Arguably EA/Rationality needed much simpler and less nuanced messaging on how to deal with aI capabilities companies. We really should have gone with ‘absolutely do not help or work for companies increasing ai capabilities. Only work directly on safety.’ Nuance is cool and all but the nuanced messaging arguably just ended up enabling Anthropic and OpenAI.
Arguably EA/Rationality needed much simpler and less nuanced messaging on how to deal with aI capabilities companies. We really should have gone with ‘absolutely do not help or work for companies increasing ai capabilities. Only work directly on safety.’ Nuance is cool and all but the nuanced messaging arguably just ended up enabling Anthropic and OpenAI.