
控制海量个人数据的科技公司可能利用这些数据影响人们的思维,Palantir 联合创始人兼首席执行官亚历克斯·卡普(Alex Karp)周二发布的访谈中发出警告。
Technology companies that control vast amounts of personal data could use it to influence people's thinking, Palantir co-founder and CEO Alex Karp warned in an interview released Tuesday.
卡普在《建设者》(The Builders)播客节目中对记者沃尔特·艾萨克森(Walter Isaacson)表示,美国人担心政府掌握他们的哪些信息,却较少关注科技公司如何使用他们自愿分享的信息。
Americans worry about what the government knows about them but pay less attention to how technology companies use the information they willingly share, Karp told journalist Walter Isaacson on The Builders podcast.
卡普说,人们倾向于将在网上分享个人信息视为一种选择,但可能忽略了在数字平台上收到的推荐内容如何塑造他们的观点。
People tend to view sharing personal information online as a choice, Karp said, but may overlook how the recommendations they receive on digital platforms can shape their opinions.
“如果由公司控制所有数据,它们将控制你的想法,或严重影响你的想法,”卡普在8月录制的访谈中说。
"If you have companies controlling all data, they will control what you think or heavily influence what you think," Karp said in the interview, recorded in August.
他主张,即使政府没有直接介入,商业数据收集也引发了严重的隐私担忧。
He argued that commercial data collection raises serious privacy concerns even when governments are not directly involved.
由卡普联合创立的 Palantir 开发由 AI 驱动的数据分析软件,供军事、情报和执法机构使用。
Palantir, which Karp co-founded, develops AI-powered data analysis software used by military, intelligence and law enforcement agencies.
被问及监控问题时,卡普为在法律授权下使用该公司工具进行了辩护,但警告称监控属于 AI 最危险的潜在应用之一。
Asked about surveillance, Karp defended the use of the company's tools when authorized by a legal mandate but warned that surveillance was among AI's most dangerous potential applications.
他说,此类系统既可能由公司部署,也可能由政府部署,并呼吁对某些 AI 技术实施限制,包括立法防止为军事用途开发的系统被部署在国内。
He said such systems could be deployed by both companies and governments and called for restrictions on certain AI technologies, including legislation preventing systems developed for military use from being deployed domestically.
卡普还指责大语言模型开发者利用客户的专有数据改进其 AI 系统,却未给提供这些信息的组织足够的控制权。
Karp also accused large language model developers of using customers' proprietary data to improve their AI systems without giving the organizations supplying that information sufficient control.
他辩称,企业实际上可能是在为 AI 服务付费,同时贡献宝贵数据,而这些数据最终可能被用来开发竞争产品。
Businesses could effectively pay for an AI service while contributing valuable data that might eventually be used to develop competing products, he argued.
卡普表示,公司和国家应对其数据、AI 模型和敏感技术系统保留更大的控制权。
Karp said companies and countries should retain greater control over their data, AI models and sensitive technological systems.
卡普表示,人工智能可能创造更多薪酬更高的工作岗位,尤其是为蓝领和职业技术工人,但也警告称,这可能扩大贫富差距。
AI's economic impact Karp said AI could create more and better-paying jobs, particularly for people in working-class and vocational occupations, but warned that it could also widen the wealth gap.
他说,拥有或开发人工智能系统的人,积累财富的速度可能远快于其他工人。
Those who own or develop AI systems could accumulate wealth far more quickly than other workers, he said.
他呼吁政府解决这种差距,同时确保税收和再分配政策改善低收入人群的生活。
He called on governments to address the disparity while ensuring that taxation and redistribution policies improved the lives of lower-income people.
不过,他反对征收财富税,理由是执行财富税可能需要对个人财产进行侵入式监控。
However, he opposed a wealth tax, arguing that enforcing it could require intrusive monitoring of personal property.