The Frontier with Brendan Foody
00:33:31 · Andy Fang, Brendan Foody
[00:33:31] S1: What do you think the timeline is? [00:33:33] S0: Three to five years for 50% of knowledge work tasks in the digital economy. Because I because of this point around virtual coworker interaction, long horizon uh reasoning, I just think that'll take a while to coordinate. And then I feel like 100% is really far out because there's so many things that are also just like intrinsically human, right? Of like people want to talk with someone else in certain settings, uh, whether it's a conversation with a therapist or a sales meeting or whatever it is. And I feel like all those innately human things will just be structurally really challenging to automate. [00:34:14] S1: I think one thing even today that we're struggling with, I think a lot of these models, they like they tell how much better they are at like handling tasks over a longer time horizon and stuff. For enterprises, it's like on the one hand, you want to use the capabilities of the best models. But on the other hand, I mean, I'm not the only person to say this, I know for a fact, but like a lot of the runaway, like dumb spend that we're seeing in AI is people running these background tasks that are aren't doing anything. And so there's like this tension of where we're like, hey, don't do like background tasks that don't do anything. At the same time, the labs are like, hey, these like long background tasks are like, you know, the frontier of intelligence. And so I think that's something even today, we're just trying to figure out how to incorporate that.
00:34:14 · Andy Fang, Brendan Foody
[00:34:14] S1: I think one thing even today that we're struggling with, I think a lot of these models, they like they tell how much better they are at like handling tasks over a longer time horizon and stuff. For enterprises, it's like on the one hand, you want to use the capabilities of the best models. But on the other hand, I mean, I'm not the only person to say this, I know for a fact, but like a lot of the runaway, like dumb spend that we're seeing in AI is people running these background tasks that are aren't doing anything. And so there's like this tension of where we're like, hey, don't do like background tasks that don't do anything. At the same time, the labs are like, hey, these like long background tasks are like, you know, the frontier of intelligence. And so I think that's something even today, we're just trying to figure out how to incorporate that. [00:34:56] S0: I feel like there's a lot of opportunity to build an agent that just does spend monitoring. Oh, I'm sure yes. How useful are all of these different agents? [00:35:03] S1: Or just tell the agent, hey, you can only spend this much.
00:35:03 · Andy Fang, Brendan Foody
[00:35:03] S1: Or just tell the agent, hey, you can only spend this much. [00:35:05] S0: I'm sure the frontier labs are jumping to ship that feature. Yeah, exactly. What about the future of knowledge work? Because I think it ties to this closely of what role do humans play in the economy? You obviously run with Tony one of the largest labor marketplaces in the world. And I feel like so many people are asking this question of what role humans will play in the economy, both in the digital world and then subsequently the physical world. How are you thinking about that? [00:35:32] S1: Right now, we think there's a lot of work that is needed for us to get more and more dashers on the platform. And even like in the context of autonomy, you know, I think um my co-founder Stanley has actually also said this multiple times. It's like we actually think in the future we're gonna need more dashers, not less with uh autonomy, you know, because you think about let's say autonomy fast forward to the future, autonomy's here. We got the drones, we got the self-driving cars, all this stuff. You're gonna need to do fleet management, you're gonna need to telly up in situations where things aren't going well, all the supply chain stuff, like figuring out how to like the manufacturing, figuring out how to just like reposition the cars and the drones. [00:36:10] S0: Yeah. I've heard the stories about how when a Waymo gets stuck in San Francisco, they have people positioned all throughout the city that can run through it. Oh, we're actually we actually power uh of any Waymo location.
00:36:10 · Brendan Foody, Andy Fang
[00:36:10] S0: Yeah. I've heard the stories about how when a Waymo gets stuck in San Francisco, they have people positioned all throughout the city that can run through it. Oh, we're actually we actually power uh of any Waymo location. [00:36:21] S1: No, we have a partnership with Waymo to close doors on Waymo's. Okay. So like someone leaves and the doors open, I guess they they figured out they will dispatch a dasher. [00:36:30] S0: You're totally right. And it ties to the elasticity of productivity in the economy. Like if you could drive down the cost of a delivery, people would just get like way more deliveries. Yeah, exactly. Similarly, if we can automate 80% of what it means to be a software engineer, you'd probably hire more software engineers because the ROI of each marginal person is so high. And so it feels like it's just been this case of Jevons paradox where everyone speculated around running out of jobs and things to do. And when we made people more efficient, unemployment rates are still lower than ever.
00:36:30 · Brendan Foody, Andy Fang
[00:36:30] S0: You're totally right. And it ties to the elasticity of productivity in the economy. Like if you could drive down the cost of a delivery, people would just get like way more deliveries. Yeah, exactly. Similarly, if we can automate 80% of what it means to be a software engineer, you'd probably hire more software engineers because the ROI of each marginal person is so high. And so it feels like it's just been this case of Jevons paradox where everyone speculated around running out of jobs and things to do. And when we made people more efficient, unemployment rates are still lower than ever. [00:37:02] S1: There's certainly gonna be jobs that are gonna change. Like, I think there is still a lot of people, even within our company, I'm sure at your company as well, who are like either, I don't know, skeptical is the right word or afraid of like really leaning in to the capabilities of the models. There's gonna be certain jobs that are gonna be like people are gonna need to learn better, and there's others where like anybody on the team can create a website now, like an internal website. Yeah. So, like, you know, we used to have people fully dedicated to like, hey, we need people to build internal tools at DoorDash. And now I don't think we need people dedicated to that anymore because everybody can do it themselves. [00:37:35] S0: What do you think of all the tech layoffs and how much of that is AI versus just over hiring? I don't think it's very AI related, to be honest. Do you see types of roles that will get displaced more?
00:37:35 · Brendan Foody, Andy Fang
[00:37:35] S0: What do you think of all the tech layoffs and how much of that is AI versus just over hiring? I don't think it's very AI related, to be honest. Do you see types of roles that will get displaced more? [00:37:47] S1: Hard to say. I will say with the transformation of AI, I think there's certain industries that got hit harder than others. So maybe that's like a contribution to it. Maybe that you could say that's AI related where certain like SaaS or whatever, maybe they got hit harder. I don't think people can credibly say, like, hey, because AI is here now, like everything every engineer is five X more productive. Like there's some engineers who are, but I don't think that's unanimously true, even at the best companies. [00:38:12] S0: What about for functions like fraud or like these more monotonous uh things? Do you feel like it's having an impact on your guys' workforce planning there? [00:38:20] S1: The way I would frame it is there's workflows that we can automate much more. And there are things that we can get those people to do that like we didn't have time for them to do previously. Yeah. Yeah. So I think that's more the way we're framing it right now.