Pause AI was relatively small in scale. I feel like AI is in great need of protest. Protesting for increased regulation and safety, layoff compensation, etc.
A lot of what EA wants in terms of AI can be protested for.
I feel like the EA community should protest more? What do you think?
I'm also wondering why the broader community hasn't had a SINGLE AI safety protest. 10% Extinction risk is pretty well repeated in the mainstream online community.
EA is a fairly small and weird social movement in the grand scheme of things. A protest movements consisting only of EAers will produce pathetically small protests, which might get some curious media writeups but will be unlikely to scare or influence anybody.
If you actually want a big protest movement, you have to be willing to form coalitions with other groups. And that means playing nice with people like AI ethicists, rather than mocking and attacking them as has been unfortunately common here.
Hello titotal. I'd prefer you didn't refer to small protests as "pathetically small". LMK if it isn't obvious why.
Can you give examples of *mockery* of AI ethicists on the forum, if that's where you mean by "here"? For sure, I think a lot of people here (including me to some degree, although "AI ethics" is a very broad thing no doubt including lots of stuff I do agree with) are not the biggest fans. And I agree that building bridges would be needed for a useful protest movement. I'm sure people say very mean things in private or over Slack. But on the forum itself people are generally pretty careful in what they say in my experience, but also, when they are harshly cr... (read more)
Thank you for your response. My impression is that big or small, every individuals additional contribution to a protest is roughly proportional to the impact of the protest. This meaning that its just as impactful for people to have small scale protests.
Good question. My hunch is that EA as a culture tends to prioritize epistemic and ethical sophistication and rigor, over direct 'political' action. And has traditionally avoided getting involved in issues that seem 'intractable' by virtue of being highly controversial and potentially partisan.
Against that background of EA's rather 'ivory tower' ethos, any direct protests may tend to be seen as rather simplistic, strident, and undignified -- even for issues such as animal agriculture where there's pretty strong EA consensus that factory farming is unethical.
But I think it's time for EAs to climb down from our AI safety debates, recognize that the leading AI companies are not actually prioritizing safety, and start getting more involved in social media activism and in-person protests.
Thank you for your response! Along with the EA community, I too am scared of doing activism for something controversial and bizarre like AI safety
Protests are usually done by those in dire need of change: minorities, poor people, people whose identity is attacked, etc. AI risks are overwhelmingly highlighted by rich white male engineers: not those who usually have a reason to go out in the streets. And as Geoffrey says, who despise those who do--it's easier to mock those who struggle when you don't, assuming that they make unnecessary noise because you don't feel at all part of their fight.
And now EAs realize that profit is taking over safety concerns--it took a lot of time! It was painful to read Altman's praising until the board shuffling at OpenAI. It's been years that people protest because greed and unequal distribution of money make their own lives poorer and harder; but now greed causes survival risks that also extend to rich engineers, so they have to do something.
Protests are usually done by those in dire need of change: minorities, poor people, people whose identity is attacked, etc.
I think this is not true. Even the 2020 BLM protests, which skewed unusually non-white, were almost 50% non-hispanic white according to Pew, the first result I found when googling the question. This matches my impression that protests actually tend to be dominated by relatively well educated white people, which makes sense, because these groups dominate most forms of political activity.
In general EAs share a lot of demographic factors with typical (US) protestors:
They diverge in some others, for example EAs being concentrated in STEM graduates over other less mathematical degrees.
minor edit: some ambiguity in how Pew is handling missing data, doesn't affect the bottom line, added second section.
Can you give examples of *mockery* of AI ethicists on the forum, if that's where you mean by "here"? For sure, I think a lot of people here (including me to some degree, although "AI ethics" is a very broad thing no doubt including lots of stuff I do agree with) are not the biggest fans. And I agree that building bridges would be needed for a useful protest movement. I'm sure people say very mean things in private or over Slack. But on the forum itself people are generally pretty careful in what they say in my experience, but also, when they are harshly critical of AI ethics, they almost never in my experience use "mockery" as a weapon. Their tone is usually pretty earnest. Also maybe my experience was an outlier here, but the one time I posted a comment about a prominent AI ethics person that was relatively harsh (I think with some justification), it got deleted by the mods on the grounds that they didn't want to "undermine the potential for further engagement" , which doesn't suggest to me that this stuff is just running wild.
To be clear, I actually think AI Ethics peoples have many substantively reasonable criticism of EA views on both X-risk and on the tolerance for right-wing extremism on race in some EA/rationalist circles in the Bay. But that is a separate question from "are EAs too mean to ethics folks", and I feel like I haven't seen much of that?