Raw LLM Responses

Great interview, Reuben. I've heard many of Deutsch's arguments from other people, so it's nice to hear them from him directly. However, he fails to answer some of your questions to a satisfactory level, and makes some leaps of logic. "Then they would be people"... and goes on to make assumptions about how AIs would be similar to humans, based on him having defined them as people. He objects to the orthogonality thesis, but the reasons are lacking. If he thinks these essays are convincing, then just give us the best arguments. He gave you an argument, and then when you werent convinced, he was essentially like "but that's not the best argument". Literally all he would need to do to debunk most AI doom scenarios would be to say "the orthogonality thesis is wrong for the following reasons", then give them (and defend them). Saying "read this and you’ll agree with me" is a convenient way of not having to defend your point of view. If the person reads it and doesn't agree, you can just claim that they don't "get it". But if you "got it", you could have just explained it in the first place. He tends to be very indirect when answering your questions. Reuben, props on the pushback. It is clear that you are holding back out of respect and politeless.