Corey Knowles

So was the idea like immediately that

were going to do fully autonomous agents or was it, were there stepping stones along the way you were hoping for?

Uh, how did that vision kind of evolve if it did along the way?

What essentially are they?

I know we're talking about agents, but what are they really doing under the hood and how autonomous are they today?

So you have some model flexibility built into here too, right?

And how do you choose the right model for a specific agent with the way these work?

Are you able to adjust them on the fly?

Are you able to, you know, I think of that whole idea of like, you know, having a basic and a power model kind of paired together where for certain parts of tasks it's using this, but

Maybe it recognizes when it's in over its head and needs some extra go-jus.

So you built your own compression approach, and then we do have where OpenAI and Anthropic have landed their own kind of methods as well.

And what else did you find around benchmarking those together?

and seeing how they compare.

I know you said yours was faster.

Stepping outside of coding, this feels like an approach that could be useful for any long context task you have.

Could this help solve, you know, memory context window issues for non-coding agents as well?

I'm suddenly drawn back to the idea that when we met in person a few weeks ago, I told you I was going to try to write with wine, and I haven't done that yet.

But that's on my list.

I may do that this week.

I absolutely should have just to see how it goes.