A customer calls a company with a routine billing question. An AI system finds the account, explains the charge and resolves the problem in seconds. The customer is pleased. The company saves money. But the employee who used to answer that call has a reasonable question: What happens to me?
I believe artificial intelligence will improve our lives and expand what people can accomplish. I also believe it will reduce employment in some fields. We do workers no favor by pretending otherwise. If we want the benefits of AI to reach everyone, we must prepare people for the jobs it changes and help them move into the opportunities it creates.
The first changes are already visible in work done primarily on a screen. Over the next few years, companies will need fewer people for routine customer support, data entry, transcription, basic bookkeeping, appointment scheduling and standardized document preparation. AI can also produce first drafts of advertising copy, translate ordinary business material, summarize meetings and handle parts of insurance claims and loan processing.
This does not mean everyone in those occupations will lose a job. Many employees will supervise AI, handle unusual cases and spend more time with customers. But when a team can process substantially more work with the same number of people, companies may hire fewer replacements as employees leave. For someone looking for a first job, reduced hiring can matter as much as a layoff.
Professional work will change too. Lawyers may use AI to review documents and draft routine agreements. Accountants may automate parts of tax preparation. Software developers may generate and test code more quickly. Physicians may receive help with records and diagnostic analysis. In each field, a qualified person must still exercise judgment and accept responsibility for the result. Yet firms may need fewer hours of junior work to complete the same assignment. That creates a difficult question: if AI takes over the entry-level tasks, how will the next generation gain experience?
Over the following decade, the effects could spread further as AI is combined with robotics and more capable business systems. Warehouses, manufacturing plants, retail operations and transportation may automate more tasks. The pace will vary. Moving a piece of information through a computer is easier than navigating a crowded street or caring for a patient. Physical work also depends on equipment costs, safety rules and public acceptance. We should be skeptical of anyone who claims to know exactly how many jobs will disappear by 2036.
Research supports both concern and caution. The International Labour Organization’s analysis of generative AI and jobs finds that exposure is widespread, especially in clerical work, but that many jobs are more likely to be transformed than completely eliminated. The World Economic Forum’s Future of Jobs Report 2025 records employers’ expectations of substantial job creation and displacement through 2030. Those figures are forecasts across many economic forces, not a count of jobs lost to AI alone.
There will also be growing demand for work AI cannot easily provide on its own. Healthcare, elder care, education, skilled trades, cybersecurity and the building and maintenance of digital and energy infrastructure all need people. The U.S. Bureau of Labor Statistics projects growth through 2035 in occupations including nurse practitioners, data scientists, information security analysts and several hands-on healthcare roles.
But a displaced billing clerk cannot become a nurse practitioner simply because a report says demand is growing. The new job may require years of education, licensing and money the worker does not have. New jobs may arise in a different city. They may pay less at first. That gap between an economic forecast and a family’s daily life is where public policy must work.
I would begin with employers. When a company introduces AI, it should identify which tasks will change and offer employees training before positions disappear. A customer-service employee may be able to move into complex case resolution, sales support, quality assurance or supervision of AI responses. A bookkeeper may learn to investigate discrepancies and advise small businesses. People already know their industry and customers. That knowledge is valuable, even when the tools change.
Education must become more flexible. Community colleges, universities, unions and employers should offer short programs tied to actual openings, with classes available outside normal working hours. Students should learn how to use AI, check its work and protect sensitive information. They also need strong foundations in writing, mathematics, practical problem-solving and working with other people. Training should lead to a credible qualification or a job interview, rather than another certificate with little value in the labor market.
Government has a role in making transitions possible. Workers who lose jobs should be able to obtain income support, career guidance and affordable training without navigating a maze of programs. Where a new occupation initially pays less, temporary wage insurance could soften the decline. Portable health coverage would make it easier to change employers or start a business. Local governments should track where displacement is occurring and direct support to communities before prolonged unemployment takes hold.
We should also make it easier to create new businesses. AI can give a small company capabilities once available only to a large corporation: research assistance, customer support, translation and software development. With access to capital, practical training and fair competition, displaced workers can become entrepreneurs and employers themselves. I have spent much of my life building companies, and I know that a new enterprise often begins when someone sees a need that established businesses have overlooked.
The broader economy will benefit if AI raises productivity and makes useful services more affordable. But productivity alone does not guarantee shared prosperity. If the gains flow only to a small number of companies while wages and opportunities weaken elsewhere, consumer spending and trust in our institutions will suffer. Business leaders and policymakers should measure success by rising incomes and new opportunities as well as by output and profit.
We have time to act, though we do not have time to waste. AI will take over some tasks, reshape many occupations and help create work we cannot yet fully describe. The outcome will depend on choices made by employers, educators, governments and workers themselves.
I remain optimistic about AI because I believe in human ingenuity. That optimism carries an obligation: when we build tools that make our economy more productive, we must also build a path for the people whose work those tools change.
— Dr. Mohan Ananda
Founder, DRAI Health
Scientist • Entrepreneur • Policy Innovator