agentiseraAI and the mental gym.
It becomes easier to get an answer. Understanding it still requires work. What happens to our will to learn when we can take a shortcut past the thinking?
Access to knowledge is a beginning. Understanding requires practice.
Sometimes we use AI as a glorified Google: ask a question, get an answer, move on. Other times we ask it to write, analyze, or take several steps toward a result. I am still learning what the tools can do, where the boundaries lie, and what they will become.
But the question I keep coming back to is less about the technology and more about us. What happens when we can get a result without having to understand how we arrived at it?
A finished report is not the same as understanding
Imagine asking AI to create a report and then passing it on without really knowing what it contains. The document exists. The task seems complete. But who has verified the claims, understood the assumptions, and judged whether the conclusions hold?
It's not just about AI being able to make mistakes. It's also about whether I know enough to detect those mistakes. A well-formulated answer can feel finished long before I have done the work required to trust it.
And what happens if the recipient uses AI to summarize the report and write a response? We could end up with AI-generated work going back and forth, without any of us having truly engaged with the content. What have we achieved then?
The intention must come from me
When I use AI, the intention must come from me. What do I want to achieve? Why does it matter? What would make the result useful, and what would make it incorrect?
Guiding a language model isn't just about getting it to give the answer I want to hear. Then it risks becoming a machine that confirms my own assumptions. I need to ask it to question the question, show alternative explanations, and help me check claims against sources that I actually review.
I can hand over a task. But I still need to own the intention and judge the result.
Where do we train our thinking?
Many of us have less physically demanding jobs than people had in the past. If daily life doesn't provide us with enough movement, we need to create space for it intentionally. The gym is one way to do that.
I wonder if we will need a similar approach to mental effort. Not because thinking functions exactly like a muscle, but because learning requires us to do something ourselves: try, grapple with a problem, make mistakes, and rethink our understanding.
If I directly ask AI every time I get stuck, I might be skipping the very part that would teach me something. Having access to an answer is not the same as being able to explain it, question it, or use it in a new situation.
What happens to the junior?
In certain areas, ChatGPT or Claude can provide a broader or more well-formulated answer than someone at the beginning of their career. That doesn't mean the model has the same judgment, context, or responsibility. But it raises a question: where does the beginner practice?
If we automate all the tasks that previously helped a junior develop, how does that person then build the experience needed to later guide and judge the more complex work? Access to an impressive answer does not automatically create the ability to recognize a bad one.
Use AI without taking all the shortcuts
I don't think the answer is to avoid AI or preserve tedious tasks for their own sake. I want to let it handle repetition and help me explore things that would be difficult to tackle alone. But I also want to retain a portion of the demanding work: formulating the problem, weighing alternatives, guiding agents, and checking what comes back.
Sometimes that might mean trying it yourself first and asking for feedback afterward. Sometimes asking for a hint instead of a solution, or closing the chat and explaining the thought in your own words.
My guess is that those who benefit the most from AI will be those who embrace the technology while simultaneously using it to deepen their own knowledge. Not just producing more, but becoming more capable themselves.
We can appreciate the moment when something finally becomes clear. I don't want to lose that by making every answer effortless. The question isn't just what AI will be able to do. It's also what I want to continue learning myself.