During Ocado’s most recent earnings call, chief executive Tim Steiner said the group’s advances in artificial intelligence and robotics had allowed it to fulfil online grocery shops at an ever faster pace.
In 2012, it took 25 minutes of human labour to pick a 50-item order. That is now down to 10. But Ocado’s technological progress means the company requires 500 fewer workers this year, after it already announced 2,300 jobs would be at risk in 2023.
The UK company’s move over many years to phase down human labour where feasible exemplifies workers’ fears about generative AI: it may boost productivity, efficiency and profitability but it can also displace staff.
Some businesses are yet to embrace the shift but many have spent more than a year experimenting and engaging in workplace pilots.
“Companies are moving from asking, ‘What is our AI strategy?’ to experimenting . . . implementing generative AI into processes,” said Karin Kimbrough, chief economist at LinkedIn. “It is starting to change the landscape of work.”
Now employees, bosses and policymakers are trying to decipher what exactly the benefits of generative AI look like.
“This latest generation of AI could change every job. I don’t think that is too much of an exaggeration,” said Peter Cheese, chief executive of the Chartered Institute of Personnel and Development, the UK’s professional body for HR and people development. “Of course you can see examples where AI in different forms is already making a difference to their workforce, but it’s still early days for many companies.”
Many employers are cutting jobs under the guise of economic and political uncertainty. But high profile examples of AI-driven lay-offs in recent months, from technology company IBM to language learning app Duolingo, are fuelling questions about whether a slash and burn of white-collar roles is under way.
The 42-year-old billionaire Dario Amodei, who runs AI developer Anthropic, has warned the technology he and peers such as OpenAI are building could wipe out half of all entry-level office jobs in the next five years. Already, graduates account for just 7 per cent of hires across the 15 biggest technology companies, with the number of new recruits down a quarter compared with 2023, according to SignalFire, a venture capital firm.
“AI is starting to get better than humans at almost all intellectual tasks, and we’re going to collectively, as a society, grapple with it,” Amodei told television network CNN last month. “AI is going to get better at what everyone does, including what I do, including what other CEOs do.”
Academics, recruiters and management consultants are split on whether talk of a bloodbath is just scaremongering or a clear-eyed view of AI’s potential to shake up the labour market. But even if AI is not destroying jobs at scale today, it is certainly redesigning them and changing the equation between work, output and headcount.
“No sector is immune [to the impact of AI],” said Peter Brown, a global workforce expert at PwC. “But it is primarily changing roles not eliminating them, enabling humans to focus on more value-add elements of their jobs.”
For now, Mike Clancy, general secretary of the Prospect trade union, which represents 160,000 members across the public and private sectors in the UK, largely agreed. He said it was important to differentiate between industries.
“If you work in air traffic control, on transmission and distribution networks, run infrastructure or manufacturing processes, you have been shifting your talents to align with new technology for years and years. AI can help, but the human aspect will have to be retained for systems resilience.”
By contrast, Clancy said “email jobs” — those reliant on text responses, from solicitors to customer service agents — would undergo “spectacular shorter-term change” even if there was a “long lead time to realising the benefits”.
Since the launch of ChatGPT in late 2022, workplace experts have been trying to ascertain whether companies will opt to amplify their capacity by empowering staff to do more work alongside AI or seek to keep the same output while reducing hiring.
Schroders is one example of the former. “We don’t see a short-term revolution and mass displacement of roles but we do anticipate these trends will drive an evolution over the next five to 10 years that will reshape workforces,” said Meagen Burnett, chief financial officer at the fund manager. Using AI to run analytics of the company’s data, create reports and provide support with queries are examples of uses that could be “transformative”. Having a more “AI literate” workforce would make staff more productive, give them an edge and help them to deliver higher-value work, Burnett added.
Biotechnology company Moderna’s recent move to combine its human resources and technology functions — opening the door to more automation — is another sign of how companies are beginning to think about workplace planning.
IBM has gone further, using AI agents to take on the work of hundreds of HR staff. Klarna, the Swedish fintech, said its AI assistant now managed two-thirds of customer service queries, drastically cutting volumes handled by humans (the chief executive later admitted such stark cost cuts lowered quality). At Google and Meta, AI is reshaping engineering, recruiting and marketing, contributing to reorganisations and headcount cuts.
Other companies are “at the start of the journey”, said James Milligan, the global head of technology at recruiter Hays. “Lots of traditional sectors outside of tech — large private sector companies, FTSE 250, Fortune 500 — are still grappling with some of the basics of good governance and data protection policies in a world of generative AI.”
AI is already shifting the skills workers need. Those fluent in the new tools or who have experience deploying them in workplaces are being promoted, paid more and recruited aggressively. LinkedIn data shows a rise in hiring for roles including prompt engineer, head of AI and responsible AI use architect.
A new PwC report, which analysed almost 1bn job advertisements across six continents, found AI-skilled workers were being paid 56 per cent more in 2024 than those without knowledge of the technology, compared with 25 per cent more the previous year.
It also found industries deemed “most exposed” to AI experienced three times higher growth in revenue per employee than those considered “least exposed”. For optimists, this is proof AI is making individual workers more valuable. “Contrary to fears about job losses, job numbers — and wages — are growing in virtually every AI-exposed occupation, including the most highly automatable jobs,” the PwC report said.
Yet there is a big question over how long this will last. PwC said AI-exposed jobs were growing more slowly (at a rate of 38 per cent) than less-exposed jobs (at 65 per cent). The number of roles such as financial analyst, legal associate or market researcher was increasing at a more sluggish pace than before.
There is also a risk that workers get left behind. According to PwC, the mix of capabilities sought by employers is changing 66 per cent faster in occupations most exposed to AI, such as financial analysts, than in those least exposed, such as physical therapists. This acceleration makes it harder to keep up with demands for new skills, especially for mid-career workers who may not be AI natives, and those not working for large companies. In all analysed countries, women held a higher share of AI-exposed jobs than men.
Workers who saw their responsibilities absorbed into new technologies were “rotating towards skills generative AI can’t do”, said Kimbrough at LinkedIn. These individuals were “disrupted” rather than “displaced”, she added, meaning they were prioritising human-centric skills.
Claudia Harris, chief executive of Makers, a tech talent and training platform, said “a two-speed economy is emerging” between the companies investing in AI and those that are not.
“The dividing lines are not traditional. This is not about innovative industries versus not. This is about companies and cultures that are able to make this huge and decisive shift.”
