Anthropic研究员Jacob Coxon因对公司风险管理的担忧而离职，指责该公司及其他领先AI企业正在"拿我们的生命赌博"1。这一举动引发了业界对AI技术生存风险的广泛讨论。随后，Anthropic的对齐负责人在社交媒体上发表观点，声称"我们真诚地相信AI可能杀死全人类"，并估计这一风险在十年内的概率超过10%1。
这些警告言论随即招致科技界的质疑。由于Anthropic即将启动IPO申报流程，外界对这些风险声明与企业融资目的之间的关系提出了疑问1。讨论人士质疑，这些"末日警告"是否本质上属于企业展示其AI技术能力先进性的营销策略1。ControlAI执行主任Connor Leahy也参与了关于AI控制问题的相关讨论1。
Jacob Coxon, a researcher at Anthropic, departed from the company citing concerns that leading artificial intelligence firms are "gambling with our lives" 1. His resignation has intensified industry scrutiny over whether AI technology poses an existential threat to humanity and whether such warnings serve legitimate safety purposes or corporate interests.
The controversy deepened when Anthropic's alignment lead posted on social media that "we genuinely believe AI could kill all humans" and estimated the probability at over 10 percent within the next decade 1. These statements have prompted skepticism within the technology sector about the authenticity of such "doomsday warnings" and their potential connection to Anthropic's forthcoming initial public offering 1. Questions have emerged regarding how the company will disclose these risk statements in its IPO filing 1. Some observers have suggested these warnings may function as marketing tactics to demonstrate the sophistication and power of AI capabilities 1. Connor Leahy, executive director of ControlAI, has participated in ongoing discussions about AI safety and control 1.
评论
还没有评论，欢迎留下第一条。