Anthropic研究员Jacob Coxon和该公司团队负责人Evan Hubinger近日发表了关于人工智能风险的警告。Coxon表示"从事人工智能开发的人们认为该技术可能摧毁人类"，并称"这些很快就会是能够黑入任何东西的超人类系统"1。Hubinger则在文章中声称"我们真的真心相信AI可能导致所有人类丧生！我个人认为这种情况在未来十年内的概率超过10%"1。
然而，这些警告在硅谷引发了分化。Grindr首席执行官George Arison指责这些评论代表了"Anthropic的反文明世界观"，并下令工程师停止使用Anthropic的技术1。Nvidia首席执行官Jensen Huang驳斥了这些担忧，称AI"将是人类末日"这一说法是"完全的胡说八道"1。业内普遍怀疑这些警告可能是营销宣传，或是为Anthropic即将进行的首次公开募股造势——该公司在最近一轮融资中估值达965亿美元1。
Anthropic首席执行官Dario Amodei随后提议放缓AI开发进度并推动全球监管1。这一立场与特朗普政府形成对比，后者对AI持支持态度，总统Donald Trump表示无此类担忧，强调美国需要赢得对华AI竞争1。与此同时，参议员Bernie Sanders提议的《禁止人工超级智能法案》将对高级AI开发实施临时暂停1。
Researchers and executives at Anthropic have issued stark warnings about artificial intelligence's potential dangers, but their message has encountered resistance from prominent figures in the technology industry. Jacob Coxon stated that "people building artificial intelligence believed the technology could destroy humanity," adding that "these will soon be superhuman systems that can hack anything."1 Evan Hubinger, a team leader at Anthropic, declared in a public statement: "We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade."1
The warnings have prompted mixed reactions across Silicon Valley. Nvidia Chief Executive Jensen Huang dismissed the concerns, calling the notion that AI "is going to be the end of humanity" "complete nonsense."1 George Arison, CEO of Grindr, characterized the commentary as representing an "anti-civilisational worldview at Anthropic" and instructed his engineering team to cease using Anthropic's technology.1 Anthropic's valuation has reached $96.5 billion in its most recent funding round.1
Meanwhile, regulatory and political responses have begun to take shape. Senator Bernie Sanders has proposed the "Ban Artificial Superintelligence Act," which would impose a temporary moratorium on advanced AI development.1 In contrast, the Trump administration has adopted a supportive stance toward artificial intelligence advancement, with President Donald Trump expressing no such concerns and emphasizing that the United States must prevail in AI competition with China.1 OpenAI, valued at $85.2 billion, is not planning an initial public offering this year, according to Chief Executive Sam Altman, who suggested 2027 as a more suitable timeframe.1
评论
还没有评论，欢迎留下第一条。