AI safety

We give you two out of three rights. Agentic systems can access sensitive information, it can execute code, and it can communicate externally. We could keep things safe if we gave you two out of those three capabilities at any time, but not all three.

if we get something in place that’s poorly targeted, that wastes a bunch of people’s time, what’s going to happen is people are going to say, “See, these safety risks, this is nonsense. I just had to hire 10 lawyers to fill out all these forms.

I do think a lot of it is eliciting powerful pre-trained models. So people are probably divided on this because obviously in principle you can definitely teach new things. But I think for the most part, for a lot of the capabilities that we most use and care about, a lot of that feels like it’s there in the pre-trained models.

One thing I really like about the character work is from the outset it was seen as an alignment piece of work and not something like a product consideration

regulation is very good for big companies because they can follow it. I can’t follow it, right? If I want to start an AI startup in Europe now, I cannot because there’s an AI regulation that makes it very complicated for me. I probably need to get notaries involved.

if you think about an MCP gateway as a way to easily control and scale access to MCP servers, especially within you're in a cloud, I guess, platform. So we may have, like, 30 or 40 MCP servers that we run

I am, like, working harder than I ever have because I have all these agents, and they, like, they need stuff from me. And I'm constantly, like and I'm doing building things that I always wanted to build, but just never had the time to.

let's just assume you're gonna have not Internet limited or no Internet access.

on like 10% of the users, I would randomly test parameters and then I would see if they would, because you can favor the photo or you can download it, I would measure if they favor it or like the photo. And then I would A/B test