比尔·盖茨警告人工智能已突破多个危险阈值

比尔·盖茨近日在接受采访时发出警告，称人工智能已经突破了多个应该被重点关注的危险阈值1。这些阈值包括生物能力、网络能力、心理社会能力、就业市场破坏能力，以及对AI的控制缺失1。盖茨对当前社会对这一问题的关注程度表示失望，指出"我惊讶于业界以外对此的关注和讨论是如此之少"1。

在安全风险评估方面，盖茨认为生物恐怖主义风险尤为严峻，其危险程度约为自然疫情风险的50倍1。他主张对能够制造新分子的AI模型进行监管1。针对AI可能带来的就业冲击，盖茨指出人工智能能够以远低于人类劳动成本的价格替代白领工作，且错误率可能低于人类1。他提议征收代币税和机器人税以帮助失业者1，认为这次技术革命不同于以往——虽然历史上没有技术导致净就业减少，但盖茨警告这次情况可能不同1。

Bill Gates has cautioned that artificial intelligence has surpassed multiple critical safety boundaries that demand urgent attention and action.1 Speaking with MIT Technology Review, Gates identified breaches across biological capabilities, cyber-capabilities, psychosocial capabilities, job-market-destruction capabilities, and a lack of control as key areas of concern.1 He emphasized the severity of bioterrorism risks posed by AI, stating that he views bioterrorism risk as approximately 50 times more frightening and more likely than natural pandemic risk.1

Gates has called for immediate regulatory and policy responses to address these emerging threats.1 He advocated for government monitoring of any artificial intelligence model capable of designing novel molecules.1 On the employment front, Gates warned that AI can now substitute for white-collar work at costs far below human labor expenses, potentially with error rates lower than humans achieve.1 To mitigate job displacement, he proposed implementing taxation mechanisms on tokens and robots, along with establishing programs to preserve employment for human workers.1

Gates expressed frustration with what he perceives as insufficient awareness of these risks beyond the technology industry itself.1 He noted that current economic statistics may be misleading regarding the unprecedented nature of this technological transition, suggesting that unlike previous technological revolutions, this disruption may result in net job losses.1 Gates stressed that effective solutions require collaborative efforts between government and industry to develop and implement adequate safety measures.1