Don’t let AI make you useless

Imagine a lawyer who used to spend hours reading contracts, identifying risks and writing legal arguments. Today, this same lawyer just asks AI to review the contract, identify the risks and prepare the response. The lawyer, by using AI tools, has become faster and more productive. Nevertheless, the lawyer is also becoming lazier and too comfortable and is at risk of becoming useless in the future. Quite specifically, being replaced by an AI-powered machine that is doing the job.
Let us further examine the case. What if this same lawyer, after becoming dependent on AI tools for the job for a long period of time (months), now receives answers from AI that the lawyer cannot recognise as wrong? Imagine that. The lawyer, having probably stopped practising the skill, is now no longer able to challenge the machine. Eventually, the lawyer becomes useless and easily replaceable.
Another example is a software developer who once understood how to build an application from the ground up. Today, AI can generate the code, explain errors and even suggest the architecture. The developer, in return, is also faster and more productive and can now deliver projects much faster. Nevertheless, what happens if the software developer increasingly accepts whatever AI produces without understanding the underlying code? A serious problem can arise where the developer may be able to operate AI but no longer be able to develop without it. Again, the importance of the software developer may be reduced.
This is the paradox of AI. The technology designed to make us more productive can, if we are not careful, gradually make some of our own skills unnecessary. This includes skills that we may seriously need when AI gets something wrong.
What does it mean to become “useless” because of AI? It does not necessarily mean losing your job. It means losing the ability to perform the fundamental parts of your profession without depending on a machine. You become excellent at asking AI for answers but weaker at producing, questioning or defending those answers yourself. And this is not just a theoretical concern.
A 2025 study by researchers from Microsoft and Carnegie Mellon University surveyed 319 knowledge workers and examined 936 real-world examples of AI use at work. It found that greater confidence in AI was associated with less critical thinking. The researchers also observed that AI shifts human effort from doing the work towards verifying, integrating and supervising AI-generated work.
So, how do we avoid becoming professionally dependent on AI? My humble opinion is to simply never outsource the skill you want to keep.
Use AI to challenge your thinking, not replace it. Before asking AI to solve a problem, try solving it yourself. Ask AI to critique your idea rather than immediately giving you one. When it writes something for you, read, verify and improve it. Occasionally do important tasks without AI simply to keep the muscle exercised.
Most importantly, remain curious about your profession. Learn the fundamentals. Understand the technology. Keep practising.
AI should become your copilot, not your brain. Because the future may not belong to people who compete against AI. It may belong to people who know how to use AI while still being capable without it.
Truth be told, AI has revolutionised the way we work and has made many people more productive. Nevertheless, the more I personally use AI and see the way people use it, the more I fear that dependence on AI may make us useless.
Until we catch up again next week, avoid becoming too dependent, but master and appreciate the professional skills you already have.