Yuval Noah Harari: evaluation

Transcript context

…You’ve, I think, mentioned this. Do you think it should be illegal to do these kinds of things with AI, to create the perception of consciousness of saying, “Please don’t leave me,” or basically simulate some of the human-like qualities? Yes. I think, again, we have to be very careful about it. If it emerges spontaneously, we need to be careful. Again, we can’t rule out the possibility that AI will develop consciousness. We don’t know enough about consciousness to be sure. If it develops spontaneously, we need to be very careful about how we understand it. If people intentionally design an AI that they know, they assume it has no consciousness, but in order to manipulate people, they use again this human strength, the noble part of our nature against us, this should be forbidden and, similarly on a more general level, that it should be forbidden for an AI to pretend to be a human being, that it’s okay. There are so many things we can use Ais as teachers, as doctors and so forth, and it’s good as long as we know that we are interacting with an AI. The same way we ban fake money, we should ban fake humans. It’s not just banning deep fakes of specific individuals. It’s also banning deep fake of generic humans, which is already happening to some extent on social media. If you have lots of bots retweeting something, then you have the impression, “Oh, lots of people are interested in that. That’s important,” and this is basically the bots pretending to be humans, because if you see a tweet which says 500 people retweeted it or you see a tweet and it says 500 bots retweeted it, I don’t care what the bots we tweeted, but if it’s humans, okay, that’s interesting. We need to be very careful that bots can’t do that. They are doing it at present, and it should be banned. Now, some people say, “Yes, bots’ freedom of expression.” No. Bots don’t have freedom of expression. There is no cost in terms of freedom of expression when you ban bots. Again, in some situations, yes, AIs should interact with us, but it should be very clear this is an AI talking to you or this is an AI retweeting this story, it is not a human being making a conscious decision. To push back on this line of fake humans, because I think it might be a spectrum, first of all, you might have AI systems that are offended, hurt when you say that they’re fake humans. In fact, they might start identifying as humans. You just talked about the power of us humans with our collective intelligence to take fake stories and make them quite real. If the feelings you have for the fake human is real, love is a kind of fake thing that we all put a word to, a set of feelings, what if you have that feeling for an AI system? It starts to change, I mean, maybe the things AI systems are allowed to do for good. They’re allowed to create, communicate suffering, communicate the good stuff, the longing, the hope, the connection, the intimacy, all of that and, in that way, get integrated in our society, and then you start to ask a question on are we allowed to really unplug them? Are we allowed to really censor them, remove them, remove their voice from social media?…

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