Thread by David Deutsch
- Tweet
- Mar 29, 2023
- #ArtificialIntelligence
Thread
Not signing. Overall, it reads to me like "we shouldn't make anything whose effects we can't prophesy, so let's begin by giving the world's totalitarian governments and organised criminals a chance to catch up".
Incidentally, it implicitly confuses AI with AGI.
Incidentally, it implicitly confuses AI with AGI.
We should expect problems. But I see no reason why we wouldn't still be facing them, only worse, in 6 months. And meanwhile we'd have forgone 6 months of benefits. Including, down the line, forgoing lives saved.
That's if the effect is as profound as many expect. That, too, is unknowable.
Darwin was right that the effect of his knowledge would be profound, unpredictable, and could include existential risk to civilisation. But wrong to delay its publication.
Darwin was right that the effect of his knowledge would be profound, unpredictable, and could include existential risk to civilisation. But wrong to delay its publication.