Ahmed El-Kishky

Yeah.

But it's demonstrably better.

Like you could tell GPT-5 tried seven times to solve that last problem.

And each time it's failed.

So you can imagine, you know, like Mustafa and Boris and Robin and Andrew sitting there submitting every, you know, like 10, 20 minutes.

I mean, OK, this one's wrong, too.

Yeah.

And all the while, the experimental reasoning model is just thinking.

It's thinking for hours and hours.

It hasn't even like outputted a single answer yet.

So it's three hours later and it's still thinking.

It's just grinding on this one problem.

And then eventually it gets to

solving it, we start seeing it, like, you know, get some answers.

We submit the first one, it's wrong as well.

And then the second one that we submit from that model is correct.

So GBD-5 tried seven times and got it wrong.

And then, you know, on the second try, the experimental model, the IMO model, just, you know, got it.

I don't really know, but it must have been within, like, 30 minutes of, like, the competition end.

Wow.