One might think that, to prepare for a country of geniuses in a datacenter, we’d be making great use of our smart fellas in a datacenter
Alas, AI for good / defensive acceleration etc. are afterthoughts in the big scheme of things still, more impactful as PR than risk mitigation
Many users criticized AI safety work as cynical PR afterthought because companies prioritize payouts and private data-center control over real preparation for automation risks, though one remained bullish if efforts increased.
No Digg Deeper questions have been answered for this story yet.
I remain bullish about the potential if we tried harder but we aren’t really (yet) https://milesbrundage.substack.com/p/reasons-that-ai-can-help-with-ai
One might think that, to prepare for a country of geniuses in a datacenter, we’d be making great use of our smart fellas in a datacenter
Alas, AI for good / defensive acceleration etc. are afterthoughts in the big scheme of things still, more impactful as PR than risk mitigation
@Miles_Brundage my concern with the data center geniuses framing remains that nobody should be able to own and control it privately.
hard to imagine a scenario where it's globally good for society.
the govs should definitely have plans for managing power concentration before rather than after.
@Miles_Brundage the pr angle runs deep when the payout timeline is "hopefully never"
@Miles_Brundage PR beats prep every time until the datacenter starts automating something we werent ready for
@Miles_Brundage the challenge might be how to balance innovation with actual application; too often, tech gets hyped without a clear framework for real-world impact, leaving potential benefits untapped.