THE WORLD’S artificial-intelligence superpowers are beginning to throw their weight around(炫耀实力，仗势欺人). The American government’s short-lived(短暂的，持续不久的) decision to withhold foreigners’ access to Fable, a whizzy(炫酷的，技术先进的) AI model from Anthropic, reflected not just a desire for control over the technology but also fear of the dangers it could pose in the wrong hands. Now China is said to be considering curbing overseas access to its most advanced models. Yet for AI to be harnessed(驾驭，利用) safely, the industry needs a plan that is coherent(连贯的，清晰的) and predictable(可预测的). Sir Demis Hassabis, the co-founder of Google DeepMind, thinks he has one.
全球人工智能超级大国开始炫耀实力。美国政府曾短暂决定限制外国人使用Anthropic公司推出的炫酷AI模型Fable，这不仅反映了对技术控制的渴望，也体现了对技术落入不法之手可能带来危险的担忧。如今，据说中国正在考虑限制其最先进模型向海外开放。然而，要安全驾驭AI，行业需要一套连贯且可预测的方案。谷歌DeepMind联合创始人戴米斯·哈萨比斯爵士认为他有一个方案。
Sir Demis has sketched out(大致勾勒，简要描述) proposals for AI regulation in the past. His latest attempt—published online on July 14th and elaborated(详细阐述) in an interview with The Economist—is more concrete(具体的，明确的). The American government, he says, should develop a system for testing the safety of new AI models before they are released. “It’s important that it’s not just an industry body,” he adds. But a regular government agency wouldn’t do either. “It would not be able to move fast enough, or have the right resources.” Instead, Sir Demis suggests taking inspiration from FINRA, the Financial Industry Regulatory Authority, a private agency in America that regulates(监管，管理) brokers and stock markets.
哈萨比斯爵士过去曾勾勒过AI监管方案。他最新的尝试——7月14日在线发布并在接受《经济学人》采访时详细阐述——更为具体。他表示，美国政府应建立一套系统，在新型AI模型发布前测试其安全性。“重要的是不能仅仅是一个行业机构，”他补充道。但普通的政府机构也不行。“它无法做到足够快速反应，也缺乏适当资源。”相反，哈萨比斯爵士建议从FINRA（美国金融业监管局）汲取灵感，这是一家私人机构，负责监管美国经纪商和股票市场。
It is a pragmatic proposal. Gone are the optimistic hopes, still held by some of his peers, of first establishing international consensus: America, he says, must lead the way. Agreement is not required because other countries will fall in line(服从，保持一致) in order to maintain access to American technology and the vast American market. One day that could even extend to China, helping to avert a race to the bottom(逐底竞争) on safety. Although Sir Demis thinks the scheme should initially be voluntary for model-makers, his expectation is that it will end up mandatory(强制性的). FINRA, by way of example(作为例子), has hard enforcement powers(硬性执法权力) delegated(授权，委托) to it by the state.
这是一个务实的提案。一些同行仍抱有先建立国际共识的乐观希望，但那种希望已经破灭：哈萨比斯说，美国必须带头。不需要达成协议，因为其他国家为了继续获取美国技术和庞大的美国市场，会纷纷效仿。有朝一日，这甚至可能扩展到中国，有助于避免安全领域的逐底竞争。尽管哈萨比斯认为该计划最初应对模型制造者自愿实施，但他预计最终会变成强制性的。以FINRA为例，它拥有国家授权的强大执法权力。