Timothy Garton Ash is right that serious artificial intelligence risks demand international action, but “AI Hiroshima” may be the wrong picture(蒂莫西·加顿·阿什说得对，严重的人工智能风险需要国际行动，但“人工智能广岛”可能是个错误的比喻。) (Would even an AI disaster on the scale of Hiroshima be enough to make humankind protect itself? I fear not, 22 August). Hiroshima was not technology going rogue(失控的；行为失常的). Human beings designed the bomb, authorised its use and dropped it. The technology worked much as intended(基本按预期地).
蒂莫西·加顿·阿什说得对，严重的人工智能风险需要国际行动，但“人工智能广岛”可能是个错误的比喻（即使发生广岛规模的人工智能灾难，是否足以让人类保护自己？恐怕不会，8月22日）。广岛并非技术失控。人类设计了炸弹，授权使用并投下了它。技术基本按预期运行。
AI may one day behave in ways we cannot control, but many of the gravest harms could happen without any dramatic moment when “the AI takes over(接管；控制)”. AI might help design a pathogen(病原体), find a vulnerability(漏洞；弱点) in critical infrastructure or improve a weapons system while people still decide what happens next. Other harms may build almost invisibly through thousands of ordinary decisions – more autonomy here, one safeguard removed there, another consequential task handed over because the system has worked well so far. By the time the danger is obvious, the important decisions may already have been made.(等到危险显而易见时，重要的决定可能已经做出了。)
人工智能或许有一天会以我们无法控制的方式行事，但许多最严重的危害可能不会发生在“人工智能接管”的戏剧性时刻。人工智能可能帮助设计病原体、发现关键基础设施的漏洞或改进武器系统，而人类仍在决定下一步会发生什么。其他危害可能通过成千上万个普通决策几乎不可见地累积——这里多一点自主权，那里移除一项保障，又因为系统迄今运行良好而移交另一项重要任务。等到危险显而易见时，重要的决定可能已经做出了。
That is also where international agreement could begin. We do not need countries to agree about superintelligence(超级智能) or the probability of human extinction before agreeing that there are some doors AI should never be allowed to open by itself(独自地；自动地). Weapons systems, critical infrastructure and biological synthesis(合成；综合) are obvious places to start. Governments could agree minimum safeguards, require clear human authority for the most consequential actions, keep records of who authorised what and share serious failures and near-misses(险情；侥幸避免的事故) internationally.
这也正是国际协议可以着手的地方。我们不需要各国先就超级智能或人类灭绝的可能性达成一致，然后才同意有些门人工智能绝不应被允许自行打开。武器系统、关键基础设施和生物合成是显而易见的起点。各国政府可以商定最低限度的保障措施，要求最重大的行动必须由人类明确授权，记录谁授权了什么，并在国际上共享严重失败和险情。