Stephen Wolfram

For me, it doesn't make a difference, so I just go with what works.

Because we had done a podcast with the physicist Stephen Wolfram.

I can't remember what his definition of one was or one plus one, but he had his own separate view of consciousness.

And we've done so many podcasts with different people.

It's very different from, it has language elements, just like we can describe consciousness.

Is consciousness a language? I think it's ridiculous.

So I'm very worried, you know, but in some sense I'm excited because it means that physicists will have a place in the future to actually do stuff.

You do talk a lot about falsifiability.

where you say, if now I can do so many different things, what are the foundational pieces that I need to understand so I can learn to push back on this? And this is where my second aha has been.

So I just watched this great YouTube conversation with Stephen Wolfram, who you might want to actually invite on your show.

And in this interview, he said something that has just, I heard it like 10 days ago, and it's just been like rattling in my brain.

You know, he designed mathematical languages to do integrals.

We've talked about this before where I think we appreciate books that are a little more bold with who their target audience is and a little more restrained with what exactly they're trying to cover.

I actually think what is ChatGPT doing and why does it work by Stephen Wolfram, a friend of the podcast, by the way, we've had him on, is a great example of kind of saying like, I know exactly who I'm talking to and I'm just going to limit the breadth of what I talk about.

So I was a little nervous in chapter one.

But chapters two and three are, I think, much more like meaty.

We get to talk about whatever we want.

Uh, but what we want to talk about right now is chapter two, but this is all the like understanding foundation models and like how they're built.

Um, A lot of overlap here with like, what is ChatGPT doing and why does it work, uh, by Stephen Wolfram? Um, I think a lot of us are familiar with kind of like the, the, like the, the base details of like an LLM models.

Like LLM models are trained, they're, they're next token prediction machines.

Is consensus reality basically just an average of what humans are experiencing, right? Like versus some other civilization that has a completely different set of senses that will experience the laws of physics completely differently because of it? And there's some, you know, I'm going with this somewhere, I promise.

There's some physical theories now, you know, by Stephen Wolfram.

There's some propositions by David Deutsch.

There's a few ideas that actually point to the fact that even computationally, if you're looking at a computational graph, just kind of like making decisions infinitely...

If the graph responsible for executing the rules themselves, so the way it appears, like the laws of physics and everything, then if you find yourself somewhere else on the graph, this is, for example, Stephen Wolfram's physics project, if you look from somewhere else in the graph, even though you're technically looking at the same laws of physics, it will look like different laws of physics to you because of the angle from which you're looking from that particular computational graph, right? So now, so just to make it clear that there are some ideas in theoretical physics that do make sense of why even consensus- Oh, but before- I like

interface

theory.

There's three or four of them.

One is Stephen Wolfram's description.

What is chat GPT doing and why does it work? That explains the mechanistic aspects of the GPT's process.

Very cool.

Comprehensible by humans and that like the humans gonna write down a couple equations, you know, like an interstellar It's like you're so on a blackboard you write down a couple equations and you save the world, right? that's like the romantic idea of like if this is half of themselves, but I realized that I didn't think that was gonna happen in my lifetime and and then actually I Actually, a lot of nature is irreducible.

And this is Stephen Wolfram's theory.

And what does that mean? It means that you won't have a simplified computational model of it.

Your model of that piece of irreducible nature will need an amount of computation for your generative model to make predictions that is similar to the generative process that gave rise to that emergent property in nature.

Like, it's weird they do all these things.

We don't even have a basic answer to that about why it's so good, right? And Stephen Wolfram's angle on a mathematician was that in creating a model of human language, we created a model of the world because human language models the world.

Otherwise, how could we describe... quantum physics, unless we could describe it with language.

And I find that a somewhat satisfying answer, but it's still hard to answer the questions about consciousness or anything else.