Andy Halliday

Yeah.

The way we would have to do, I think, in custom GPTs is you'd have to delete that document.

That's the way you do it.

We would do the document and upload it again.

It's the process right now.

Cool.

Can I move on to some updates on some news that we talked about yesterday and some news today that sort of moves the chips forward on the board here?

We talked yesterday about the whole idea of trying to mimic biological computing in order to achieve energy efficiency in silicon.

And I talked about Naveen Rao's startup, Unconventional AI, which has gotten $475 million to go after this question.

Like, how do we imitate neurons more directly in the physics of silicon rather than

just run in computation, which I've understood how to do on GPUs since the beginning of video games, right?

So now a different approach to the design of chips.

Well, we then kind of leaked into the whole idea of wetware and the companies that are out there using actual living cells as components in the compute infrastructure.

So now...

A stealth company just emerged yesterday called the Biological Computing Company.

And it has a $25 million seed funding round led by primary venture partners and is going to commercialize neuron-powered AI.

Now, when they say neuron-powered, they're talking about real neurons.

Okay.

The co-founders are both neurosurgeons.

And they grow neurons on electrodes and then encode data.