Fixing the AI industry’s PR problem - 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):
请您阅读我们的用户注册协议隐私权保护政策,点击下方按钮即视为您接受。
FT商学院

Fixing the AI industry’s PR problem

Doomers will say it’s hopeless but here are five ideas to put out the fire
00:00

{"text":[[{"start":5.54,"text":"The public reputation of the big AI companies is an absolute dumpster fire. According to Gallup, 9 per cent of Americans think AI will do more good than harm, versus 39 per cent who take the opposite view. To see how that translates to feelings about OpenAI or Anthropic, try talking to, oh, anyone who doesn’t own a stake in one or the other."}],[{"start":27.18,"text":"What can the companies do to regain trust? What is the right PR strategy for AI? Doomers will say it’s hopeless. There is no way to put lipstick on a pig of a technology that will make a few people rich while draining the livelihoods, privacy, freedom and natural resources of the many. For the utopians, productivity and abundance will soothe all fears in a year or two. I don’t know where I am on that spectrum, but I have been talking with a lot of PR professionals lately, and have been thinking about corporate reputation for years. Here, then, are five ideas for the industry, if it wants to put out the fire."}],[{"start":63.46,"text":"Start with the basics. Be companies, not cults. Enterprises that are worth almost a trillion dollars cannot operate like founder-led start-ups. They need to be institutions where the top job goes to whoever can do it best, under the supervision of a serious board and accountable to investors. OpenAI and Anthropic, in other words, cannot afford to be identified with Sam Altman and Dario Amodei in the way that Meta is identified with Mark Zuckerberg or SpaceX is identified with Elon Musk. The public has to know that the boss is just a person, one who can be sacked if things go wrong. In the meantime, Altman and Amodei need to knock it off with the essays and manifestos. They’re creepy, and the company’s work needs to speak for itself."}],[{"start":107.5,"text":"Hand in hand with this, don’t think you can pull a Travis Kalanick. Uber’s co-founder drove an (app-hailed) tank through the objections of regulators and taxi drivers, betting that happy riders would form a pro-Uber constituency that would carry the day eventually. It worked. But AI is much, much bigger than ride-hailing. Many people’s professional lives are going to be rearranged by force (it’s already happening), and there will be a lot more of them than there ever were taxi drivers."}],[{"start":137.62,"text":"Meanwhile, know that your reputation for both character and competence is under threat. I borrow the distinction from Rupert Younger of the Oxford University Centre for Corporate Reputation. He points out that a poor reputation for character is a relatively quick fix, usually involving getting rid of senior leadership (see above). But when a company gains a reputation for incompetence, it’s a long way back. When an experimental OpenAI model recently escaped its sandbox, accessed the internet and broke into another company’s servers, OpenAI said it took its responsibility to prepare for such incidents “seriously” but described the incident as a technology or industry problem — not an OpenAI problem. The incident was “something we expect to become more commonplace”, and “evidence that, without proper safeguards, highly capable AI agents are now able to work around technical controls”."}],[{"start":189.06,"text":"Next time, instead of generalities, Altman could say this: “We have zero tolerance for our models breaking the law, autonomously or not, and all of our team members — myself included — will be held accountable for seeing that this does not happen again.” If you can’t say something like that, then either (a) you have made a technology you can’t control, in which case you are bad at your job or (b) you like that it is out of control, in which case you are a bad person. Competence and character."}],[{"start":216.22,"text":"Be assured that your foundation/think-tank is worse than useless as a PR strategy. Hiring economists and philosophers to study how AI can be used for good may ultimately expand the horizons of knowledge. But given how low trust has fallen, everyone will suspect your foundation is a sales front or an effort to control the regulatory agenda. What the public needs is not white papers. It is to see you doing good: projects where your workers help families and small businesses prosper. Build something that changes the world now."}],[{"start":250.62,"text":"That is a specific instance of a general point: stop talking so much about the distant future. That is the timescale that Silicon Valley prefers. Back here on Earth, we are trying to get to Christmas. No one cares about a productivity boom that may take a couple of years to arrive. Focus your PR efforts on what is happening today."}],[{"start":270.66,"text":"Will AI companies take this, or any other advice? Not while their valuations are still screaming higher. Money confirms all strategies, even bad ones. But reputation always matters eventually."}],[{"start":285.98,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1788423395_2789.mp3"}

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。

一周展望:日本央行担心通胀超调有没有道理?

《市场前瞻》是英国《金融时报》的未来一周市场情况指南。

科技巨头用担保工具将3000亿美元AI敞口移至表外

华尔街找到新途径,将科技巨头的信用优势转化为更低成本的资金,以支持AI基础设施建设。

无人驾驶出租车冲击重要岗位

克拉克:坐在后座的我们往往看不到出租车司机这份工作的诸多好处。

特朗普称美国已与丹麦达成协议,以取得对格陵兰安全事务的“控制”

丹麦政府表示,协议最早下周即可签署,并将尊重该地区的主权。

特朗普禁止美国主要新闻媒体进入白宫

总统禁止CNN、MS NOW和《政客》参与报道,进一步加大对媒体的打压。

导弹和无人机袭击加剧,沙特拉响空袭警报

也门胡塞武装重新点燃冲突以来,沙特当局首次在首都发布警告
设置字号×
最小
较小
默认
较大
最大
分享×