Corey Knowles

It's building a very specific product for a very specific audience.

So they're kind of like speed running it.

And yeah, I think it's a bit of a glimpse of the future.

So by the way, this is, I think, for what you should do if you want to use Codex for something like a production series project.

I think if you're using Codex for something that's much more prototype-y, experimental, throwaway, then I think if it's just a single-player tool that doesn't have any user data in it, you could just put it out there.

It's fine.

Put it up next day as projects coded with Codex, then you can like deploy them on it to like Vercel or something.

Yeah.

That's the one place where you have considerations.

It's just like if you're asking users to upload data, then you've got to make sure you do right by making sure that it's secure enough for whatever you're asking for.

Or more than a year?

I forget what month we launched the research preview in, but it's one of the M months.

Yeah, like, okay, so this is something I care about a lot, actually.

So, you know, our mission at OpenAI, I might not get the exact wording right, but our mission at OpenAI is to ensure that AGI benefits all humanity.

And I'm

I believe we're going to have amazing AI capabilities and AGI, and we already do already have amazing AI capabilities though.

The AI part is already so powerful, but actually there's still a ton of work to just bring those capabilities to all humanity right now.

We've got new models coming, but even if we didn't, I could just, as a product person, work for years to help make sure that people get the benefits.

If you think about

AI adoption, I think there's broadly three phases.