Zuzanna Stamirowska

Yeah.

Such that you actually create a framework space on this that you can kind of adapt to new situations.

Okay.

This is a very big difference of having like evolving memory, contextualized evolving memory.

That's your own, right?

Yeah.

And having a brain, let's say, which is set once.

So I guess right now we're kind of, you know, pushing.

In fact, it's not through LLM's per se that we are pushing and you would even have folks from OpenAI, you know, saying some things that the way forward is through reasoning, right?

Yeah.

So how far can we get with reasoning?

So

I wouldn't, so reasoning is reasoning, it's even less related to transformer per se.

So there I wouldn't like put a bar necessarily, but just, I mean, given the math, like the memory is not there.

So it's difficult and kind of somehow tiresome to actually try to trick transformer into having memory.

So what I like to talk about is like epicycles.

So you guys, like before we had Copernicus and the proper theory of solar system, people were observing the moon.

And to make sense of the observations, they're trying to kind of design some sort of orbit that would be maybe like this, because that was the only way that they could explain the observations, right?

Yeah.

It was like cumbersome, pretty ugly if you think about this, but then every time they got a bit better, they were getting, you know, like, well, champagne or, you know, they would party.