zlacker

>>aliabd+(OP)
Ilya just self-immolated the hottest company of the decade with an 80 billion dollar valuation, all to make a point that he thinks AGI is too close and that safety needs to come first before all else. That’s quite the statement!

>>dougmw+d3
At this point whenever I see someone concerned about "safety" with respect to AI, I assume it's either someone who wants to:

a) Make sure it's sufficiently neutered as to not say something offensive or incongruent with certain ideologies, make certain topics off-limits

or

b) Scare lawmakers into making it impossible for anyone to catch up to OpenAI et al, so they can reap the benefits of cutting-edge AI while keeping the public placated with incremental improvements to outdated and sufficiently hobbled models