Andy Halliday

Currently, they're demonstrating that to enhance existing AI models with wet compute, so to speak.

But still, it's being held on electrodes.

It's a San Francisco startup, and it wants to launch hybrid neuron silicon clusters for cloud environments by 2027.

Wow.

Wow.

I did not see that coming.

And none of us did because it's been behind the scenes.

But now they can actually do that.

They're not conscious yet.

So anyway, the cell lines that are being used in this, I, I, I'll be very interested in a report back here when I find out, but I don't, I wouldn't believe that they're using human brains neurons, right?

I'm sure they're probably using some animal neuron because they, they don't really care so much about whether the neuron is, you know, a, from one mammal or another, frankly.

Yeah.

Our our intelligence comes from the fact that our brain has a lot of neurons up here in the front part, which doesn't exist in most other mammal brains.

Right.

The neocortex.

Right.

To close out, I have one more update from what we discussed yesterday because things are happening fast.

You want to stay up to date here.

So we talked about Z.AI's, I'm sorry for the Canadian audience, Z.AI's model GLM5, which really came up to the sort of the level of performance in reasoning and encoding of the major frontier models.

It's kind of leapfrogged up there.