Reconciling Effective Altruism and E/acc

🌈 Abstract

The article discusses the tensions and disagreements between Effective Altruism (EA) and Effective Accelerationism (E/acc), and how they can unite against the AI ethics tribe.

🙋 Q&A

[01] Reconciling Effective Altruism and E/acc

1. What are the key differences between EA and E/acc?

• EA is more focused on top-down planning and legibility, while E/acc is about bottom-up, market-based adversarial competition as a check on centralized control.
• EA is seen as more affiliated with the Democratic party, while E/acc is more right-coded.
• EA tends to have a more polite, debate-oriented approach, while E/acc is more confrontational and unequivocal in defending tech.

2. What are the key areas of agreement between EA and E/acc?

• They agree on most issues other than AI safety, being mostly techno-optimists and accelerationists.
• They both need to unite against the AI ethics tribe, whose concerns are better appreciated by regulators.

3. Why does the author believe EA and E/acc should reconcile?

• EA needs someone to defend it, as it won't defend itself.
• E/acc needs EA to be able to self-critique and self-regulate until it can figure out AI alignment.
• Together, they can present a united front against the AI ethics tribe, while privately disagreeing on the best approach to AI alignment.

[02] Uniting against the AI ethics tribe

1. What are the concerns of the AI ethics tribe?

• They are concerned about broader inequality and perpetuation of group disparities, as well as issues like misinformation and hate speech.

2. How do the concerns of the AI ethics tribe differ from those of EA and E/acc?

• EA and E/acc are not as concerned about the problems of capitalism that the AI ethics tribe is focused on.

3. Why does the author believe EA and E/acc should band together to fight the AI ethics tribe?

• The AI ethics tribe's concerns are better appreciated by regulators, who are most likely to set strangling AI regulations.
• EA and E/acc need to present a united front in denouncing concerns that could prematurely strangle AI development, while privately hashing out the best approach to AI alignment.

</output_format>