Why Tech Writers Should Care About AI and Critical Thinking

Researchers discovered that using AI made work feel easier — maybe too easy

There’s a new Microsoft Research paper called The Impact of Generative AI on Critical Thinking. It basically asks: now that AI can write our drafts, have we forgotten how to think?

The researchers surveyed 319 knowledge workers — the kind of people who use “ideate” unironically — and found that the more we trust AI, the less thinking we do.

Confidence in the machine goes up; cognitive effort goes down. Apparently, ChatGPT is like that friend who always offers to “help” and somehow ends up doing the whole group project.

What The Research Uncovered

The researchers discovered that confidence is a funny thing — especially when you hand it over to a machine. The more people believed the AI knew what it was doing, the less they felt compelled to double-check its work. It’s like trusting a toddler with scissors because he looks confident.

Meanwhile, people who actually trusted themselves — their own judgment, experience, and ability to spot nonsense — behaved differently. They stayed skeptical, poked at the AI’s output, and asked, “Really?” instead of “Cool, done.” In other words, they thought like adults in a room full of autocomplete suggestions.