Hinton talked with Steven Bartlett (not to be confused with the Bartlett and Cochrane coefficients), and
Hinton talked with Steven Bartlett (not to be confused with the Bartlett and Cochrane coefficients), and very patiently explained, answering the questions of a man who published…
Hinton talked with Steven Bartlett (not to be confused with the Bartlett and Cochrane coefficients), and very patiently explained, answering the questions of a man who published… Hinton talked with Steven Bartlett (not to be confused with the Bartlett and Cochrane coefficients), and very patiently explained, answering the questions of a man who published a book with a gold cover and big gold letters in caps THE DIARY OF A CEO, why AI is dangerous, how we can prepare for it, and why Yann LeCun and the rest of the industry talk about it less. The points are the usual ones, thank goodness Hinton exists: - The military part of AI is exempt from regulation even by the European AI Act that gets in everyone's way, so those German humanoid robots at 16,000 euros each might end up on the battlefield with a six-barreled machine gun on their shoulder before the Chinese ones; they might even be called Arnold in honor of the great Austrian. - Humans + AI are stronger than just AI; bad people + AI is very, very bad. Hinton gives a good benchmark: a medium-sized cult can afford a biological weapon, smaller than Aum Shinrikyo, and of course smaller than, say, the Syrian Alawites - We have never been in a situation where there's something smarter than us on the planet. Stronger, yes. Longer fangs, yes. Heavier and falls off a cliff impressively when scared — yes. Smarter — no. All in all, an hour and a half of conversation about AI in the way it's usually not talked about — with a good understanding of the situation and without much optimism. Host: Ark Business Where: Online · Telegram / YouTube Materials - ▶ Recording - 🔗 neura-robotics.com View the original announcement →