The Microsoft co-founder's comments are being shared online, but the message behind them goes beyond a programmer being overtaken by AI.

At the Royal Society in London, Bill Gates talks to The Observer's James Harding about AI. The interview published on October 25, 2026 recorded a notable observation: at the end of 2025, he realized that the programming skills he had pursued since the age of 13 were basically outdated.

Harding asked if one could still be the best coder in the world. “No”, Gates replied, then said that time had passed.

The answer gives the discussion of AI a very unique nuance. The character speaking is the co-founder of Microsoft, who is famous for software. When he himself looks back at a skill that was once the foundation of his career with that reticence, the story easily touches the worries of those who have spent many years becoming good at a profession.

What Gates worries about is the speed of human adaptation In an article on Gates Notes dated August 26, 2026, Gates admitted that AI still makes mistakes, but said that reliability is improving rapidly. According to him, software in particular may have more demand as costs decrease, helping to limit net job losses if people can still do better at some stages such as design.

He is especially worried about young people as opportunities in entry-level positions narrow. Gates writes: “Waiting until everyone is out of work or underemployed will be too late.”

He also describes competitive pressure: a company using AI to lower costs can force competitors to follow suit. From that perspective, retraining and support for workers needs to begin before widespread replacement takes place.

The point worth discussing here is the gap between two times: when a task is automated and when the person who used to do that task finds a new job. In that gap, rent, children's tuition and living expenses still have to be paid. A course can open doors, but completing a course in itself is not a promise of employment.

This is also the reason why it is necessary to separate the ability to learn tools from the ability to change careers. A person can quickly learn how to use an AI assistant, while gaining experience in another field requires time, practice opportunities, and a place that accepts new hires.

These are signals that need to be monitored, not yet conclusive of mass job loss A study by Erik Brynjolfsson, Bharat Chandar and Ruyu Chen at Stanford, updated August 2026, analyzed ADP payroll data through June 2026. The group has not seen evidence of widespread job displacement across the economy.

However, employment of 22-25 year olds in occupations strongly influenced by AI is 19% lower than what would be assumed if they grew at the same pace as those less affected. This is not the rate of 19% of young workers being fired. The difference mainly comes through a reduction in new recruitment. The authors emphasize that this is a descriptive observation and does not prove cause and effect.

This distinction has implications for how the warnings are read. A forecast about the future should be presented as a forecast. A change in employment data needs to be properly interpreted. If the two things are merged into one, the reader can easily understand that the most worrying scenario has happened and there is no longer any possibility of change.

Who pays for the transition period?

In an interview with The Observer, Gates used the phrase “Hire robots, fire people” to describe the economic dynamics that worry him. He proposed the idea of ​​dedicating some jobs to humans and taxing AI, including a token-based option, to help fund the safety net.

Those proposals open up a larger debate about who is more proficient in using technology. If productivity increases, will the results come from lower prices, corporate profits, worker income or shortened working hours? Each allocation leads to very different experiences for the same technical advance.

For a working person, the concern may be whether his or her position still exists or not. For businesses, it is quality and cost. For policymakers, questions also include funding sources to support career changers and how to preserve opportunities for those who have not yet entered the labor market.

Returning to Gates's admission, its weight lies in the feeling that a once-proud ability suddenly no longer warrants the same position as before. But from that feeling to a better future is still a gap that needs to be filled with specific decisions: what to study, how to organize work and what resources to support affected people.

Source: The Observer, Gates Notes and Stanford research. Quotes translated from English.