Skip to main content
Edwin Chen

Edwin Chen

Jun 24, 2026

7:13
W- And what does that make you think? Okay, you, you wanna be a f- you wanna be a mathematician when you grow up, Fields Medalist sort of saying, "I'm relieved that it's not good enough," but you're, I, it's, you're s- you're talking as if, like, you feel pretty confident that it will be good enough in the next couple years.
7:29
Yeah.
7:30
So my belief is that if you really believe in scaling laws, and I do, it's that it almost seems like there's nothing that humans can do that AI won't soon be capable of.
7:46
And if you think about that very deeply, I think, uh, you almost have to worry about what would that mean for humanity? Like, what would that mean for the role of humanity in the universe? Like, a couple years ago, you know, we think about humanity and human intelligence as playing this very, very unique role in the galaxy.
8:06
But then AI comes along and shows us that, as far as we know, we can create something that's actually smarter than us and better in many ways.
8:13
And so you can sort of imagine one path where humanity as a species falls into a paralysis because people believe AI will do everything better anyways.
8:23
Like yeah, all these kids who formerly would have really wanted to grow up to do mathematics, maybe now they believe that, "Okay, AI will just do it better than me anyways.

22 MINS LATER

30:39
So I guess the obvious environment is, like, using a computer, but what would be an example of an environment that's non-obvious that's teaching models things we might not think of?
24:04
You've never raised? Wow.
24:05
So we're still happily bootstrapped in [laughs] and growing in, I think, the, the best way possible.
24:10
So, uh, [laughs] I'm actually really, really happy that, that we've never raised.
24:13
And I mean, yeah, to your point, I think it's never been easier to raise money in AI.And that, that's kind of the problem.
24:20
Like, when you raise a billion dollars, you get all these growth targets from your investors that incentivize volume over quality.
24:26
You get all this board pressure to spend your time optimizing for your next fundraise instead of the product that you're building.
24:31
Like, like all of my friends who are, who are, uh, CEOs of other companies, they're like, "Oh, yeah, I have to spend the next few weeks just prepping a board deck." And, uh, they're, like, always jealous of the fact that I [laughs] I, I don't have to do the same thing.

19 MINS LATER

43:32
So how do you think about efficiency in company building?
7:24
What are like the best companies doing to figure out like, hey, you know, on a month to month basis, like are our models getting better?
7:29
So historically...
7:31
All the companies that paid really, really hard attention to benchmarks, like these very, very academic benchmarks that had been created by the research community.
7:39
And not all the labs realized the problems with basically benchmarks.
7:45
it's almost very easy to optimize for benchmarks.
7:48
Models, at the end of the day, they're very good at hill climbing, very concrete, very specific, very narrowly defined objective functions.
7:54
And so what would end up happening is that the models would make a ton of progress on the benchmarks.

23 MINS LATER

31:15
What are some of the key vectors where that divergence exists?
19:31
And... that kind of to what you're saying we are not ready for what is about to be unleashed from where you sit because you're on this cold face in a way that most people are not Do you think AGI is a thing that will happen? And two, do you have a sense of timeframe, recognizing that any prediction is going to be wrong because that's just how predictions go? Yeah,
19:53
I do believe that AGI will happen.
19:55
Like I see the trends, even today, when we add more and more of the data that we create, we teach the models in ever more sophisticated ways.
20:03
Yeah, we basically always see them improving.
20:05
So I do see this path to AGI.
20:08
I don't believe that will be as soon as everybody else thinks.
20:12
Like, honestly, I think a lot of reason that people have such short timelines is because there's this Silicon Valley impetus towards hype and the need for fundraising.

13 MINS LATER

33:10
he was like britain is becoming a vassal state of like a handful of silicon valley tech empires basically and i think implicit in that is this idea is that there's a real group think out here and that this place is fundamentally both boring and uniform it's like keeping up with the joneses but like to a really really high level oh here's my latest startup there's not a lot of conviction other than i just got to get my next round i got to get my valuation up i had to get more clicks Does that kind of ring true to you? And do you think that is dangerous? Again, when we go back to this idea of AGI and who was building these things and who they're building them for, is that a problem?
5:30
I'm curious just do you think this will happen more and more as a result of AI? And also just where has AI most helped you, uh, find leverage to be able to do this?
5:40
Yeah, so we hit over a billion in revenue last year with under 100 people, and I think we're going to see companies with even crazier ratios, like 100 billion per employee in the next few years.
5:50
AI is just going to get better and better and make things more efficient, so that ratio just becomes inevitable.
5:56
Like I- I used to work at a bunch of the big tech companies, and I always felt that we could fire 90% of people and we would move faster because the best people wouldn't have all the distractions.And so when we started Surge, we wanted to build it completely differently with a super small, super elite team.
6:11
And yeah, what's crazy is that we actually succeeded.
6:15
And so I think two things are colliding.
6:18
One is that people are realizing they don't have to build giant organizations in order to win.

17 MINS LATER

23:12
Can you just talk about that?
12:16
And so can you talk a little bit about what technology you actually have to build to manage quality for a generic, complicated expert task involving language?
12:29
So let me start actually by contrasting with... a very old school take on data and data quality and how we just think about the problem differently.
12:41
And so, yeah, so let me give you an example.
12:43
So let's say you wanted to train a model to write an eight line poem about the moon.
12:48
And the way most companies think about it is, okay, well, let's just hire a bunch of people from Craigslist or through some recruiting agency.
12:53
Let's ask them to write poems.
12:55
And the way to think about quality, again, going back to coming to image annotation days, it's like, okay, is this a poem? Is it eight lines? Does it contain a word moon? They just check all these boxes in the same way you might check, okay, this is a cat, this is a dog.

27 MINS LATER

40:04
But how would you collect the data? Would you get a Stanford professor and have it just score responses? I guess a scoring response would be easier than just generating responses.
17:30
If you assume eventually some form of superhuman performance across different model types relative to human experts, How do you think about the role of humans relative to data and data generation versus synthetic data or other approaches? Like at what point does human input sort of run out as a useful point of either feedback or data generation?
17:52
So I think human feedback will never run out.
17:54
And that's for a couple of reasons.
17:55
So, I mean, even if I think about the landscape today, I think people often overestimate the role of synthetic data.
18:02
Personally, I think synthetic data actually is very, very useful.
18:05
We use it a ton ourselves in order to supplement what the humans do.
18:08
Again, like I said earlier, there's a lot of cruft that simply isn't worth a human's time.

8 MINS LATER

26:26
Going back to something Elad mentioned, like where should we expect to see like Surge investing over time? Like, what do you think you guys will do a few years from now that you don't do today?
16:32
Can you take me to the founding moment for you, leaving the last company and deciding that you were going to go all in on Surge?
16:38
Yeah.
16:39
So I used to work as an ML engineer at a bunch of the big companies and the problem I just kept running into was that it- it just kept on being impossible to get the data that we needed to train our models.
16:49
For example, I used to work on our ad- search and ad systems at Twitter, and one of the thing- first things I wanted to do was build a sentiment classifier.
16:55
Yeah, it's a super simple problem.
16:56
All- all you need is 10,000 tweets labeled as positive or negative to train your models, but our human data system at the time was literally just two people we'd hired off of Craigslist working nine to five.
17:06
Even just in order to get started, we had to wait a month, then we had to wait ano- another month for them to label the tweets inside a spreadsheet because the tools that we had were just terrible.

26 MINS LATER

42:45
algorithms.If compute continues to prove to be the unlock, where throwing more compute at it unlocks more and more performance, does that denigrate data quality in the prioritization stack?

We value your privacy

We use cookies to understand how you use our platform and to improve your experience. Click “Accept All” to consent, or “Decline non-essential” to opt out of non-essential cookies. Read our Privacy Policy.