The 'greybeard' engineers rehired by Ford who could be proof AI can't replace humans

By TOM LEONARD, US CORRESPONDENT Published: 23:07, 7 July 2026 | Updated: 23:25, 7 July 2026 'Artificial intelligence is a fantastic tool, but it's only as good as the information you use to train it,' said senior Ford executive Charles Poon.  'Mistakenly, we thought that by just introducing AI and ingesting the design requirements that we had, that would produce a high-quality product.' The embarrassing volte-face comes just a year after Ford's chief executive, Jim Farley, criticised the quality of America's blue-collar workers and predicted that AI 'will leave a lot of white-collar people behind' and replace 'literally half' of all jobs in the US. And Ford is far from alone in having to rehire the employees they laid off in favour of AI, as executives across the business world discover that humans can't be replaced by machines quite so easily after all. This so-called 'AI boomerang' phenomenon can be seen in a 2025 report from the global market research company Forrester Research, which revealed that 55 per cent of employers regretted their decision to get rid of staff because of AI.  Orgvue, a British software company, found the same percentage of international businesses it surveyed had second thoughts after making redundancies to make way for AI. Separate research found nearly a third of businesses which cut jobs primarily due to productivity gains from AI later rehired them for the exact same role. And this trend is only likely to become more pronounced, it's claimed. Another study recently predicted that half of all companies that replaced customer service or operational employees with AI will be forced to restaff those roles – albeit under different job titles – by next year. Ford is far from alone in having to rehire the employees they laid off in favour of AI, as executives across the business world discover humans can't be replaced by machines quite so easily after all Ford's chief executive, Jim Farley, previously criticised the quality of America's blue-collar workers and predicted AI 'will leave a lot of white-collar people behind' This is rather a different story to the one trotted out by AI industry bosses who insist humanity needs to prepare for a future in which machines can – and will – do so many jobs it will make the Industrial Revolution look like a footnote. 'Companies that moved too quickly on AI are now seeing where it falls short in practice,' Megan Slabinski, an executive at consulting firm Robert Half told business news website Fast Company. While switching to AI brought early 'efficiency gains', it created 'gaps in quality, oversight and decision-making' and was 'not the be-all and end-all solution some initially believed it would be,' she said. Ms Slabinski is hardly the only one suggesting AI has been wildly over-sold. Previously, customer service was widely considered an obvious candidate to be turned over to the machines.  After all, even if a customer chatbot said the wrong thing, how serious could it be compared, say, to a Ford Puma starting to fall apart on the M25? In fact there's been a string of expensive foul-ups. In March, tech giant Meta announced that users of its social media platforms Facebook and Instagram who found themselves locked out of their accounts would no longer have access to a customer service representative but instead an AI-powered chatbot. The decision was an open invitation for internet scammers to try their luck at tricking these computer replacements.  They were spectacularly successful: in May, hackers managed to take control of 34,000 Instagram accounts by exploiting a bug in Meta's AI customer service software that allowed them to change someone's password by simply asking the chatbot to do it. Among those affected was the former White House social media account for Barack Obama which started posting bizarre Instagram messages mocking President Trump and claiming the White House was 'under Shiite control'. Experts say scammers exploit how AI chatbots are encouraged to ingratiate themselves with users. A US car dealer chatbot was even manipulated into selling a Chevrolet SUV for $1. The AI 'revolution' faces other setbacks. Tech giants such as Meta, Amazon and Uber have drastically reined in how much they allow employees to use it because it cost so much money.  In May, for example, it was revealed that an unnamed company had 'accidentally' incurred a $500million bill with AI company Anthropic in a single month. But is this U-turn just a temporary setback – are those Ford 'greybeards' merely training the company's AI to be better at their jobs so they can be replaced again? Not necessarily. Apart from the cost issue, critics flag up a more fundamental fact about AI that makes it inherently unsuitable for replacing human intelligence in most jobs: that the two things are just intrinsically different. Existing 'generative' AI models create 'content' such as text, images and video based on patterns they've learned from crunching vast amounts of data that's been fed to them. These chatbots may be able to have impressively cerebral online conversations with people – and flatter or even flirt with them – but it's deceptive. Generative AI has no power of reasoning and can only predict what is the most probable human answer. And as it doesn't truly know facts, it continually makes errors, known as 'hallucinations'. AI is fine for doing work in which the output is either right or wrong and can be checked through some sort of automated process, but most jobs aren't like that and so require what is known as 'human-in-the-loop' review, or a real person.  (Computer coding is one of those types of jobs, which is why few expect Silicon Valley to restore the many staff laid off there).  Meta boss Mark Zuckerberg revealed he didn't believe AI would lead to mass job losses Last week, even Meta boss Mark Zuckerberg revealed he didn't believe AI would lead to mass job losses. AI optimists counter that the technology is getting better at an astounding pace and will inevitably overcome its issues. But sceptics insist the basic problem remains: artificial intelligence simply cannot be trusted to be let loose on most jobs without a grown-up present. Many people who would rather not see their jobs thrown on the scrap heap of technological progress will no doubt be praying they are right.