Andy

And what's going to happen here, I'm pretty sure, is

you know, like the screen on the left, which is the Cerebrus powered one is, is changing dramatically while GPT three 5.3 codex is still after one minute.

Just thinking.

Yeah.

It's, it's, it's slow.

Yeah.

It's basically running on a different chip architecture that Cerebrus built, which optimizes for inference.

So, you know, just like Grok, the GROQ company, you know, built a special chip, a language processing unit, LPU.

Cerebrus has wafer scale chips.

inference engine and it's a huge massive chip stack that you know is goes actually beyond i think the capabilities of nvidia's forthcoming vera rubin stack and and it's already out there and there are small data centers that can run using cerebrus and it's not inexpensive i i expect but that's what we're seeing if we're seeing just how powerful and quick ai can be when it's running on

optimized chip architecture.

That's a beautiful example of the narrow sort of niche applications where you need Cerebra Speed.

And at the other end of the spectrum, we're seeing a migration towards local models that are going to run very slowly, even on your phone at 40 tokens per second.

But you've got time to take a sip of your coffee and breathe a couple of breaths before this thing spits out an enormous long thing that you have to try to read and comprehend.

So there will be need for what we just saw, but I think it's going to be at the tip of the pyramid, the top unique use cases.

SpaceX AI is a different company.

Probably in development, but we haven't heard about it yet.

Right.

Take the example...

Take the example of a hackathon, Beth.