tech
I’d Rather Risk Cancer Than See AI Move This Fast
I’d benefit if AI cured cancer. And I still want AI progress to slow down.
TL;DR
- The author, an AI professor with a high genetic risk for cancer, reflects on her mentor's vision of AI curing cancer.
- She questions the timeline and feasibility of AI curing cancer soon, citing limitations in AI's applicability to biological systems compared to domains like chess or coding.
- Existing AI tools are powerful but underutilized, suggesting that advancements in AI are not solely dependent on creating more powerful systems.
- The rapid release and subsequent issues with Anthropic's Fable 5 model highlight society's unpreparedness for the broader repercussions of advanced AI.
- The author argues that speeding up societal adaptation to AI is not a viable solution and that slowing AI development is necessary to address potential risks like unemployment, inequality, and surveillance.
- Ultimately, she prioritizes a careful approach to AI development, even if it means delaying a cancer cure, to safeguard human meaning and societal well-being.