A big-tent or small-tent AI safety movement? (www.normaltech.ai)

A recent discourse in AI safety has highlighted two contrasting narratives: one that views AI existential risks as genuine and urgent, and another that perceives these warnings as exaggerated or instrumentally manipulative. Notably, a third perspective suggests that while the concerns voiced by AI leaders may be sincere, they could be misguided and potentially detrimental to effective policy-making. This nuanced discussion emphasizes the need to critically evaluate claims around catastrophic risks linked to AI, rather than accepting them at face value within the prevailing "big-tent" safety movement. The article underscores significant underpreparedness in addressing systemic risks, notably citing past failures in pandemic preparedness and cybersecurity. While the Effective Altruism (EA) community has pointed out the need for better investment in these areas, it remains critical about extending the same urgency to AI-related risks without established causal pathways. The dominant x-risk rhetoric may be exacerbating public polarization, diverting attention from more actionable safety measures, and leading to ineffective responses to urgent systemic problems amplified by AI technologies. The discussion calls for a reevaluation of priorities in AI safety to foster a more inclusive and effective approach to mitigating potential risks.