In today’s world, where AI is used almost everywhere, we sometimes forget to ask ourselves a crucial question: How does it affect us?
Has it had a negative or positive impact on us?
Have you used AI to learn about a topic or to get a more detailed explanation of it?
Good, then you’ve learned something.
Or have you used AI to have it do the work for you without learning anything in the process? In that case, you may have gained convenience, but at the expense of knowledge.
The impact of AI on us depends on how we use it.
It can be a powerful tool for learning and productivity, but it can also make us lazy and lead us to rely too heavily on technology.
The key is to find a balance and use AI in a way that complements our own abilities rather than replacing them.
For example, when AI is used to summarize articles or videos for you, it has a significant impact on the way you think.
Your attention span is becoming shorter, and your tolerance for complexity, nuance, and cognitive friction begins to erode.
Fortunately, there’s one invaluable skill that most people overlook: reading.
I’m talking about deep, focused reading of long-form, where you sit quietly for an hour or two and fully immerse yourself in a single flow of thought.
If you rely exclusively on AI summaries, you outsource the mental effort of understanding.
But it is precisely this intellectual effort, the cognitive friction that arises when you grapple with challenging literature and build mental models from the ground up, that is where deep understanding takes place.
AI can provide you with the conclusions, but it cannot impart understanding.
In an age of hyperefficiency and instant answers, deep reading is no longer just an intellectual pastime. It is cognitive resistance.
This is how you protect your ability to think independently, grasp nuances, and concentrate, while the rest of the world practices merely scratching the surface.
The same goes for software engineers. I’m a software engineer myself, so this topic is very important to both me and, hopefully, all other software engineers.
A software engineer takes their craft very much to heart.
Now, I could write another article on this topic, but I’ll try to keep it brief.
Coding agents are useful, but you should also be aware of their consequences.
Using coding agents to write boilerplate code can save time and make you more productive. However, it can also lead to a decline in your problem-solving skills and understanding of the underlying code by relying on them too much and also the other effects I've mentioned before.
When you use coding agents, you outsource the mental effort of understanding.
But it is precisely this mental effort that is crucial. It is crucial because it is the only way your brain can learn and build new patterns.
If you want to learn something, like playing the piano, you have to put in mental effort and practice consistently; that’s how your brain learns.
However, if you let coding assistants do the work for you and do nothing else, you’re outsourcing the mental effort of understanding.
This can be particularly detrimental for emerging software engineers. They’re in the process of building their fundamental skills, and relying too heavily on coding agents can hinder their growth and their understanding of core concepts.
When you use coding agents, use them to learn, not to replace your own understanding. Use them as a tool to support your learning process, not as a crutch to bypass it.
Ask them to help you understand, rather than just provide answers, by asking the right questions. For example, “Can you explain why this code works this way?” or “What principles underlie this approach?”
A software engineer should always have a solid understanding of the fundamentals. If you don’t know them, learn them. If you dive into other topics without knowing the fundamentals, it’s like building a house without a solid foundation, you’ll end up with a mess.
Understanding the fundamentals is the key to mastering any topic. It’s the foundation for everything. Try to use coding agents only in moderation. Try to code for two days a week without them, and see how that affects your understanding.
Finally, remember that coding agents are just tools and cannot replace your own understanding. Use them wisely and continue to challenge yourself to grow as a software engineer.
Thank you for reading. In the next article, we'll look at how software engineers can use AI correctly.