As the United States approaches an election year in 2024, citizens brace for the onslaught of robocalls from political campaigns asking for their vote. What people might not expect, however, is to hear a politician who is not known to be multilingual speaking in a language other than English. It turns out, this type of surprise phone call is the result of AI voice cloning and translation technology, and it’s becoming more common.
Last month, New York City Mayor Eric Adams made headlines when he used AI, including a newly released tool from ElevenLabs, to make calls to constituents in Spanish, Yiddish, Mandarin, Cantonese, and Haitian Creole — despite not speaking any of those languages.
Adams views the AI-modified calls as a way to more effectively reach diverse communities, telling City Hall reporters, “I have to be able to speak to people in the languages that they understand, and I’m happy to do so.” A spokesperson for the mayor’s office said that native speakers listened to the recordings before they went out to ensure the translations were accurate.
But some tech ethics groups feel that the calls were misleading because they didn’t disclose the use of AI. Executive Director of the Surveillance Technology Oversight Project Albert Fox Cahn told the Associated Press, “The mayor is making deep fakes of himself. This is deeply unethical, especially on the taxpayer’s dime. Using AI to convince New Yorkers that he speaks languages that he doesn’t is outright Orwellian. Yes, we need announcements in all of New Yorkers’ native languages, but the deep fakes are just a creepy vanity project.”
Opinion journalists are also weighing into the debate. NPR’s Scott Simon said on Weekend Edition, “Mayor Adams’ voice making robocalls in fluent Mandarin may seem more absurd than harmful. But imagine the real damage that could be done if various operatives begin to use deep fake technology to make politicians and public figures seem to say, in voices well-known and familiar to us, things that they never really said in any language?”
Recorded phone calls aren’t the only potential application of AI voice translation in politics; the technology could easily be used in video campaign ads and other types of messages. And with plenty of companies offering similar tools, including Spotify and Google, there’s a good chance Americans will be hearing more politicians say, “Yo apruebo este mensaje.”
随着美国即将迎来2024年的选举年,公民们准备迎接来自政治竞选活动的机器人电话的冲击,要求他们投票。然而,人们可能不会想到的是,听到一位不懂多种语言的政治家用英语以外的语言说话。事实证明,这种类型的惊喜电话是人工智能语音克隆和翻译技术的结果,并且越来越普遍。
上个月,纽约市市长埃里克·亚当斯(Eric Adams)使用人工智能(包括ElevenLabs新发布的工具)向西班牙语、意第绪语、普通话、广东话和海地克里奥尔语的选民打电话,尽管他不会说这些语言。
亚当斯认为,人工智能修改的电话是一种更有效地接触不同社区的方式,他告诉市政厅的记者,“我必须能够用他们理解的语言与人们交谈,我很高兴这样做。”市长办公室的一位发言人说,母语为英语的人在出门前会听录音,以确保翻译准确无误。
但一些科技伦理组织认为,这些电话具有误导性,因为它们没有披露人工智能的使用情况。监控技术监督项目执行主任阿尔伯特·福克斯·卡恩告诉美联社,“市长正在深深地欺骗自己。这是非常不道德的,特别是在纳税人的钱。使用人工智能来说服纽约人,他说的语言,他不完全是奥威尔式的。是的,我们需要用所有纽约人的母语发布公告,但这些深度伪造只是一个令人毛骨悚然的虚荣项目。”
舆论记者也在参与这场辩论。NPR新闻的斯科特·西蒙在周末版节目中说:“亚当斯市长用流利的普通话打电话似乎更荒谬,而不是有害。但是,想象一下,如果各种各样的特工开始使用深度伪造技术,让政客和公众人物似乎用我们熟悉和熟悉的声音说一些他们从未用任何语言说过的话,可能会造成什么样的真正损害?”
录音电话并不是人工智能语音翻译在政治领域的唯一潜在应用;该技术可以很容易地用于视频竞选广告和其他类型的信息。随着包括Spotify和谷歌在内的许多公司提供类似的工具,美国人很有可能会听到更多的政治家说,“Yo apruebo este mensaje”。
以上中文文本为机器翻译,存在不同程度偏差和错误,请理解并参考英文原文阅读。
阅读原文