Charlie—I appreciate your point about the lack of funding for AI-related protests.
There seems to be a big double standard here.
Many EA organizations are happy to spend tens of millions of dollars on ‘technical AI alignment work’, or AI policy/governance work, in hopes that they will reduce AI extinction risk. Although, IMHO, both have a very low chance of actually slowing AGI development, or resulting in safe alignment (given that ‘alignment with human values in general’ seems impossible, in principle, given the diversity and heterogeneity of human values—as starkly illustrated in recent news from the Middle East.)
But the same EA organizations aren’t willing, yet, to spend even a few tens of thousands of dollars on ‘Pause AI’ protests that, IMO, would have a much higher chance of sparking public discourse, interest, and concern about AI risks.
Funding protests is a tried-and-true method for raising public awareness. Technical AI alignment work is not a tried-and-true method for making AI safe. If our goal is to reduce extinction risk, we may be misallocating resources in directions that might seem intellectually prestigious, but that aren’t actually very effective in the real world of public opinion, social media, mainstream media, and democratic politics.
Lightspeed Grants and my smaller individual donors should get credit for funding me to work on advocacy which includes protests full-time! Sadly afaik that is the only EA/adjacent funding that has gone toward public advocacy for AI Safety.
Charlie—I appreciate your point about the lack of funding for AI-related protests.
There seems to be a big double standard here.
Many EA organizations are happy to spend tens of millions of dollars on ‘technical AI alignment work’, or AI policy/governance work, in hopes that they will reduce AI extinction risk. Although, IMHO, both have a very low chance of actually slowing AGI development, or resulting in safe alignment (given that ‘alignment with human values in general’ seems impossible, in principle, given the diversity and heterogeneity of human values—as starkly illustrated in recent news from the Middle East.)
But the same EA organizations aren’t willing, yet, to spend even a few tens of thousands of dollars on ‘Pause AI’ protests that, IMO, would have a much higher chance of sparking public discourse, interest, and concern about AI risks.
Funding protests is a tried-and-true method for raising public awareness. Technical AI alignment work is not a tried-and-true method for making AI safe. If our goal is to reduce extinction risk, we may be misallocating resources in directions that might seem intellectually prestigious, but that aren’t actually very effective in the real world of public opinion, social media, mainstream media, and democratic politics.
Lightspeed Grants and my smaller individual donors should get credit for funding me to work on advocacy which includes protests full-time! Sadly afaik that is the only EA/adjacent funding that has gone toward public advocacy for AI Safety.