Rizwan Virk

It's just too many particles.

But then they came up with this idea of 3D modeling and rendering.

And you've got things like what's obfuscated, don't bother.

Even if it's behind this curtain, we don't need to necessarily render it.

So computer science is often about optimization.

And there's something called lazy evaluation in programming.

And lazy evaluation means you don't evaluate, you don't actually run the code unless you need to.

So if you say x equals 2 to the 1 million power, but then you never use that variable x, that computation might take a long time to do.

I mean, it's not that hard for a computer to do, but you can think of a more complex computation.

But if it's never used, it'll just wait around until somebody needs x, and then it'll go and it'll compute it.

Now, in this case, because you've been living here, it's probably cached, so it's already there.

Yeah, right, okay.

Because that's the thing you might look at next, whereas you're not going to go to Mogadishu immediately.

You'd have to get on a plane.

Exactly.

I mean, it's like saying, can you run chat GPT on a 286 or a Pentium processor?

No, you need the GPUs from NVIDIA.

But more than that, you need the LLM.

training and the algorithms, which couldn't have been done.

And that's where it gets even a little bit stranger, okay?