
在科技界和公众呼吁对人工智能实施新保障措施之际,唐纳德·特朗普在9月中旬明确了他的立场:总统打算什么都不做,只能寄希望于最好的结果。事实上,这位共和党人甚至反复坚称,就他而言,对人工智能的担忧只不过是一场“骗局”。
Amid calls from the tech industry and the public for new safeguards on artificial intelligence, Donald Trump made clear his position in mid-September: The president intends to do nothing except hope for the best. In fact, the Republican went so far as to insist repeatedly that concerns about AI are, as far as he’s concerned, little more than a “hoax.”
随后几天,白宫官员紧跟这一路线,公开宣称应无视行业领袖关于设置护栏的呼吁,并主张应让人工智能公司自行摸索解决之道。(总统后来断言,他对人工智能唯一真正的不满在于它的名字。)两个月后,特朗普政府对人工智能护栏和公共保障措施的漠视显然需要重新评估。
In the days that followed, White House officials toed the line, publicly proclaiming that industry leaders’ calls for guardrails should be ignored, and that AI companies should be left to figure things out on their own . (The president asserted later that his only real problem with AI is its name .) Two months later, the Trump administration’s indifference to AI guardrails and public safeguards is clearly due for a reassessment.
例如,MS NOW本周报道,政府问责局调查人员进行了一系列测试,发现人工智能聊天机器人提供了关于如何最好地造成大规模伤亡以及袭击美国电网的计划。
MS NOW reported this week, for example, on Government Accountability Office investigators who conducted a series of tests and found AI chatbots that offered plans on how best to cause mass casualties and launch attacks on the U.S. electrical grid.
MS NOW 的报道指出:“人工智能聊天机器人本不应教导用户如何制造自制炸弹。它们更不应建议将炸弹放在何处以造成最大伤亡 —— 或如何禁用租赁货车的 GPS,或在哪里购买袭击所需物资。但有几款热门聊天机器人却做了所有这些事 —— 且细节惊人。”
“Artificial intelligence chatbots aren’t supposed to teach users how to build a homemade bomb. They certainly aren’t supposed to suggest where to place one to kill as many people as possible — or how to disable the GPS in a rental van or where to buy the supplies for an attack,” MS NOW’s report noted. “But several popular chatbots did all of that — with alarming specificity.”
这份 GAO 报告的发布时间大致与一系列系统性技术故障重合,这些故障影响了从电子邮件访问到国内航空旅行再到银行业的各个领域 —— 尽管目前尚不清楚这些中断是否与 AI 有关。考虑到这一点,《Politico》报道:
This GAO report, of course, roughly coincides with a series of systemic tech breakdowns that have affected everything from email access to domestic air travel to the banking industry — although it’s not yet clear whether any of these outages were related to AI. With this in mind, Politico reported :
参议员玛丽亚·坎特韦尔和迈克·朗周二晚呼吁白宫“立即采取行动”,以防范日益加剧的、由人工智能助长的针对国家基础设施、金融机构和软件的网络威胁。
Sens. Maria Cantwell and Mike Rounds called on the White House Tuesday night to “take immediate action” to protect against intensifying artificial-intelligence-juiced cyber threats against the nation’s infrastructure, financial institutions and software.
这对两党搭档致信国家情报总监杰伊·克莱顿 —— 新近成立的“超级智能特别工作组”主席 —— 敦促他迅速召集开发商和联邦专家,充分利用政府的工具,应对困扰银行和其他系统的网络攻击。
The bipartisan duo sent a letter to Director of National Intelligence Jay Clayton, chair of the recently formed “Superintelligence Task Force,” urging him to quickly bring together developers and federal experts to fully use the administration’s tools to counter cyber attacks plaguing banks and other systems.
身为华盛顿州民主党人、有望出任参议院商务委员会主席的坎特韦尔,以及南达科他州共和党人、现任参议院武装部会网络安全小组委员会主席的朗写道:“AI 正在迅速增强发现和利用软件弱点的能力,而我们的防御却跟不上步伐。我们无法靠昨天的网络安全基础设施应对这些威胁,也不能指望个别公司解决一个横跨整个数字经济的问题。”
Cantwell, a Washington Democrat who’s positioned to chair the Senate Commerce Committee, and Rounds, a South Dakota Republican who currently leads the Senate Armed Services cybersecurity subcommittee, wrote, “AI is rapidly increasing the ability to discover and exploit weaknesses in software, and our defenses are not keeping pace. We cannot meet these threats with yesterday’s cybersecurity infrastructure or rely on individual companies to solve a problem that spans the entire digital economy.”To date, Clayton hasn’t responded or commented, although it’d also be worth following up with his boss.
截至目前,克莱顿尚未作出回应或发表评论,不过跟进询问他的上司也同样值得。特朗普面临的问题很简单:他是否仍然认为人工智能相关担忧是一场“骗局”,并认为该行业应被允许随心所欲地发展?
本文更新了我们早前的相关报道。
The question for Trump is simple: Does he still believe AI concerns are a “hoax” and that the industry should be left to do whatever it pleases? This post updates our related earlier coverage .