Sound government budgeting matters, obviously, but in the long run innovation and productivity growth matters far more to our wellbeing. So ask yourself: Will historians in 100 years’ time really be poring over Rachel Reeves’s quest for “fiscal headroom” or parsing migration stats to see who were net contributors to the deficit in 2025? No. More likely, they’ll look to 2025 as another milestone year in the development of the nascent artificial intelligence technology — a precursor to transformative economic and social changes on par with the industrial revolution.
Will 2025 be the Year of AI Workers?
A big challenge in writing about economics in Britain is the urge to analyse everything through a political lens. Macroeconomic weakness gets ascribed overwhelmingly to the tax and spending policies of governments. If an industry falters, we invariably blame ministers and shout for them to “do something”. Quite often, we talk up the state of the public finances as if they alone define the country’s economic health.
Granted, many Britons take a dim view of such grand tech predictions. Over Christmas, I harangued 30 friends and family members about whether they regularly use ChatGPT or similar AI tools. None did. Many had tried the programs when launched two years ago, found the “hallucinations” offputting, and decided not to bother. Which astonished me, because I use these ever-improving AI tools daily to assist in tasks like editing, building to-do lists, shopping and meal planning, and evaluating data. And even these uses scratch the surface of what AI is capable of.
Look closely and AI is already changing our world. AI video tools cook up eerily lifelike footage, algorithms scan faces in airports, and software sniffs out spam so you never see it. Self-driving cars in the US are already breezing around certain cities. AI hunts down fraudulent charges before you even notice them, flags abnormalities on your X‑rays, and is transforming the process of writing computer code. The AI revolution is no distant prospect; it’s here and accelerating.
Sam Altman, chief executive of OpenAI, kicked off 2025 with a bracing blog post about what we might expect this year. He is convinced his firm already knows how to build out artificial general intelligence: a program that doesn’t just match general human intellect but surpasses it. He predicts the first “AI agents” will “join the workforce” this year, boosting output for companies that can harness them effectively. What’s more, Altman sounds bullish on the likelihood of “superintelligence”, programs that blow past current human intelligence entirely, opening the prospect of solving problems that have thus far evaded us as a species, while propelling scientific discovery.
Britons appear more anxious about this mood music than people in other nations. A Boston Consulting Group survey last year showed that, on balance, 12 per cent more of us were concerned than excited about AI — a similar result to other western European countries and Australia, but the opposite of the upbeat reaction in east Asia. I’ve noticed that people comfort themselves by fretting about abstractions like the existential risks of killer-robots or else dismiss AI as a pointless gadget that will merely encouraging chatting to bots all day.
Yet neither reflex does justice to reality already taking shape. AI is transforming tasks within many jobs, will eliminate plenty of routine-intensive roles (even high-skilled ones), and is generating new products and services (like GitHub Copilot for programmers) that couldn’t have been foreseen a few years ago.
These developments shape my new year’s resolutions. First, I’m determined to worry less over every tedious Westminster tussle on budgets and fiscal rules — sideshows compared to the seismic economic shifts AI and fertility decline could bring. Second, I want to get a grip on these emerging technologies and figure out how to use them more effectively. It seems unlikely that economic scribblers will be spared in the coming revolution.