Will AI turn us all into hipsters and artisans? - 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。
人工智能

Will AI turn us all into hipsters and artisans?

There is good reason to be dubious about the notion that automation will supplant all demand for human labour
00:00

{"text":[[{"start":6.65,"text":"My first chopping board was something cheap I got from Ikea. The most recent one I bought was made by a local carpenter I met at a craft fair. He gave me a card with his contact details, so that I could pop down to his workshop if I had any problems with warping, or needed it re-sanded at some point. "}],[{"start":24.6,"text":"The point of this story — other than to out myself as the dreadful hipster that I now realise I am — is that it sums up one view of how AI and automation might change the world of work."}],[{"start":36.2,"text":"This vision was articulated last month by the economist Alex Imas, who has just been appointed “Director of AGI Economics” at Google DeepMind. In an essay titled “What will be scarce?” he argued that even in a scenario in which advanced AI and automation could one day produce most goods and services more cheaply than humans, the demand for human labour would not necessarily disappear. "}],[{"start":60.2,"text":"Instead, he said, as people got richer, they would want to spend more of their money on “the human-intensive, provenance-rich, sometimes artisanal part of the economy where the human aspect is part of the value of the good or service itself”. In other words, AI might make dreadful hipsters of us all."}],[{"start":79.15,"text":"It is clearly true that wealthy people tend to spend more of their money on things that are inherently labour-intensive — a personal trainer over a 20-person gym class, say, or a painting over a mass-produced poster. This is one good reason to be dubious about the notion that automation will ever supplant all demand for human labour and create a “world without work”."}],[{"start":100.10000000000001,"text":"But there are a couple of caveats to the idea that we could all one day just make paintings and chopping boards for one another. "}],[{"start":106.35000000000001,"text":"The first is distributional. If the productivity gains from automation are not widely shared, but concentrated in a relatively small group at the top, then a large number of people could find themselves competing for the chance to please that small group of elites in labour-intensive ways. Needless to say, this feels like a much less idyllic scenario than one in which we’re all shopping in each other’s craft markets."}],[{"start":131.35000000000002,"text":"The second issue concerns transparency. In the creative sphere, for example, it’s clear that most people place a higher value on human-crafted art, music and writing than AI versions. In one experiment, participants bid for physical art prints that varied randomly in the described level and type of AI involvement. The study found that even “trace amounts of AI involvement were sufficient to substantially devalue the artwork relative to the case where only a human artist was involved”."}],[{"start":160.20000000000002,"text":"The problem is that many people are not very good at telling the difference. Another research study, which showed people samples of AI-written messages, found “strong negative effects on social impressions when disclosing that a message was AI-generated”, but when AI use was not highlighted recipients did not exhibit any scepticism."}],[{"start":179.9,"text":"This presents the risk that true “artisans” will still be undercut or drowned out by machine-made rivals, even though there is genuine consumer demand for what they make."}],[{"start":190.05,"text":"In some spheres, technical solutions will hopefully develop. University of Chicago professor Ben Zhao, for example, has made a new tool which can detect if a song was made with AI. But in writing, AI detectors remain imperfect."}],[{"start":203.10000000000002,"text":"There could be behavioural solutions, too. It seems likely that demand for live music performances will grow, for example. Similarly, I have interviewed illustrators who have begun to make videos for customers of their creative process."}],[{"start":217.50000000000003,"text":"Some writers, meanwhile, are beginning to change the language they use in an attempt to create some distance between their own work and LLMs."}],[{"start":226.85000000000002,"text":"I did this myself recently. I have just finished writing a book, and one of the sentences I wrote some time back — a really nicely crafted sentence that I was proud of — suddenly struck me as the sort of structure which LLMs now spit out with annoying frequency. What once felt pithy and elegant now felt commoditised and cloying. At the last minute, I changed it."}],[{"start":249.50000000000003,"text":"Was it the right decision? I’m still not sure. If we human writers are already changing our work in order to move out of the way of the machines that are trying to imitate us, where will that take us?"}],[{"start":262.05,"text":"There might well be demand for more artisans in an increasingly automated future. But don’t expect it to mean a simple life."}],[{"start":276.3,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1778849006_7397.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。

现代战争的血腥一如往昔

科技的进步并没有减少俄乌战争中的伤亡,武装无人机和AI正把前线变成险恶的杀戮地带,惨烈程度堪比一战。

帕拉贝利斯医药公司于与再生元达成交易次日披露IPO计划,上市热潮升温

成立已有十年且资金雄厚、从格雷格•维尔丁在哈佛实验室孵化出的“不可成药”生物技术公司——帕拉贝利斯医药公司,正寻求成为今年第12家进行首次公开募股的药物研发企业
19小时前

英伟达部署900亿美元助推AI繁荣

黄仁勋正成为依赖其芯片的AI相关公司的最大资助者之一。这些支出涉及逾145家公司,从AI模型开发商、云服务提供商到基础设施供应商不一而足。

Lex专栏:股市投资者信心爆棚,但现金见底

鉴于标普500指数高度依赖以人工智能为驱动的公司,股市出现小问题和大问题的可能性都很大。

FT社评:埃博拉疫情暴露全球应对大流行病准备不足

援助资金减少以及特朗普政府对全球公共卫生理念的敌意,正危及我们所有人。

“四大”急聘AI专业人才,岗位数量盖过传统审计师

全球最大的几家会计师事务所正竞相适应颠覆性的技术变革。
设置字号×
最小
较小
默认
较大
最大
分享×