Three out of these four theories have an interesting pattern to them. With good explanations in epistemology, we’re saying conjectures and their refutations and error correction is how we improve knowledge. With genetic evolution, genetic mutations, variation and natural selection weed out the ones that didn’t work. Then there’s memetic evolution on top of that, where we have ideas and then criticism weeds out the ideas that don’t work. More

Popper has a book called “Objective Knowledge,” and it’s subtitled “An Evolutionary Approach.” And that’s no accident, either. There’s a symmetry between the theory of epistemology and the theory of evolution as we understand it. Before we understood what is known as the Darwinian theory of evolution, the only idea that people had was that these entities had to be created. More

In almost all cases, you only ever have one theory on offer. In the case of gravity, there literally is only one theory on offer at the moment. There’s general relativity. Previously we did have two theories. We had Newtonian gravity and we had general relativity—but we did a crucial experiment. More

Initial guesses at what ‘knowledge’ was all about amounted to what is known as the “justified true belief” vision of knowledge, and it’s still the most prevalent idea today. Anyone who calls themself a Bayesian is a justified true believer. This is the misconception that knowledge is about trying to justify your beliefs as true. More

If you’re an academic, being able to explain all of the problems that are out there and how dangerous these problems are and why you need funding to look at them in more depth appears to be the intellectually serious position; whereas, someone who claims that we can solve it sounds a little bit kumbaya. More

A lot of the theories as to why we’re imminently going to create an AGI are based in a naïve extrapolation of computational power. It’s almost an induction of more and more computational power. They say, “AI has already gotten good at vision and beating humans at chess and at video games; therefore, it’s going to start thinking soon.” More

There’s also Solomonoff’s theory of induction. I don’t know if you’ve looked at that at all? I’ve heard of it, but I haven’t investigated it. I’m going to mangle the description. It says that if you want to find a new theory that explains why something is happening—in this case something that’s encoded as a binary string—then the correct one is a probability-weighted theory that takes into account all possible theories and weighs them based on their complexity. More

There’s some deep symmetry between multiverse theory and Feynman path integrals, right? You’re absolutely right. Feynman believed in multiple histories, but it’s an open question whether he thought these were actually physically real things or merely mathematical objects. He was relatively silent on the matter. Feynman was a realist and an absolute genius—probably the second greatest physicist of the 20th century after Einstein—but he made one of the worst quips. More

Going even further, it’s not just science. When we look at innovation, technology and building—for example, everything that Thomas Edison and Nikola Tesla did—this came from trial and error, which is creative guesses and trying things out. If you look at how evolution works through variation and then natural selection, it tries a lot of random mutations and filters out the ones that didn’t work. More

There’s another example like this. You can do this with a saucepan at home. Put a beaker of water on a heat source, then put a thermometer into that water and turn on your heat source. As time passes, record the temperature of the water.  You’ll notice the temperature of water increase. More