Lenny Rachitsky: uncertainty

Transcript context

…Yeah. Okay, last question. Being at OpenAI, I can't not ask about your AGI timeline and how far you think we are from AGI. I know this isn't what you work on, but there's a lot of opinions, a lot of, I don't know, timelines. How far do you think we are from a humanly human version of AI, whatever that means to you? For me, I think that it's a little bit about when do we see the acceleration curves go like this? Or I don't know which way I'm mirrored here. When do we see the hockey stick? And I think that the current limiting factor, I mean, there's many, but I think a current underappreciated limiting factor is literally human typing speed or human multitasking speed on writing prompts. And like you were talking about, it's like you can have an agent watch all the work you're doing, but if you don't have the agent also validating its work, then you're still bottlenecked on can you go review all that code? So my view is that we need to unblock those productivity loops from humans having to prompt and humans having to manually validate all the work. So if we can rebuild systems to let the agent be default useful, we'll start unlocking hockey sticks. Unfortunately, I don't think that's going to be binary. I think it's going to be very dependent on what you're building. So I would imagine that next year, if you're a startup and you're building new pieces, like some new app or something, it'll be possible for you to set it up on a stack where agents are much more self-sufficient than not. But now let's say, I don't know, you mentioned SAP, let's say you work in SAP, they have many complex systems and they're not going to be able to just get the agent to be self-sufficient overnight in those systems. So they're going to have to slowly maybe replace systems or update systems to allow the agent to handle more of the work end to end. So basically my long answer to your question, maybe boring answer is that I think starting next year, we're going to see early adopters starting to hockey stick their productivity. And then over the years that follow, we're going to see larger and larger companies like hockey stick that productivity. And then somewhere in that fuzzy middle is when that hockey sticking will be flowing back into the AI labs and that's when we'll basically be at the AGI tier.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.