This post was originally published on this site.
The astonishing speed at which generative artificial intelligence (AI) programs such as ChatGPT have taken off has shocked even tech insiders, writes Jonathan Margolis.
Only 10 years ago, leading experts were not concerned about the effect on the world of work. “It’s nothing new,” said the then head of AI for Amazon. “It won’t be replacing anyone.” When I emailed him this week, he didn’t answer.
Futuristic fictions such as The Jetsons imagined that all hard physical work would be done by robots, liberating us to do the interesting, creative and brainstretching, stuff. But it’s working out almost the other way round.
As the science fiction and fantasy author Joanna Maciejewska said in 2024: “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing.”
Generative AI has come to be used routinely in graphic design, fashion, copywriting, to create news and features, and social media content, drafting scripts, generating story ideas and editing in the entertainment industry, counselling, architecture, game design, photography and even writing computer code.
The fifth version of ChatGPT is due to launch in a few months; it is said to be twice as clever as the best current version. No wonder almost everyone I know in creative jobs, along with every solicitor and accountant, is worried.
Surely soon it will be people who work with their hands – sculptors, hairdressers, plumbers, carpenters – who are the new, prosperous, stable middle class. And, in this brave new world, what should we be advising young people to do as a career?
Having written about “futurology” for 25 years, I now think there is a lot to be said for aiming to be an articulate, honest and efficient plumber, builder or similar, especially in a wealthy area, where people pay for quality.
While AI can write passably adequate marketing and PR guff, company reports, probably OK soap operas or humdrum local news reports, it’s top-notch tradespeople who can show off the peerless superiority of our brains.
Last summer, in the midst of a big house renovation, I accidentally ordered a £3,000 custom pair of French doors 6cm too wide. When they arrived, I asked Nick, our talented builder, if there was any way to get us out of trouble.
In half a day, with no drawings, no augmented reality, just a circular saw and lots of MDF, he fashioned a solution so good it looks like the door was meant to be the width it was. There’s no way AI could have done that.
As one AI academic, Michael Osborne, a professor of machine learning at the University of Oxford, puts it: “AI is particularly bad at understanding or empathising with physical reality.
“The tangibility of something like a door is not something it can comprehend because it itself doesn’t exist in the real world.”
The futurist and award-winning author Martin Ford has written four books on the subject of AI encroaching on traditional work. He has been telling people for 15 years that plumbing or another skilled trade is a sound career in the new world. But what about going into building AI?
Ford has a warning on this: “It was only a few years ago that the advice was to learn to code. This was the future that guaranteed a job. And yet it turns out that one of the best uses of the new large language models is for computer programming.
“Even some AI companies are saying that within a year or two, they may be able to replace their software developers. And these are not average programmers – they’re the very best.”
Ford believes three categories of job will boom. “Firstly, jobs that require mobility, dexterity and problem-solving skills in a very unpredictable environment. So yes, electrician, plumber, nurse, hairdresser, skilled construction worker, physical therapist, farmer. To automate a job like that would require a science-fiction robot, like C-3PO in the Star Wars movies.”
The second category, he says, will be jobs that require genuine creativity. “If you’re something like a research scientist or an architect and you’re really creating something new, AI is likely to be a tool that amplifies your ability rather than something that literally replaces you.
“And the third area is jobs that involve building sophisticated relationships with people. So nursing again – that has to have empathy for patients – psychologist, business consultant, where you have to get deep into the mind of your client.
“Even an investigative reporter, where you need to go out – not just looking for information online, but talking to people and convincing them to talk. These are uniquely human skills.”
Osborne explained how he and a fellow academic first proposed the same three “protected” job areas in a paper back in 2013; they identified these fields as creativity, social intelligence and manual dexterity. “When we revised last year, we came to the same conclusions.”
His theory is that the people most affected by AI will be those in the mid-ranks of their speciality. Low-skilled people – who, say, aren’t good at writing emails – will be enhanced by AI. High-skilled people will look better because however good AI is, they will still perform better.
“But in the middle, you’ll be put under pressure by this intrusion of new entrants using AI tools.”
If your kids – or you – are still reluctant to train as a plumber, stunt performer, underwater welder, bodyguard or firefighter – all jobs currently recommended as AI-proof – how can you pivot?
The key guideline that emerges is to work with AI’s data processing, not against it, and to do that by developing talents that are still uniquely human.
Anything which requires emotional intelligence, empathy and communication skills will almost certainly continue to be human territory; so leading teams, making informed decisions, interpreting and refining AI’s suggestions, negotiating and persuasion are all expected to remain fruitful.