Fiercely
Independent

Bill Gates Calls for ‘Human Reserved’ Jobs to Shield Workers from AI

by Tim Adams

Microsoft co-founder Bill Gates. Image: Wiki Commons

Bill Gates has called on policymakers to set aside a category of jobs that artificial intelligence and robots would be barred from performing, even in cases where the technology is fully capable of doing the work.

The Microsoft co-founder is proposing barring AI from certain professions and taxing automation to slow job losses.

The proposal, laid out in a lengthy essay published this week on his Gates Notes site, marks one of the most concrete policy ideas yet from the Microsoft co-founder on how governments might cushion the economic disruption AI is expected to cause.

Gates calls his proposal “Human Reserved” — a designation that would carve out certain occupations or tasks for people only, regardless of whether AI could technically perform them.

He likens the idea to nature reserves: land that could be developed but is deliberately left alone because society judges the loss too great.

Under his framework, the designation could take two forms. Some jobs might be permanently reserved for humans because people are simply better suited to them — he points to childcare and jury duty as clear examples.

Others might receive more temporary protection, aimed at workers too far into their careers to realistically retrain. Gates has argued it isn’t reasonable to tell a lifelong construction worker in his fifties to pivot into elder care and expect him to find the transition fulfilling.

Caregiving, teaching, and medical roles are among the professions he flagged as candidates for protection, though he stopped short of proposing an outright ban on AI in fields like education and healthcare.

Instead, he suggested those professions could remain human-staffed while workers use AI tools to boost their own output — a hybrid approach rather than exclusion.

Gates also acknowledged the idea is not universal. He noted that a country such as Japan, facing a shrinking workforce and a shortage of young caregivers, might welcome robotic caregiving in a way that other nations with different demographics would not.

Advertisement

Alongside the jobs proposal, Gates renewed his call for a “robot tax” — a levy on companies’ use of AI software and automation comparable to the payroll taxes employers pay on human workers.

His argument is that the current tax code actively rewards automation: a business that hires a person pays payroll tax on their wages, but one that buys a robot can typically write off the purchase as a business expense right away.

Leveling that playing field, he argues, would slow the financial incentive to replace people with machines.

Gates has described the scale of tax reform he has in mind as more sweeping than anything attempted in his lifetime — a heavy lift, he concedes, at a moment when political polarization is also at a high point.

The essay represents a notable change in Gates’ public stance on AI and employment. Three years ago, he characterized the coming disruption to jobs as “bumpy” but ultimately manageable, framing AI mainly as a productivity tool.

In recent interviews, including with Axios and Semafor, he has struck a more alarmed tone, saying AI has crossed concerning thresholds not just in labor markets but in areas such as cybersecurity and biosecurity risk.

He has also predicted that AI’s economic impact will be defined less by its speed than by its sheer breadth, arguing that virtually no occupation will be untouched and that humans will eventually be outperformed by AI in the majority of jobs.

Gates has been candid that his proposal raises as many questions as it answers. He has not specified exactly which jobs would qualify for “Human Reserved” status, who would make that determination, or how regulators would prevent companies from working around the designation.

Critics have pointed to the same practical hurdles: without clear enforcement mechanisms, a voluntary or loosely regulated set-aside could prove difficult to police.

Even so, some analysts have noted the idea would be relatively straightforward for lawmakers to adopt in principle, since it doesn’t require new technology or infrastructure — only a policy decision about which parts of the economy to shield.

The proposal lands amid growing public anxiety in the United States and elsewhere about AI’s effect on employment. Recent labor data has shown mixed signals: overall hiring has continued to grow, even as entry-level positions and certain sectors, such as call centers, show measurable strain from automation.

Our biggest stories, straight to your inbox

Sign up now to get the latest news from around the world

Popular

Podcasts »Visit ChannelC