The CEO chatbot era is coming | CEO聊天机器人时代即将到来 - FT中文网
登录×
电子邮件/用户名
密码
记住我
请输入邮箱和密码进行绑定操作:
请输入手机号码,通过短信验证(目前仅支持中国大陆地区的手机号)。如果该手机号尚未注册,将自动创建 FT中文网账号。短信可能需要几分钟送达,验证码15分钟内有效,请耐心等待:
请您阅读我们的用户注册协议和隐私权保护政策,点击下方按钮即视为您接受。
FT英语电台

The CEO chatbot era is coming
CEO聊天机器人时代即将到来

Many executives will be following Meta’s plans for an AI version of Mark Zuckerberg with interest
许多高管将密切关注Meta打造马克·扎克伯格人工智能版本的计划。

Chief executives are on a constant quest for the grail of eternal omnipresence. Could artificial intelligence give it to them? Mark Zuckerberg seems to think so.

The Meta chief executive, occasionally lampooned for his robotic stage presence, is helping to train and test an animated AI version of himself to interact with his employees, the Financial Times has reported. The Wall Street Journal also has written that the Facebook founder is building a separate AI agent to help him pull information from different layers of the company. As one FT reader commented: “When is Meta going to build a human version of Mark Zuckerberg?”

Other corporate leaders will be watching the outcome of these trials avidly, though. Many think their ability to realise their grand strategies is limited only by hours in the day. It is why personal productivity is an executive obsession. It is why lots of executives boast that they subsist on tiny sips of sleep, in the mistaken belief that increasing their waking hours will improve their output. In this nonstop effort to manage time more efficiently, the needy employee seeking feedback can be a nuisance.

Taken literally, the chief executive’s promise “my door is always open” was always a hollow one. No boss of a company with more than a few dozen staff can hope to be accessible to everyone at all times. But generative AI can be trained to offer plausible responses, in the voice and manner of its trainer, is a superficially empathetic coach and therapist, and has all the time in the world.

That Zuckerberg should wish to experiment with such technology is understandable — and self-interested. His company employs nearly 79,000 people, only a fraction of whom are ever likely to have meaningful face time with the boss. Meta is pressing all of them to use AI internally to improve their own efficiency. The technology lets Zuckerberg scale up his presence and spread his wisdom, while making better use of his precious schedule and sending a message to his staff and customers that they should use its AI tools to do the same.

But by delegating part of his role to a machine, he raises the question of what CEOs actually do, and how much of it is time well spent. And the obvious danger is that by depending on technology, Zuckerberg and his imitators will give up walking the floor and listening to their people and narrow the understanding of their business. Agentic oversight is the AI version of chief executives’ beloved online “dashboards” — a remote managerial panopticon with a 360-degree view of operations — or their top-down “Dear All” memos.

In just such a post in March 2023, Zuckerberg repeated the mantra that the group was “building the future of human connection”. He also floated the hypothesis “that it is still easier to build trust in person and that those relationships help us work more effectively” as he laid the groundwork for calling staff back into the workplace for at least three days a week.

Yet he is modelling a future that involves AI agents interacting on behalf of their absent human sponsors. That would be a reversal of the legacy of Silicon Valley founding fathers Bill Hewlett and David Packard of Hewlett-Packard. They used to drop in on staff unprompted to ask them what they were working on, a technique they called “management by walking around”. “MBWA assumed that there was no substitute for personal involvement and one-on-one communication to make sure everyone from the bottom to the top of the company had the support to perform at their best,” according to the HP corporate archive.

The Zuckerberg approach risks absolving leaders of the need to convey tough decisions in person. Previous bursts of tech-fuelled workplace efficiency led to cases of remotely ordered firings by fax, then email, then text message. The corporate world can only be a few clicks away from the first AI-mandated mass sacking.

This might fit the image Zuckerberg has cultivated since Donald Trump returned to office. He has reshaped himself as a hard-nosed, old-style boss, appearing on Joe Rogan’s podcast last year to berate corporate America for being “culturally neutered”, while calling for more “masculine energy” in the workplace. It was just one of many signals that the “touchy-feely” leadership of the pandemic was finally over. But the alternative — subordination to the synthesised feedback of an animated bot — could be much, much worse.

版权声明:本文版权归FT中文网所有,未经允许任何单位或个人不得转载,复制或以任何其他方式使用本文全部或部分,侵权必究。

紧张局势升级之际,美国在阿曼湾袭击一艘商用油轮

中央司令部称,该船试图突破美国对伊朗港口实施的海上封锁;这是美军一个月来的首次袭击。

亲欧盟团体主张:若政治立场一致,英国可在一届议会任期内重返欧盟

“英国最佳”表示,伦敦与欧盟保持一致将加快任何重新加入进程;安迪•伯纳姆正准备提出“选项”。

你的AI治疗师存在的问题

数百万人正求助于聊天机器人解决心理健康问题,但仅限于语言交流的科技无法治愈他们。

摩根大通押注欧洲资产管理业务增长

负责该行欧洲资产管理业务的帕特里克•汤姆森表示,他“看好英国”。

德国前情报机构负责人落马

奥古斯特•汉宁被指控十多年来非法购买国家机密。

如果AI更青睐由AI撰写的简历?

种种迹象表明确实如此,而一家律师事务所的经历颇能说明问题。
设置字号×
最小
较小
默认
较大
最大
分享×