We are asking the wrong question by asking whether AI will take all the jobs. The real shift is how it is re‑slicing white‑collar work, squeezing some careers and supercharging others.
I was back on BBC Radio Bristol to speak with Claire Carter about the real “AI job story” behind The Guardian’s recent article on white‑collar workers leaving their professions for the trades.
The people in that piece are already feeling the impact: work drying up, rates being cut, and “AI editor” roles where they are paid less to fix machine‑generated output that can actually take longer to do well. For them it is an AI downgrade – more work, less money, and a broken sense of career identity.
Zooming out, the evidence paints a more nuanced picture. Studies suggest around a quarter of jobs sit in roles where a large share of tasks could, in principle, be done by AI‑enabled systems, particularly in routine administrative, call‑centre and some junior professional work.
At the same time, reports from organisations such as the Tony Blair Institute argue that AI is more likely to reconfigure tasks than wipe out whole occupations, with peak unemployment effects in the UK likely measured in hundreds of thousands, not tens of millions, and unwinding over time as new jobs emerge.
Globally, surveys show workers are rightly uneasy – with more than half of people worried that AI could replace jobs – but concern is not the same as destiny.
In the interview I argued that the real divide is not “white‑collar versus blue‑collar”, it is replaceable tasks versus complementary tasks. Roles that are screen‑based, rules‑based and repetitive are under genuine pressure, while work that is physical, caring, relational or deeply contextual remains far harder to automate at scale.
That is why some professionals in the article are swapping into carpentry or plumbing, and why local services, health, education and skilled trades can look oddly resilient while parts of marketing, content and back‑office work are squeezed.
For leaders, the challenge now is to ensure AI becomes an upgrade, not a downgrade. That means redesigning jobs so that AI takes on the dullest 20–30% of the workload, and humans lean into judgement, creativity and relationships.
It means investing in reskilling for those in highly automatable roles, and being honest when business models are shifting rather than simply pushing people into lower‑paid “AI support” work.
Above all, it means moving the conversation away from “Will AI take all the jobs?” towards “How do we design work so people and AI each do what they are best at?”
If you are digitally curious about what this AI “job swap” means for your organisation, your sector, or your own career, the BBC Bristol segment is a great place to reflect on how you and your organisation will get work done in the age of AI.
