's. That's probably like, more of what they were thinking in terms of actually profiting on. On, like, digital addiction. It doesn't seem like it's going very well for 4o.
- [22:28]
B
Seems to be extremely addictive.
- [22:30]
A
Yeah.
- [22:30]
B
So much so that very.
- [22:31]
A
For a very small amount of people, like. Absolutely.
- [22:34]
B
Yeah. Yeah. We don't know how many people.
- [22:36]
A
Yeah.
- [22:37]
B
And we hope you have a wonderful Veterans Day and we'll see you tomorrow.
- [22:43]
A
See you tomorrow. Goodbye.
- [22:44]
B
Cheers.