Letter: ChatGPT outage taught this AI dabbler a valuable lesson - 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

Letter: ChatGPT outage taught this AI dabbler a valuable lesson

From Rowan Hill, Hazel Grove, Cheshire, UK

Like a lot of people, I’ve been dabbling with OpenAI’s ChatGPT with varying seriousness for about 12 months.

“Make me a 12-week marathon training plan”, “Create me a six-week marathon training plan”, “Make a photo of me as the next Pope”, “Is it possible to run a marathon without any training?”

When the platform went kaput this month, I was midway through writing an application for a voluntary role with an academic magazine. I’ve scoffed at accounts of students using artificial intelligence to get through college, deriding them for only cheating themselves. In my own much superior and honest scholarly pursuits, I only ever used it to source articles . . . or plan . . . and sometimes reference . . . or clean up those long sentences that don’t really go anywhere, nor make the point you want to make, but roll on and on for far too long, eating up word count.

As a result of the blackout, however, I ended up with two very different halves of an application. The first, produced with ChatGPT’s help, sounded like something I’d written, if you ignored a few of the more exotic word choices and overlooked the box-ticking style of the answers. It did everything the question had asked of it, and it wasn’t necessarily a fabrication.

The second wasn’t as polished. Pulitzer-worthy accounts of my leadership and prowess in adversity made way for honest accounts of my achievements. Sentences were clipped and to the point. At the end, I realised there were a few ideas that I’d missed but just couldn’t shoehorn them in. But it was definitely me talking about me.

An obsession with performance is continuously being maximised, from schools to the workplace to public services. But at what cost? Tim Harford’s Undercover Economist column “Shared values over profit” (Spectrum, June 7) and his account of Alasdair MacIntyre’s “tale of two fishing crews”, illustrated the perils of this fascination excellently. The first crew is driven by nothing but profit. The second allows time for the cultivation of character and human connection within the team, yet often comes out on top.

ChatGPT’s outage opened my own eyes to how I’d sleepwalked, slowly but surely, into believing maximum utility was somehow better than sticking to my values. This dilemma isn’t just an issue for business; it’s a challenge for individuals too.

Rowan Hill

Hazel Grove, Cheshire, UK

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。

皮特•赫格塞斯:Maga在五角大楼的代言人

特朗普的国防部长赫格塞斯一直站在Maga文化战争的最前线——但他能带领美国与伊朗一战吗?

如何把握大宗商品超级周期

金属、小麦、石油等实物资产可在动荡时期帮助分散投资组合风险——但它们适合散户投资者吗?

从我最大的投资失误看本轮软件股抛售给我们的启示

当你持有的股票下跌时,承认自己判断有误,而不是去寻找证明自己正确的理由。

AI模型就像资本主义,最好与良知一起使用

一些企业认为,用户会更偏爱“预装道德”的产品。

英国按揭贷款机构倒闭震动华尔街

巴克莱、杰富瑞等大型机构合计向Market Financial Solutions提供了20亿英镑融资,正在评估相关的损失。

AI让投资者陷入“幻觉”

围绕下一场重大技术革命的竞争性愿景正把市场搞得晕头转向。
设置字号×
最小
较小
默认
较大
最大
分享×