{"text":[[{"start":7.55,"text":"Companies investing most heavily in AI are adding workers faster than their peers, according to new research that challenges predictions of broad AI-driven job losses."}],[{"start":17.4,"text":"White-collar worker numbers increased 10.2 per cent overall at companies that used generative AI most intensely in the first two years after they first adopted the technology, the research found, with gains across occupation types and seniority levels. Entry-level employment increased by 12 per cent."}],[{"start":36.45,"text":"However, among organisations that adopted AI but to a lesser degree — defined as the bottom two-thirds of spending per worker — there was no significant change in worker numbers compared with a control group."}],[{"start":null,"text":"
"}],[{"start":49.300000000000004,"text":"The findings run counter to claims that AI adoption will spur widespread job losses, even as tech groups including Oracle and Atlassian have cited AI investment when announcing lay-offs. Instead, they suggest the technology may be associated with faster hiring — but only for companies investing heavily enough to realise productivity gains."}],[{"start":71.05000000000001,"text":"The study, co-authored by researchers at US tech start-ups Ramp and Revelio Labs, covered almost 22,000 US companies and is the first to combine organisation-level headcount and AI spending data."}],[{"start":84.4,"text":"Ara Kharazian, chief economist at Ramp and co-author of the paper, said the data shows companies that use AI are growing faster but the benefits are unevenly distributed."}],[{"start":95.45,"text":"“There’s clearly some kind of learning curve — the [headcount] gains don’t show up for at least six to 12 months — and it’s subject to a minimum threshold. You only get these gains if you are a high-intensity adopter, and that typically comes with a decent amount of investment, beyond a couple of dollars a month on ChatGPT.”"}],[{"start":114.15,"text":"The research linked data from Ramp, a payment processor, showing how much companies paid to AI vendors, with workforce records compiled by Revelio Labs from online public profiles such as LinkedIn."}],[{"start":126.55000000000001,"text":"Because AI adopters tended to be more technical, higher-paying and more likely to have received venture capital backing than non-adopters, the study compared early AI adopters with companies that adopted the technology later."}],[{"start":138.8,"text":"One labour economist told the FT that the results, while interesting, should be interpreted with caution, particularly as the groups using AI most in the sample tended to be smaller. “‘Intense AI adopters grow faster’ and ‘small fast-growing start-ups buy a lot of AI quite early’ seems hard to separate here,” he said."}],[{"start":159.15,"text":"Kharazian at Ramp warned almost all headcount gains were among companies in the tech sector and the study covered only white-collar workers."}],[{"start":167.8,"text":"Academic research has painted a mixed picture of AI’s effect on the labour market. Stanford research published in November found a 16 per cent reduction in early-career employment in jobs exposed to AI."}],[{"start":179.9,"text":"However, a paper by Harvard economists last year, covering 280,000 companies, found declines in junior employment among AI adopters, while senior roles were largely unaffected."}],[{"start":191.4,"text":"Oracle said last week it had cut 21,000 jobs over the past year and warned its investment in, and use of, AI could lead to further reductions. Snap, Block and Cisco are among other tech companies to recently link thousands of job cuts to AI."}],[{"start":212.9,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1782806396_7529.mp3"}