Illia Polosukhin

And there's a lot of tooling, again, blockchain can support on provenance and tracking and which data was used for specific answers and where it came from, et cetera.

All of those are important, but the fundamental piece is we actually don't have open source models right now.

We have open weights models and we have no source for them.

So that is a fundamental piece that we need to solve.

Yeah, totally.

Yeah, I also say open source and then correct myself.

Yeah, it's hard.

There's Allen and there's now work from Stanford as well on this.

Yeah.

So, I mean, we are building a lot of infrastructure for this and we will have kind of a core product that we want to surface all of this to the consumers and companies.

But it is an open infrastructure and we totally expect a ton of people to build different aspects of this as well.

and we already have like you know few partners who are really interested in different aspects like indeed the data kind of this content creation etc is definitely a component that you know few people interested in and participating in different ways so yeah i think it will be a combination of like kind of core product that is private that people can trust and use and you know can um really leverage and then uh kind of surrounding tools and different versions of of this uh

for different areas.

You vibe coding with Claude?

What are you doing?

I actually vibe coding with V0.

I mean, I don't actually write any production code nowadays.

And so I do a lot more just like, this is how I think this should look.

And it's been really like, I'm a back end machine learning engineer.

My front end's always sucked.