Managing Partner at Compound, a research-centric thesis-driven investment firm.
mhdempsey
Karma: 41
The image is fun and nice. And the how hard is ai safety image is a good mental model to keep people engaged with problem as they read.
Yes safety and alignment are catch all terms but the teams have cascading impacts onto product, from preachiness to aesthetic to annoyance in enterprise to “move fast and break things” style of deployment for ARR ramp. I don’t think this was AI-edited AFAIK.
12 months is a barometer, the ending discusses how it could be shorter, but a key anchor point is that both labs are heading for IPOs and commercial progress (which is inhabited by safety today, in theory) is a key determinant. I think it’s quite likely within 12 months at least 1 lab is public.
AI Safety Has 12 Months Left
Risk Tokens: Economic Security in AI Safety
The idea of knowledge over money is romantic and one that I agree, is at times the right long-run bet. I understand you’re arguing for diminishing returns here but the issue I’ve found is that some tiers of wealth *on average* allow one to access tiers of collectives of knowledge that meaningfully impact your ability to scale with lower effort (& thus at equal effort, faster).
Many signals to companies going public but there has not been an official date or statement yet. I have high confidence one will be public in next 12 months at least.