Yuval Noah Harari: belief

Transcript context

…Well, I mean, I guess, instructive there is just having the humility to look around, to think about living beings that operate at different timescale, at different spatial scale. I think that’s all useful when starting to analyze artificial intelligence. It’s possible that even the larger language models we have today are already conscious. I highly doubt it, but I think consciousness, in the end, it’s a question of social norms. Because we cannot prove consciousness in anybody except ourselves. We know that we are conscious, because we are feeling it. We have direct access to our subjective consciousness. We cannot have any proof that any other entity in the world, any other human being, our parents, our best friends, we don’t have proof that they are conscious. This has been known for thousands of years. This is Descartes, this is Buddha, this is Plato. We can’t have this sort of proof. What we do have is social conventions. It’s a social convention that all human beings are conscious. It also applies to animals. Most people who have pets firmly believe that their pets are conscious, but a lot of people still refuse to acknowledge that about cows or pigs. Now, pigs are far more intelligent than dogs and cats, and according to many measures, yet, when you go to the supermarket and buy a piece of frozen pig meat, you don’t think about it as a conscious entity. Why do you think of your dog as conscious, but not of the bacon that you buy? Because you’ve built a relationship with the dog and you don’t have a relationship with the bacon. Now, relationships, they don’t constitute a logical proof for consciousness, they’re a social test. The Turing test is a social test, it’s not a logical proof. Now, if you establish a mutual relationship with an entity and you are invested in it, emotionally, you are almost compelled to feel that the other side is also conscious. When it comes again to AI and computers, I think, again, I don’t think that at the present moment computers are conscious, but people are already forming intimate relationships with AI and are therefore almost … It’s almost irresistible. They’re compelled to increasingly feel that these are conscious entities. I think we are quite close to the point when the legal system will have to take this into account. Even though I don’t think computers have consciousness, I think we are close to the point the legal system will start treating them as conscious entities, because of this social convention. To you is a social convention, just a funny little side effect, a little artifact, or is it fundamental to what consciousness is? Because if it is fundamental, then it seems like AI is very good at forming these kinds of deep relationships with humans, and therefore it’ll be able to be a nice catalyst for integrating itself into these social conventions of ours.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.