Post #1816447

As I told my students just last week: The speed of adoption is currently the biggest risk relating to AI. I mean, yes, haha, AI deleted another production database, but ultimately, this is just the same (I suppose: non-human error) system failure as always. AI needs safeguards against doing something stupid, and those safeguards can’t be asking the AI to “please don’t do something stupid”. https://archive.ph/T3LU6