From Cognitive Offloading to Cognitive Surrounding

Recently, I developed a Laravel package called laravel-env-settings. While building it, I found myself thinking about something much bigger than the package itself: how we, as developers, should work with AI.

AI can dramatically accelerate development, but there is a danger in simply offloading everything to it. If we continuously ask AI to build things we don't understand, we create a growing comprehension debt—the gap between what we understand and what AI is capable of producing.

Instead, I believe we should move from cognitive offloading to cognitive surrounding: use AI as part of our development environment, learn from what it produces, question it, and gradually make the gap smaller and smaller.

I'd really like to hear your opinion on this topic.

There's definitely a practical limit: we don't have time to learn everything, and if we became too restrictive about only using technologies we already fully understand, we might end up building something outdated—something that doesn't fully accommodate today's needs, possibilities, and expectations for creativity in the market.

At the same time, I think we should avoid letting AI continuously introduce new technologies into our projects without our control. When building something, we should keep the technology domain relatively constrained to what we already know and have a good command of, while allowing AI to occasionally introduce a few new technologies when they genuinely make sense.

Then, instead of blindly adopting them, we learn enough about them to understand what we are building and continue moving forward.

Otherwise, if we keep delegating everything to AI, the gap between what we understand and what AI can produce can grow silently—and eventually become a very difficult comprehension debt to pay back.

What do you think?