The Frontier with Brendan Foody
00:28:05 · Andy Fang, Brendan Foody
[00:28:05] S1: Pretty much. And then, but we also have our own like harness where like we can also you can spin up like a agent on a cloud instance and kind of do stuff. But like I think that one requires a little bit more wrangling. [00:28:18] S0: How do you think about incentivizing AI adoption? Because there's obviously one school of thought of token maxing and like how do you just get people making as many queries using as many tokens as possible? Yeah. Which I think there is some value in, but obviously there is often a disconnect between usage versus actual outcomes. And then the other school of thought is how do you just look at the outcomes that people are accomplishing? But it's hard to know exactly if you set higher expectations for someone. You know, it's hard it's hard to like calibrate those expectations to what they should be with and without AI. Yeah. So how do you how do you think about those two schools of thought of token maxing versus focusing on outcomes and how you guys incentivize AI adoption through DoorDash?
00:28:18 · Brendan Foody, Andy Fang
[00:28:18] S0: How do you think about incentivizing AI adoption? Because there's obviously one school of thought of token maxing and like how do you just get people making as many queries using as many tokens as possible? Yeah. Which I think there is some value in, but obviously there is often a disconnect between usage versus actual outcomes. And then the other school of thought is how do you just look at the outcomes that people are accomplishing? But it's hard to know exactly if you set higher expectations for someone. You know, it's hard it's hard to like calibrate those expectations to what they should be with and without AI. Yeah. So how do you how do you think about those two schools of thought of token maxing versus focusing on outcomes and how you guys incentivize AI adoption through DoorDash? [00:29:02] S1: I would say we definitely focus much more on outcomes. So, like, for example, like we will try to solicit like concrete anecdotes of where people have used AI to make their job a lot better. Whether it's, hey, I got this project done like three times faster, or it's like someone on the recruiting team like kind of vibe coded their own internal like pipeline uh website tool that like saved them a lot of time, or kind of like what you mentioned, like a schedule action type workflow. And so that's what we're really looking for is like, hey, let's try to like and what we also try to do is celebrate those. So like if we identify that, then we share it in a forum and like that way people can kind of learn from example. And so that's kind of where we're at right now.
00:29:02 · Andy Fang, Brendan Foody
[00:29:02] S1: I would say we definitely focus much more on outcomes. So, like, for example, like we will try to solicit like concrete anecdotes of where people have used AI to make their job a lot better. Whether it's, hey, I got this project done like three times faster, or it's like someone on the recruiting team like kind of vibe coded their own internal like pipeline uh website tool that like saved them a lot of time, or kind of like what you mentioned, like a schedule action type workflow. And so that's what we're really looking for is like, hey, let's try to like and what we also try to do is celebrate those. So like if we identify that, then we share it in a forum and like that way people can kind of learn from example. And so that's kind of where we're at right now. [00:29:42] S0: So sort of rewarding the behavior culturally through shout-outs, probably performance reviews, all that kind of stuff so that people focus on aiming higher and accomplishing more with AI. Yeah, exactly. Nice. Andy, there's so many people talking about recursive self-improvement and super intelligence within two or three years. How do you think about that and how much AI is overhyped versus underhyped?
00:29:42 · Brendan Foody, Andy Fang
[00:29:42] S0: So sort of rewarding the behavior culturally through shout-outs, probably performance reviews, all that kind of stuff so that people focus on aiming higher and accomplishing more with AI. Yeah, exactly. Nice. Andy, there's so many people talking about recursive self-improvement and super intelligence within two or three years. How do you think about that and how much AI is overhyped versus underhyped? [00:30:05] S1: You can't argue with the amount of spend that companies are I mean, for us 20x in like six months, like that's definitely not hype. But I don't know. This RSI thing, at least the people I've talked to, they don't think it's coming that soon. I do know for a fact that for some of these labs, it's like, I guess for all the like the top labs now, it's like a top priority for them. I think the potential there is obviously tremendous if we can get to RSI. Yeah, I don't know. It's just like things are changing so fast, it's kind of hard to like wrap my head around. [00:30:32] S0: So maybe two questions. First of all, is how long until AI will be better than humans at 50% of tasks in the digital world? And then second of all, how long until hu AI will be better than humans at 100% of tasks in the digital world? So like digital super intelligence.
00:30:32 · Brendan Foody, Andy Fang
[00:30:32] S0: So maybe two questions. First of all, is how long until AI will be better than humans at 50% of tasks in the digital world? And then second of all, how long until hu AI will be better than humans at 100% of tasks in the digital world? So like digital super intelligence. [00:30:49] S1: I'll go back to the tax thing, like filing my taxes thing. Like, I don't know. Some people at the labs might think that Fable or Mythos or Soul is good enough to file your taxes. I think for me, my I think there's a couple definitions there. It's like, is AI good enough to do it like in tandem with a human, or is AI good enough to do it fully autonomously by itself? Maybe it's like a hey, I I my con I just need access to the data and then I could do it. Maybe that's the only blocker. But like I think for me, to do 50% of tasks well, I think people's jobs are just gonna evolve, I guess.
00:30:49 · Andy Fang, Brendan Foody
[00:30:49] S1: I'll go back to the tax thing, like filing my taxes thing. Like, I don't know. Some people at the labs might think that Fable or Mythos or Soul is good enough to file your taxes. I think for me, my I think there's a couple definitions there. It's like, is AI good enough to do it like in tandem with a human, or is AI good enough to do it fully autonomously by itself? Maybe it's like a hey, I I my con I just need access to the data and then I could do it. Maybe that's the only blocker. But like I think for me, to do 50% of tasks well, I think people's jobs are just gonna evolve, I guess. [00:31:23] S0: Yeah. Well, it's crazy how dragged the frontier is, where yeah, on one hand, the models are discovering these new math theorems that like you know no one's ever been able to solve. Totally. And on the other hand, they still can't do our taxes, right? Or so many of the basic things that we try to automate throughout a company. I I also have been thinking a lot about how we interact with coworkers. Sure. Because if I think about all of the tasks that I have to do in my job, and I imagine it's probably the same for you, probably 60 to 80 percent of those involve interacting with other people. Yes. If we think about how the models are used today, it's almost non-existent insofar as interactions with other people. And so it feels like that's gonna be one of the next giant evolutions of what tasks the models need to learn how to do.