Is anyone safe from AI?

The U.S. job market remains stable: the unemployment rate was 4.3% in April, and the economy added about 115,000 new jobs, suggesting that fears of AI replacing humans were exaggerated and that AI has become a tool that helps professionals rather than replacing them.

But that is only on the surface, as shifts do produce, especially in the tech sector, where, according to Layoffs.fyi, more than 165,000 workers have been laid off over the past year. Even big names are cutting the staff: Microsoft shut 15,000 jobs last year, Amazon reportedly eliminated 30,000 positions in the last six months, Meta laid off more than 1,000 employees and may cut even more, and Oracle also recently laid off thousands of workers.

The financial sector is also under pressure. Bank of America cut about 1,000 employees, Citi plans to reduce around 20,000 staff, Morgan Stanley cut about 2,500 workers, and Wells Fargo cut about 4,000. Bank executives say AI helps automate tasks, from processing documents and credit applications to preparing financial reports and communicating with clients.

Now the head of NVIDIA claims that some of the most secure and valuable jobs in the AI era may be electricians, plumbers, construction workers, mechanics, and technicians, the people who build and maintain the physical infrastructure that AI depends on.

So, is it time to go back to studying?

Indeed, if we ask AI about the prospects of a company, the S&P 500, or the Nasdaq, it can give a detailed and convincing answer in seconds. The problem is that AI does not actually think. It predicts patterns based on existing data. In situations without clear answers, this can be counterproductive.

The same applies to coding. AI can speed up development, but it can also make mistakes that cause major losses. Earlier this year, reports surfaced about an AI coding agent that accidentally deleted an entire company database within seconds while trying to automate tasks.

That is why human oversight still matters. Someone needs to understand what to check and which questions to ask. Otherwise, the result may look correct while being completely wrong underneath.

There is also a chance that governments will intervene in AI usage by promoting gradual adoption, retraining workers, and helping people adapt to new industries to prevent social unrest and reduce inequality. 

On a personal level, the only real option is to keep learning and stay flexible. In the AI era, adaptability may matter more than specialization.