Only 4 percent of jobs rely heavily on AI, with peak use in mid-wage roles
- Reference: 1739259069
- News link: https://www.theregister.co.uk/2025/02/11/ai_impact_hits_midtohigh_wage_jobs/
- Source link:
The research found that roughly 36 percent of occupations incorporate AI for at least 25 percent of their tasks. These findings align with previous reports showing that [1]few businesses have fully embraced the technology.
Of those seeking AI assistance, about 37 percent work in software engineering roles, 10 percent toil in fields related to media, the arts, and design, and nine percent are involved in education and library services.
[2]
Occupational roles in which AI is least useful tend to involve physical labor, such as transportation and material moving, healthcare support, and farming, fishing, and forestry occupations.
[3]
[4]
The study found 57 percent of AI use goes towards augmenting human work, and 43 percent automates work.
To understand how AI is impacting the economy, Anthropic, one of the top contenders in the commercial AI space behind OpenAI, has launched the [5]Anthropic Economic Index .
[6]
Backed by [7]a research paper [PDF] on the subject, the initiative looks at the impact AI is having on specific occupations, as defined by the US Bureau of Labor Statistics – which had not been disbanded by DOGE the last time we checked.
AI, the outfit says, is affecting people's jobs, an issue that has prompted numerous studies in recent years and raised thorny socio-political questions about labor costs, wages, and the consequences of automation.
AI boosts women's employment in educated economies
In an economic research paper titled " [8]AI and Women's Employment in Europe ," authors Stefania Albanesi, António Dias da Silva, Juan F. Jimeno, Ana Lamo, and Alena Wabitsch find that AI can boost women's employment in countries where women have high labor force participation and strong educational attainment.
"Our findings are consistent with the notion that the diffusion of AI-enabled technologies may benefit women’s employment and that this benefit may be amplified by improvements in educational attainment," the authors report.
Women are more likely than men to work in occupations with higher exposure to AI, the authors note, citing prior research. However, higher levels of education help mitigate potential negative impacts in these roles.
As France convenes its [9]Artificial Intelligence Action Summit to discuss these issues, Anthropic is offering data on actual AI usage based on the prompts it receives from users of its software, with appropriate privacy protections, we're assured.
To date, AI's impact is not particularly broad but it has transformed certain jobs.
"Only ∼4 percent of occupations exhibit AI usage for at least 75 percent of their tasks, suggesting the potential for deep task-level use in some roles," the paper states, citing foreign language and literature teachers as an example. "More broadly, ∼36 percent of occupations show usage in at least 25 percent of their tasks, indicating that AI has already begun to diffuse into task portfolios across a substantial portion of the workforce."
[10]
While computer and mathematical occupations queried Claude the most, that job category accounts for only about 3.4 percent of US workers. The category office and administrative support was responsible for just 7.9 percent of prompts to Claude but represents the highest percentage of US workers, 12.2 percent, within the total workforce.
[11]Microsoft 365 price rises are coming – pay up or opt out (if you can find the button)
[12]Meta's plan to erase 5% of workforce starts today
[13]Cloudflare hopes to rebuild the Web for the AI age - with itself in the middle
[14]France, UAE to drop €50B on AI mega-datacenter. Still nowhere near America's $500B bet
AI usage peaks in mid-to-high-wage occupations, notably [15]IT-related jobs . But it drops off at both extremes of the salary spectrum - such as highly paid roles like physicians and lower-wage jobs like restaurant workers.
"Our empirical findings both validate and challenge previous predictions about AI’s impact on work," Anthropic's paper states. " [16]Webb [2019, PDF] predicted highest AI exposure in occupations around the 90th wage percentile, while we find peak usage in mid-to-high wage occupations, with notably lower usage at both extremes of the wage distribution. This pattern suggests that factors beyond technical feasibility – such as implementation costs, regulatory barriers, and organizational readiness – may be tempering adoption in the highest-wage sectors."
The Anthropic researchers also point to predictions from a [17]2023 research paper suggesting 80 percent of US workers could have at least 10 percent of their work tasks affected by the introduction of LLMs. The AI giant says their data indicates only about 57 percent of occupations are using AI for at least 10 percent of their tasks, though they suggest that number could rise as AI adoption expands. ®
Get our [18]Tech Resources
[1] https://www.theregister.com/2025/02/06/lenovo_ai_report/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Z6st0v9jyF4FcyWCI7U2OQAAAEo&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Z6st0v9jyF4FcyWCI7U2OQAAAEo&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Z6st0v9jyF4FcyWCI7U2OQAAAEo&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.anthropic.com/economic-index
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Z6st0v9jyF4FcyWCI7U2OQAAAEo&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[7] https://assets.anthropic.com/m/2e23255f1e84ca97/original/Economic_Tasks_AI_Paper.pdf
[8] https://www.nber.org/papers/w33451
[9] https://www.elysee.fr/en/sommet-pour-l-action-sur-l-ia
[10] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Z6st0v9jyF4FcyWCI7U2OQAAAEo&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[11] https://www.theregister.com/2025/02/07/microsoft_365_price_rises/
[12] https://www.theregister.com/2025/02/10/meta_to_toss_5_of/
[13] https://www.theregister.com/2025/02/10/cloudflare_q4_2024_ai_web/
[14] https://www.theregister.com/2025/02/08/uae_france_dc_ai/
[15] https://www.theregister.com/2025/02/10/january_it_hiring/
[16] https://www.michaelwebb.co/webb_ai.pdf
[17] https://arxiv.org/abs/2303.10130
[18] https://whitepapers.theregister.com/
What a surprise
Pseudo-AI is useless in customer-facing jobs, and even more in highly specialised/critical jobs like surgury.
So pseudo-AI can basically only be used in administrative tasks and, even then, it's not used all that much.
Well gosh, I do believe I've read that some lawyers found out that AI won't write their judicial papers all that well.
I'm waiting for the report that says that companies have banned the use of AI for internal reports.
Ah, who am I kidding ? That will never happen.
Well, Zoom and similar conferencing tools have automatic minuting now. I troubled to read through some and while it's mostly correct, the most common mistake was for the transcript summary to attribute 'next step' actions to the wrong person. This will undoubtedly be improved on in the future - so long as anyone bothers to complain to the tool makers about it.
Speech recognition systems are "AI" now, are they?
Isn't everything?
The physical job replacement by technology has already happened / continues to happen. It has been many years since a team of men with shovels has been replaced by one man in a JCB. The analogue for physical jobs is mechanisation rather than AI. To take it a step further and replace the man driving the JCB might be 90% possible with perhaps simple feedback automation not necessarily AI, but there will always be that last bit that is the hard bit.
Like automatic landings for aircraft, been happening since the 50s, or for civil aircraft the 1960s. Still 60 years on we have 2 human brains watching its every move.
So AI finds a place in admin, but there is still that last 10% that is hard to fix - bosses may consider that the 10% error is worth it for a large reduction in staff costs. Customers are likely to disagree because that's a lot of customers for a big business and a lot of unwanted unhappy noise flying around the internet.
4% is far far higher than I would have imagined.
so far the only people using AI that I've seen in industry have been a certain subset of job candidates trying to cheat their way through the take-home technical tasks. Easily caught because the tasks I've given are designed to fool AI and be done properly by a human and the cheaters never even check to see if their code executes properly.