Zuckerberg氏、オープンAIはAGI安全性より優れると主張 — 米国は規制緩和すべきと訴え
Zuckerberg氏は6500語のマニフェストを発表し、外国のライバルに打ち勝ち、AIを数十億人へ民主化するためにオープンソースのスーパーインテリジェンスを主張。データ利用とモデル蒸留に関する規制緩和を米国に明確に求めた。
meta ceo mark zuckerberg published a 6,500-word manifesto outlining his vision for open-source superintelligence, arguing that delivering 'personal superintelligence' to billions and small businesses will drive economic prosperity. he positions meta as a responsible steward, pledging to counter pr crises by investing in local energy infrastructure and high-paying jobs where it builds datacenters. zuckerberg explicitly urges the us government to reduce oversight, warning that current restrictions on data use and model distillation prevent the us from competing with foreign labs that 'hold several advantages.' this signals meta's intent to commoditize advanced ai through open releases while lobbying for a deregulated environment to train massive models.