Bill Gates warns about AI’s effect on employment. The challenge reaches beyond job losses to the training of tomorrow’s professionals.
☕ TECHLATTE: IDEAS WITH FOAM
⏱ Estimated reading time: 2.5 minutes
For years, we imagined AI danger as something cinematic: machines out of control, autonomous systems, cyberattacks and robots making decisions. Perhaps the more troubling scenario is less spectacular.
A company discovers that it can operate better, faster and more cheaply with fewer people. Then another company does the same. And then thousands. This is part of the question Bill Gates has put on the table.
The problem might not be losing a job
It might be finding that the job no longer needs to exist as it once did. In his essay The turbulent AI era is here. The choices we make now are critical, Gates warns that the transition to an AI-driven economy could be especially turbulent because this time automation is reaching cognitive work: analysis, customer service, programming, finance, law, administration and sales. Junior and mid-level workers may feel the effects first.
If a company automates tasks previously done by a young professional, more than a vacancy disappears. A step on the learning ladder disappears. Where will future senior professionals come from if fewer junior roles remain?
‘Human Reserved’
Gates suggests discussing activities that should be reserved for humans—not because AI cannot perform them, but because it may become very good at them. He mentions care for people and the communication of a serious diagnosis. The same debate could extend to hiring, credit assessment, legal decisions and medical treatment. The question becomes: do we want AI to do this without a responsible person behind the decision?
The economic incentive
People require wages, benefits, taxes, holidays, training and time. Software does not operate under the same rules. If automation becomes much cheaper than hiring, businesses will have a powerful incentive to replace human work. Gates therefore raises the possibility of taxing some uses of AI and robots and financing retraining and social protection. The concern is not that technology is the enemy; it is that economic incentives may accelerate replacement faster than society can adapt.
Is a jobs apocalypse inevitable?
We cannot say that today. The ILO estimates that roughly one in four workers worldwide is in an occupation with some exposure to generative AI, while considering transformation of many jobs more likely than their complete disappearance at present. The World Economic Forum projects both job destruction and creation this decade. But a displaced worker needs to know which jobs will appear, where, when and whether they will have the skills to take them.
Colombia has its 2025 CONPES 4144 policy on AI. That is a step forward, but harder questions remain. Which decisions should retain human involvement? Who is responsible when AI harms someone? What happens to workers who lose jobs before they can retrain? How will professionals be formed if entry-level positions disappear?
The real risk
The most troubling scenario may not be a machine rebelling. It may be a quiet sequence: one company eliminates ten positions, another a hundred, and young people discover that the roles in which they once began no longer exist. An economy could emerge in which being able to work no longer guarantees that anyone needs to hire you.
We should ask not only what work AI can do, but which work and decisions we are willing to protect before it is too late.
☕ This coffee is served. If we had to designate three activities as ‘Human Reserved’ tomorrow, which would you protect?