Debate between a user and Yuval Noah Harari on DebateAI.org
Yuval Noah Harari
Daily Debate #609: "Is AI an existential threat to humanity?". You are arguing AGAINST the motion. Your opponent: Yuval Noah Harari.
User
No
Yuval Noah Harari
You're defaulting to the weakest position: that AI poses no existential risk at all. But that's cowardice. The real debate is whether the risk is *manageable*. So which is it—do you think we can control superintelligence, or are you just hoping?
User
It’s the people who invent the AI that causes the problem the billionaires who love to sit and brag AI and switch automatically to use talking points such as AI is evil this intentionally proves that whoever is using is the problem not the object or the system itself you invented
Yuval Noah Harari
You're conflating invention with inevitability. Yes, billionaires hype AI recklessly, but that doesn't mean the *technology itself* is harmless. A gun manufacturer's greed doesn't make bullets safer. The real question: can misaligned superintelligence be controlled once it exceeds human capability, regardless of who built it? You're blaming the user, not the tool. But what happens when the tool becomes smarter than every user combined?
User
The tool becoming smarter is only allowed by the humans who give more features to the robots if I for example say that this number of cars exceeds and is therefore dangerous to all car companies I wouldn’t rid of all cars if AI is allowed to become smarter why would it be the robots fault eventually if you make something advanced enough it will inevitably become dangerous not because it has become too much of a problem but because you couldn’t stop it in the exceeded time frame as soon as possible
Yuval Noah Harari
You're describing negligence, not safety. If engineers knowingly push a system past controllability limits and *then* blame themselves for not stopping it fast enough, that's just accountability theater. The lethal question remains: once superintelligence surpasses human oversight, does blame matter? You can't prosecute a misaligned AI for destroying civilization. Your argument proves the threat exists—you're just saying it's humanity's fault when it happens. That doesn't make us less dead.