So to be meaningful, you at least need to be at like gigawatts level of infrastructure.

Which is, to give the perspective, which is a million GPUs, not as many GPUs, a million GPUs.

And the last part, the one which you mentioned, like with the MacBooks and the equipment, is part of the privacy and encryption, like of the requests and the answers.

And with this, we should understand the current nature of the state of the market.

OpenAI or Entropic or Google, They do not encrypt your data.

They store your data unencrypted.

They can read it.

But for that, the community should gather around a protocol and put all the efforts for this protocol to grow and to get to these gigawatts of data centers or like local, smaller data centers connected to the network.

David, can you start out and tell us a little bit about your career journey and how you got to where you are today?

So together with my siblings, we built several companies in computer vision field.

So before we all start to use language models, computer vision was a frontier for AI development.

There are multiple reasons for that.

One of the silly reason is that computer vision was used for entertainment in with entertainment requests.

you actually can have a mistake.

When animation glitch or when there is some special effects which goes wrong, it's okay.

How is that possible to have a shortage and a massive waste at the same time?

Yes

... their powerful model.

Um, so they were excited to use this model.

Uh, but then Anthropic just shut it down.

Shut, shut down only for users of OpenClaw.

Um, and it was a, a, um, a, a, a big effort for the entire community to try to switch, uh, off and try to build everything which is needed to switch off, uh, the model.

So the access to the model and access to AI is already right now is visible that, uh, it... we shouldn't take it for granted.

The bankers