Tom Bilyeu

I feel like for qualia to exist, it must be processed.

I will even grant that the processing is simply the Markovian dynamics of moving from one thing to another, the switching of states, fine.

But it is moving from one state to another, which I will call that processing.

Yeah, I try not to kill the audience with the things I just can't remember.

Do you have an example of something that's non-computational?

I think you gave one earlier, but I forget.

Yeah, that's right.

By its nature.

In fact, probably most.

So the question is- So when you say non-computational, you mean something that ends up in that loop?

So the halting problem is- And most things are like that.

There's no Turing machine that will know which one is going to halt?

When somebody says it's Turing complete, does it mean that it halts appropriately or is that something else entirely?

That's something else entirely.

Got it.

Okay, that's crazy.

Persistence.

So this is, I think, a key part of my thesis, which is persistence is the hard problem for AI right now.

And when I look at the thing that holds AI back from being something that people can implement into their workflows today, it's because it can give you really amazing stuff, but it cannot give you really amazing stuff over and over from different angles and different setups.

People are working on it and