Einar Volsett

So the actual use case for that particular domain is quite narrow, even though it produces a lot of interesting and challenging problems academically.

Potentially, it's a similar stuff.

This is typically they're stationary, the satellites.

And so in that case, yeah, you might form a mesh network between the satellites in order to, you know, in order to coordinate on certain things.

But again, it's a slightly, it's a slightly convoluted problem.

And it's also like, well, if they're going to be stationary, chances are you could just put some base stations in strategically placed and you'd solve 99.9% of the problems.

No, no, what I mean is if you need to communicate, if you need to coordinate between the satellites, you're probably better off just having all the satellites communicate down to Earth and then have a central point where you do all the coordination and send it back up again, rather than effectively trying to have a peer-to-peer network in space.

Because that's not going to be as effective, I don't think.

Although I haven't thought much about it.

But yeah, I mean, every now and again something comes up and I'm like, I could potentially have used this specific weird algorithm stuff.

Half an hour later, I'm like, no, actually, there's a much easier, faster and better solution, which doesn't require any of this, you know?

Yeah.

After Cornell, like physically or like mentally the shift away from academia?

What are you asking me here?

Both.

Well, I wanted to do something else.

And so part of Paul Graham's writings was a big part of it.

And then what I did... Sorry, what was a big part of it?

Paul Graham's writings.

He's SS.