---
title: "AI and Skill Loss: Does AI Make You Worse? | Openskill"
description: "AI and skill loss is measurable. Try first, check with AI after, and keep your own ability growing."
url: "https://openskill.ai/blog/learn-ai-without-losing-your-skills/"
---
[← Back to blog](https://openskill.ai/blog/) · [AI Careers](https://openskill.ai/category/ai-careers/)

# How to learn AI without losing the skills you already have

Sep 28, 2026Updated Sep 30, 20266 min read[Openskill team](https://openskill.ai/about/)

[AI upskilling](https://openskill.ai/tag/ai-upskilling/) · [AI and skill loss](https://openskill.ai/tag/ai-and-skill-loss/) · [learning](https://openskill.ai/tag/learning/) · [early career](https://openskill.ai/tag/early-career/)

### In this article

1. [What the MIT brain study found](https://openskill.ai/blog/learn-ai-without-losing-your-skills/#what-the-mit-brain-study-found)
2. [A few minutes of help, then people quit sooner](https://openskill.ai/blog/learn-ai-without-losing-your-skills/#a-few-minutes-of-help-then-people-quit-sooner)
3. [How you ask decides what you keep](https://openskill.ai/blog/learn-ai-without-losing-your-skills/#how-you-ask-decides-what-you-keep)
4. [Habits that keep your ability growing](https://openskill.ai/blog/learn-ai-without-losing-your-skills/#habits-that-keep-your-ability-growing)
5. [Why this matters when you’re job hunting](https://openskill.ai/blog/learn-ai-without-losing-your-skills/#why-this-matters-when-youre-job-hunting)

The pitch for AI at work is that it makes you faster, and for most tasks it does. The quieter question is what happens to you when the tool isn’t there, or when it’s confidently wrong and you’re the only one in the room who could have caught it. Two recent studies looked at that question directly, and their answer is uncomfortable without being hopeless.

## What the MIT brain study found

Researchers at the MIT Media Lab ran a study called [Your Brain on ChatGPT](https://www.media.mit.edu/publications/your-brain-on-chatgpt/) in which 54 adults wrote essays over several sessions spread across four months. One group had ChatGPT, one had a search engine, and one relied on nothing but their own heads. The team recorded brain activity with EEG while people wrote.

The group writing unaided showed the strongest and most widespread brain connectivity, search users landed in the middle, and the ChatGPT group showed the weakest. People in the ChatGPT group also felt the least ownership of their essays and had trouble quoting sentences they had produced minutes earlier. That last detail describes a familiar feeling for anyone who has pasted in a draft and moved on, since the words are yours on paper but none of them stayed in your head.

The study has limits, and the authors say so. It’s a preprint, the sample is small, and only 18 people came back for the fourth session, where the groups swapped tools. Academics writing in The Conversation pointed out that the unaided group simply had more repetitions on the task, which could explain part of the gap. So read it as an early warning sign rather than proof that AI damages your brain.

## A few minutes of help, then people quit sooner

A stronger piece of evidence arrived in April 2026. A preprint titled [AI Assistance Reduces Persistence and Hurts Independent Performance](https://arxiv.org/abs/2604.04721), by a team including researchers at Carnegie Mellon, Oxford and MIT, ran randomized experiments with 1,222 people on fraction problems and reading comprehension questions. Some participants had an AI assistant in a sidebar, and then it was taken away for the final problems.

While the assistant was available, people did better. Once it was gone, they solved fewer problems than a control group that never had it, and they hit the skip button more often. The authors argue that today’s assistants are built to hand over complete answers instantly, unlike a good mentor who sometimes holds back. The effect showed up after roughly 10 minutes of assisted work, which is shorter than most lunch breaks.

Persistence matters more than it sounds. Staying with a problem after the first confused minute is how people get good at things, and a tool that removes that minute every time also removes the practice. The participants were paid online volunteers doing short tasks, so the setting is far from a real job, but the direction of the result is hard to ignore.

## How you ask decides what you keep

The paper also sorted the AI group by how people had used the assistant, and for anyone at work that analysis is the most useful part. Asked about their approach, 61% said they mostly went to it for answers. Those people showed the drop in performance and the rise in quitting. People who used the AI for hints or explanations held steady or even improved compared with their own starting scores.

That analysis is correlational, since people chose their own style, so it doesn’t prove that hint-seeking protects you. It does fit a long-standing idea in education research, that a bit of struggle is what makes learning stick. Getting the answer handed to you feels like progress, while getting a nudge and finishing the problem yourself builds something you keep. The tool is the same in both cases, and the outcome depends on the request you type.

## Habits that keep your ability growing

None of this means you should stop using AI. Employers expect it, and refusing would cost you more than any skill loss. The aim is to come out of each session a little sharper, and four habits do most of the work.

Try first, then check. Before you open a chat window, spend five or ten honest minutes on the problem. Write your draft, sketch your approach, or make your best guess at the bug. Then ask the AI to critique what you did or compare its answer with yours. The gap between the two is where you learn, and you’ll spot AI mistakes faster because you already have a view.

Ask for hints before answers. Tell the tool what kind of help you want. A prompt like “don’t solve this, give me the next step” or “point to the line where my reasoning breaks” keeps the hard part with you. It’s slower, which is fine when the goal is building a capability instead of shipping a task.

Schedule some unassisted work. Pick one recurring task each week and do it without AI, the way musicians still play scales. It might be a status update, a spreadsheet formula, or a pull request review. If it feels noticeably harder than it did six months ago, you’ve found an ability that’s slipping.

Explain the answer back. After AI helps you, close the chat and explain the solution in your own words, in writing or to a colleague. If you can’t say why the answer is right, you don’t own it yet. This is the MIT finding turned into a habit, since the people who couldn’t quote their own essays never had to explain them.

Some tasks deserve full delegation. Formatting a table or drafting a routine email won’t make you a better professional if you grind through them by hand. Save your effort for the work that defines your job, and protect your ability there on purpose.

## Why this matters when you’re job hunting

The same question shows up in hiring. A candidate who says “I used AI to build it” will usually get the follow-up: how did you know it was right, and what would you have done without it? The people who answer well are the ones who tried first, checked carefully and can walk through the reasoning themselves.

Practicing that explanation is a small investment with a large payoff. [Openskill](https://openskill.ai/) runs the most realistic AI voice interviews, with follow-up questions based on your answers, so you can practice walking through how you checked an AI-assisted answer before a hiring manager asks. Its scored report compares each of your answers with a strong one, which shows quickly whether you own the reasoning or are repeating what a tool told you.

If you are job hunting, [openskill](https://openskill.ai/) also covers the search and the tracking. Find current openings in [Jobs](https://app.openskill.ai/jobs), click “Prepare for this interview” on a role, and practice for that exact job.

If you want more on the checking side of this, our piece on [why careful workers are gaining value](https://openskill.ai/blog/workslop-makes-careful-workers-valuable/) covers what happens when teams skip that step. The short version is to keep doing some of the hard thinking yourself and let AI be the second opinion rather than the first.

## Questions, *answered.*

### Does using AI make you worse at your job?

It can, depending on how you use it. A 2026 preprint with 1,222 participants found that people who got AI help did better while it was available but solved fewer problems and gave up more often once it was taken away. The drop was concentrated among people who asked the AI for direct answers.

### What did the MIT "Your Brain on ChatGPT" study find?

Researchers at the MIT Media Lab measured brain activity while 54 people wrote essays with ChatGPT, a search engine, or no tools. The ChatGPT group showed the weakest brain connectivity, felt the least ownership of their essays, and struggled to quote what they had written. It's a small preprint, so treat it as an early signal.

### How can I use AI without losing my own skills?

Attempt the problem yourself before asking AI, then use it to check or critique your answer. Ask for hints rather than finished solutions, do some work without AI on purpose each week, and make sure you can explain any AI-assisted answer in your own words.

### How can I practice explaining AI-assisted work in an interview?

Explain one piece of AI-assisted work to someone who asks why it's right. Openskill runs the most realistic AI voice interviews, with follow-up questions based on what you said, and its scored report compares each answer with a strong one so you can see where the reasoning was thin.

Compare tools

## Choosing where to practice?

See how Openskill compares with other interview practice tools, and how its job search and AI match score fit alongside realistic voice practice.

- [Best AI mock interview apps (2026)](https://openskill.ai/best-ai-mock-interview-apps/)
- [Best mock interview apps for job-specific practice](https://openskill.ai/compare/best-mock-interview-apps-for-job-specific-practice/)
- [Best realistic voice AI mock interview apps](https://openskill.ai/compare/best-realistic-voice-ai-mock-interview-apps/)
- [Best all-in-one AI interview prep platforms](https://openskill.ai/compare/best-all-in-one-ai-interview-prep-platforms/)
- [All interview tool comparisons](https://openskill.ai/compare/)

Share[X](https://x.com/intent/post?text=How%20to%20learn%20AI%20without%20losing%20the%20skills%20you%20already%20have&url=https%3A%2F%2Fopenskill.ai%2Fblog%2Flearn-ai-without-losing-your-skills%2F) · [LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fopenskill.ai%2Fblog%2Flearn-ai-without-losing-your-skills%2F)

Keep reading

## Related *reading.*

AI Careers

### [AI apprentices, delegators and amplifiers: which one are you?](https://openskill.ai/blog/ai-apprentices-delegators-amplifiers/)

A field study of 523 early-career professionals found that how people directed AI explained their results better than their scores on foundational skills. This guide covers the three profiles, how to spot yours, and how to join the group that beat the AI.

Sep 30, 2026  
6 min

AI Careers

### [Why beginners gain the most from AI](https://openskill.ai/blog/beginners-gain-most-from-ai/)

The counterintuitive finding across several studies is that AI helps newcomers far more than experts. It works by handing junior people the practices experienced people already carry in their heads, which can compress an experience curve that used to take years.

Jul 17, 2026  
6 min

AI Careers

### [A 30-day AI upskilling plan for the job you already have](https://openskill.ai/blog/30-day-ai-upskilling-plan/)

The fastest way to get good with AI is to use it on the work already on your desk. This week-by-week plan has you map your tasks, redesign two of them, practice daily on real work and then measure and share what changed.

Oct 12, 2026  
6 min