That’s not quantum advantage, because it only compares the quantum algorithm to one classical algorithm.

]]>I really wish Professor Aaronson can comment on Chollet’s views by watching his videos (https://www.youtube.com/watch?v=mEVnu-KZjq4) after all Professor Aaronson is a Bayesian and so an AI researcher in spirit.

]]>Chollet is a very very smart guy, one of the few people I follow. Chollet, in a sense, solidified the final pieces of my metaphysics. Chollet talked about topology-based cognition versus geometry-based cognition, and I think he’s right. Read what follows.

I think the metaphysics of reality (including the nature of complexity, computation, time, intelligence and values/alignment!) is all so simple. What’s remarkable is how simple it all is. And by implication, how incredibly poor us humans really were at philosophy. I mean it could probably have been worked out long ago. And it should have been.

—

So, I guess the core postulate I would make is that there is a ‘space of possible worlds’ (so this is modal realism), and this is the scaffolding to the ‘computational layer’ of reality, which is still under construction (this is a growing block universe picture). The construction of reality is not yet complete, and computation is the incomplete layer! This computational layer amounts to paths through the space of possible worlds.

Computation is equivalent to *time*. So the concept of time was a computer science concept all along! Time cannot be understand within the context of the ordinary physics picture. And time manifests itself in 3 ways: as causality, as compression and as compositionality. And that’s equivalent to the geometry, iterated functions and topology, respectively, of the paths through the space of possible worlds. So *particular* worlds are becoming actualized from the space of *possible* worlds, and this happens at the computational level. And this is the ongoing construction of reality.

The math and physics layers of reality are the backdrop (the possible worlds) of reality, from which emerges a 3rd layer, the computational layer, which is time (particular worlds or paths through the space). To exist *is* to be computable, to be reducible, to be comprehensible! And the more comprehensible something is, the greater the measure of its existence.

Cognition (or mind) is the mapping of the paths through the space of possible worlds, based on measures of the 3 arrows of time, which for causality, compression & compositionality, are probability, complexity & truth values, respectively.

For the 3 arrows of causality, compression & compositionality, the corresponding cognition is perception, action & reasoning, respectively. And, this matches what Chollet said about geometry-based cognition (which is perception) versus topology-based cognition (which is reasoning)

—

The picture I’ve outlined is essentially the physical interpretation what homotopy type theory and the univalence axiom says. And I think that’s it. That’s everything.

]]>Also, I was looking at revenue from quantum computer use in a recent Gartner report, and D-Wave as of Q4 owns 92.2% of worldwide QC revenue. So there’s that.

]]>