Alexandr Wang

And so I think a decade looking back, it'll be obvious that intelligence became abundant and that agency became abundant. The current trends we're on are just gonna keep continuing. And this will be very strange. I mean, I think for the history of humanity, groups of smart people getting together towards a shared goal was the bottleneck of progress. The United States of America, in some sense, was an example of this. It's like the United States of America is formed from a group of very smart people getting together and having a vision for the future that they wanted to enact. And that's the story of nearly every company in America and the story of every YC company.

And that's gonna change. Like all of a sudden the scarce resource isn't gonna be intelligence or agency. I really think it's gonna be vision and ambition. It's like, do you have a clear view of what you want the world to look like in the future? What is the like one way in which you wanna put your finger on the scale for how the future of the world will develop and how the world will look like in five to 10 years that it does not look like today? And you have the ambition and drive to like go through all the crap to make that happen. And AI will make that easier.

Like agents and AI makes that maybe 10 times or a hundred times easier than it was a decade ago. But the flip side of that is that all of a sudden you can dream bigger.

You said your favorite interview question is, what's the hardest you've ever worked on something? Why do you like that question? There really are two kinds of people in the world, having like an internal versus an external locus of control. So if you have internal locus of control, you believe the things that happen in your life are actually more a product of like what you do. If you have an external control, it's the opposite. You believe that the things that happen to you are mostly the outcome of things outside of your control, sort of like the world's very deterministic and you're sort of like a pinball on a big pinball machine.

I only want to work with people who have an internal locus of control. One way to like index off of that is seeing how hard do people work at things that matter to them. If they have an internal locus of control, then they're going to like work their ass off to make sure that the things that matter to them happen in the best possible way. If they have an external locus of control, things that matter to them, but they sort of like throw their hands up and let the world sort of take the wheel.

When OpenAI first produced GPT -4 in the lab, China were nowhere near that. But just even the past few months, there was a Chinese company, ZeroOne.ai, that produced a model Yi Large that is one of the best models in the world. So it's behind GPT -4O, Gemini, and Cloud3 Opus. And it's the next model right behind that in the leaderboards. So we've already seen them meaningfully catch up. Chinese LLM and AI capabilities are, I would say right now, pretty close to neck and neck with US capabilities. And I think if you plot the path ahead, based on everything we've talked about with data, they have a clear shot at racing forward.

And what we've seen even in the past few years on solar, and then most recently EVs, you're seeing this pattern play out over and over again, where the CCP approach to industrial policy is not the most innovative. But once an industry has been established, and it's about turning the crank, they are better at turning the crank than any other economy in the world.

If your recruiting organization looks like a college admissions office, you should be pretty scared. It's very hard for somebody at a big tech company to have any sort of real impact. A lot of the people they hire just end up working on like a teeny piece of a teeny piece of a teeny piece of a teeny piece of a problem. The people who get selected into these very large brand name tech companies are those over optimizing for brand and status relative to impact. By contrast, small startups are literally the exact opposite. Like you're joining a small startup because you're like, wow, I see the five people working on this thing.

I know I can have an impact. It's not going to be a cool thing. Like I'm not going to tell my friends about working in a startup and they're going to think like, oh, wow, that's really awesome. And you really want these people who don't care about status, care a lot about impact. I think big tech companies negatively select for that.

and potentially even more of a military asset than nukes. It worries me a lot. At its core, this AI technology has the potential to be one of the greatest military assets that humanity has ever seen. Let's say you had AGI and you have one country with AGI and another country without AGI. Which one will win in a war? Well, probably the one with AGI is gonna like figure out how to produce all the weapons or we'll figure out a brilliant military strategy or we'll be able to hack the other country systems. And it is potentially one of the greatest military assets the world's ever seen.

I think one of the major things that people, when they think intuitively about AI, that they kind of get wrong, you realize, oh, it just hallucinates all the time. And then you throw your hands up and you're like, oh, this technology is fundamentally limited. The tricky thing about AI is it's like very hard technology to bet against because every prior instance where you would like use an early original level, so if you use GPT -2 and you said, oh, this thing can count to 10, there's no future here, or GPT -3, you would use it and it can't solve a simple math problem. You're like, you throw your hands up, this isn't going to go anywhere.

A lot of people, even in the AI industry, don't actually believe in model improvement. It's a shame, honestly, because I think it's hard to imagine how the models will get a lot better, but they will. We need to be thinking about a world where we're just on this continued track of all improvement.

The biggest thing is you just have to really, really, really care. And I think that that trait is really, really important. And, you know, it's like just in varying degrees for different people. So I wrote this post many years ago called Hire People Who Give a Shit. And it really is pretty simple. You notice, I notice, you know, when you interview people or when you interact with people, you can tell people who are just sort of like phone it in versus people who sort of like they like hang on to their work as like, you know, it's like it's like the so incredibly monumental and forceful and important to them that they do great work.

And it's sort of like eats at them when they don't do great work. And when they do great work, they're sort of so satisfied with themselves. And so there's sort of this like the magnitude of care. And one of the greatest indicators of like, A, just like how much I enjoyed working with people or like, frankly, how successful they were at scale was really just this like, what is, you know, to what degree of their, what degree of their soul is invested in, into, into the work that they do. And so I think that that, you know, if you were to pick one thing, that that probably is the sort of like unifier in some way.

It's like, you know, I care a lot. I care a lot about every decision we make at the company. You know, I still review every hire at the company. You know, I, we have this process where I approve or reject literally every single hire at the company. Um, uh, and, and so I care immensely. And then this, and then like, I work with all these people who care immensely. And then that enables us to really sort of like, we, um, we feel much more deeply what happens in the business. And as a result, we sort of like, uh, you know, we'll change course more quickly. We'll learn more quickly.

Um, we will, uh, we'll take our work more seriously. We'll adapt more quickly. And I think that that's been quite important to the, to the success that we've had. Um, broadly speaking, you know, we have this value at our company, um, quality is fractal. Um, and, and I do believe that like high standards sort of like, um, they trickle down within an organization. And, um, you know, it's very rare that you see an organization where like, where like standards, um, increase as you get lower and lower down in the organization. You know, most of the time when people realize their manager or their manager's manager, or their like director or whomever, don't really care, then they sort of like, you know, that, that removes the sort of like the, like deep desire to, to need to care.

Um, and so it's like incredibly important that, that high standards, um, and, and this sort of like this deep, um, uh, sort of care for quality is like this, is this like, um, deeply embedded sort of, um, tenant of the entire organization.

One of the things that happened in the autonomous vehicle craze is that there were a lot of promises that were being made that were divorced from the technical reality. And so a lot of the prominent autonomous vehicle companies, a lot of the prominent organizations were making bolder and bolder promises to be able to raise money. At first, they weren't super divorced, but over time, they became more and more divorced from the technical realities. And that resulted in this very painful trough where the promises weren't met. And so it felt like the entire industry was falling apart. And actually, at the end of the day, you know, now we have Waymos driving around San Francisco, perfectly proper L4 autonomous vehicles driving around Tesla autopilot has gotten really good.

And so if we had more measured promises along the way, I think now we would feel amazing about autonomous vehicles. Whereas instead, we went through this huge op, this big down, and maybe it's sort of like on the upswing again. And I think this is one of the big concerns I have about generative AI, which is we have these really big promises that get made about the technology that get divorced from technical reality.

One of the secrets to scale, and I think this applies across almost every industry, but the problem that we originally were solving, which is around building really high -quality data sets, was something that most machine learning teams knew was very important. And so most people agree that it was very important. But it wasn't necessarily the sexiest problem that every AI scientist wanted to spend their days and nights working on. And this is like collecting images, labeling them. Yeah, collecting images and labeling them and making sure that was done very, very high quality. It was a very, frankly, very operational problem. And we've built a lot of technology to support it.

But again, at its core, it's a very operational problem. And there's one article that was pretty seminal for me early on, which was an article by Paul Graham, or an essay by Paul Graham, I should say, called Schlepp Blindness. And I encourage everyone to read it if you get a chance. But basically, the idea was that most people avoid thinking about the really sort of just like hairy, ugly, difficult, annoying problems that exist in the world, but they're really, really important. And he actually uses Stripe as one of the examples in his essay. But these problems are everywhere, the sort of ugly, hairy problems that everyone knows are important, but aren't sexy to work on.

And if you can identify what those problems are, then those generally make really exciting startup ideas. So that was a lot of the original pitch was like, you know this is important, but you probably aren't the most excited to work on it. And then we were just, I think we even very early on, our whole team was super duper scrappy. And so I think a lot of teams and a lot of customers that we work with, like saw our velocity and our product velocity and how fast we were moving and thought to themselves like, hey, yeah, these guys, even if they don't have the perfect part today, they're going to get to a product that we're going to be able to rely on really, really quickly.

It has been true for a long time that the United States has been ahead, and that's been true for maybe the past decade. That being said, the very recent event on Christmas Day, about a month ago, DeepSeek released a model, which, by the way, I think is symbolic that the Chinese lab releases an earth -shattering model on Christmas Day when the rest of us are celebrating the holiday. And they released it to much fanfare, and then they followed up with their reasoning model, DeepSeek R1, which is the one that we evaluated as top of the leaderboard. You know, the reality is yes and no. So, you know, the Chinese labs, they have more H -100s than people think.

And these are the highest -powered NVIDIA chips that they were not supposed to have. Yes. My understanding is that DeepSeek has about 50,000 H -100s, which they can't talk about, obviously, because it is against the export controls that the United States has put in place. And I think it is true that, you know, I think they have more chips than other people expect, but also, on a go -forward basis, they are going to be limited by the chip controls and the export controls that we have in place.

What are some jobs that could be at risk because of AI, right? Because a lot of our fans are probably just blue -collar listeners, like just people that work in, like you're not going to, you're still going to need a plumber, you're still going to need an electrician, you're still going to need anything where you have to physically do something, you're probably still going to need. Yeah, for sure. And then even stuff where you're like, let's say you're mostly just working on a laptop, and even for those jobs, like it'll just change. Like instead of being, I literally do the work on a laptop, it'll almost be like everybody gets promoted to being a manager.

Like, because I'm going to be managing like a little pod of 10 AI agents. But can you have that many managers, you think? Yeah, because I think that what's, the other thing that's going to happen is the economy is just going to grow so much. Like there's going to be, like industries are going to pop off in crazy, crazy ways. And so, you know, the limit is going to be how many AIs can you have? And then you're going to be limited for, in terms of the number of AIs you have by the number of managers that you have. But yeah, so it's all about like everybody leveling up to be managers.

And then also everybody, you know, just way more ideas are going to start happening. Like way more ideas are going to start becoming a reality. That's a good point. Yeah, you don't think about that. Once some of the things at the first level of certain businesses are handled more easily by AI, then you're going to be able to have more people operating at a higher level.

Why did you just call it like a single piece of art and then you need a single perspective or point of view? I think one criticism of products that come from larger organizations, and I think this like is even now starting to plague a lot of the AI labs, is that they become kind of this Frankenstein, almost like Cronenberg amalgamation of a bunch of different people's points of view and visions and beliefs. And it's sort of like PM hell in some sense, where like every PM has like a thing that they're trying to get into the product and jam into the product. And then all of that kind of like blends together into this like smoothie almost.

People feel it as users and consumers. You can feel when a product is just like lots of different people worked on it and there were like clearly different people who were gold differently. So they just kind of like jammed it all together. Muse had to be the exact opposite, which is like there's one clear point of view. There is one cohesive experience that we're trying to deliver to the world. I love that you said that. It just happens to be pure coincidence. Like I don't know when this is going to come out, but we're recording this on the 15th anniversary of Steve Jobs' death. And when you think of somebody like a consumer products that had a single point of view, you obviously think of Steve first.

I've read every single book on Steve in the history of Apple and I've done like, I don't know, 15 or 20 episodes on them for my other podcast founders. But one of my favorite lines to describe Steve's impact was it says that Apple is just Steve Jobs with 10,000 lives. The products are to his taste, his perspective, his point of view. What he wanted to see in the world is exactly what the end user got.