Is AI Making Our Next Techs Smarter or Just Faster?

Young technician in a dim server room thinking, lit by a laptop, with an experienced colleague in the background

There is a question I keep hearing from MSP owners, usually late in a conversation and a little quietly. Is AI making people less intelligent? More specifically, are the junior techs walking through our doors arriving with the foundations they need to grow, or are they arriving already wired to hand every problem to a chatbot? It matters because the skills shortage isn’t going away. Every hire has to become someone the business and its customers can rely on.

AI amplifies the habits people already have

I don’t think AI on its own is making anyone dumber. Shallow learning existed long before Copilot. What AI does is amplify whatever habits a person already has. A curious junior with AI learns faster than any junior I could have hired ten years ago. An incurious junior with AI simply gets to the wrong answer with more confidence.

Picture a new tech handling a ticket. A user can’t get into Outlook, and there is an Entra sign-in error on the screen. The tech pastes the error into Copilot Chat, gets a plausible fix and applies it. The user is back in. Ticket closed. But ask that tech why it failed, whether it was conditional access, a stale token or something else, and they can’t tell you. Worse, they may have weakened a control without realising. The tool did its job. The fundamentals simply weren’t there to check it.

Test for fundamentals when you hire

That means the interview has to change. If candidates can use AI on the job, a good CV and a polished take-home task tell you very little. Here is what I would look for instead:

  • Ask them to explain, with no tools, what happens between typing a web address and a page loading. DNS, routing and certificates should come up in some form.
  • Have them whiteboard how email gets from one mailbox to another, and how sign-in and identity fit together in Microsoft 365.
  • Give them a vague fault and ask them to walk through a troubleshooting path out loud. I care more about the order of their questions than the final answer.
  • Show them an AI answer that is subtly wrong and ask them to find the problem.

That last one tells you a lot. I’m looking for curiosity, and for someone who explains the why rather than reciting the what. I wrote earlier about why everyone needs basic technology troubleshooting skills, and for anyone in a technical role that bar is far higher.

Make Copilot a tutor, not an answer machine

Once they’re on board, the same tool that can hollow out learning can accelerate it. Encourage juniors to ask Copilot to explain a concept, then quiz them on it. Get them to ask “why would this fix work?” before they apply it. Have them summarise what they learned in a short note in Teams or OneNote that a senior tech can read. Treat Copilot as a patient teacher available at two in the morning, not as a vending machine for fixes.

Hire the habit, then feed it

AI is not the enemy of capable technicians. It’s a multiplier, and it multiplies whatever it’s given. So when you hire, test for foundations and curiosity, not just speed. Then give those people AI as a tutor. The MSPs that do this will grow techs who understand the work. The rest will have staff who can close tickets but can’t explain them, and customers eventually notice the difference.

Leave a Reply