AI

Right now, I feel like the whole world is surrounded by AI. Everyone is talking about AI—whether it's influencers, founders, investors, or even normal people. There's this story that with AI, we can make money, startups can get funding, companies can gain market share, and governments can enhance their efficiency. It seems like with AI, we can access information faster and more easily than before. We can even earn more money.

However, I don't think so.

It seems like we can get information more efficiently with AI — something that has already summarized the context and hands us complete notes. The appeal is obvious: I can dump all my course materials into a chat window and walk away with a ready-made lecturer — one that already knows the key points, simplifies the hard parts, and tells me exactly what to study before an exam.

The worse: there are even many influencers who feel like cheaters—they use AI to summarize information, reduce the cost of creating posts on social media, and present it as "fully organized notes" or something similar, without even processing the information through their own brains. With so much AI-generated information in my daily life, I start to question what is actually being learned.

But why do I feel like I never actually learn anything new?

I still have so many things needed to learn. A lot of the information feels vague—it's like I've heard it before, but I can't explain it clearly. I still have a lot of homework before the deadline. With AI, it feels like I can finish things faster than before, but why does it still take me so long—and I end up gaining less knowledge than before? I am not learning anything new.

AI has deprived me of my ability to truly learn new things and to process information deeply.

Take a look at one of my recent commits:

Who else on Earth can map out an entirely new product requirement and implement it flawlessly in under ten minutes? Nobody used to be able to do that. Yet today, that is the standard we are told to simply accept.

Am I really learning anything new here? No. I am just using AI to generate the thoughts through code for me. Previously, if I wanted to build a large-scale feature like this, I would have had to spend a significant amount of time reading documentation, understanding the codebase, searching for related resources in the open-source community, and then writing the code by hand. Now, I can just let AI generate it for me.

With AI, I seemingly save a lot of time building large-scale features, but in reality, I am not learning anything new. I am merely using AI to generate information. Worse yet, I am losing my sense of oversight because I find myself blindly clicking the "accept" button without even reading the code, turning AI's output into an addictive shortcut.

This concern isn't unique to me. Lee Robinson—a developer educator, formerly at Vercel and now at Cursor—recently warned that tons of AI-generated code is becoming a liability, and that engineers still need to be responsible for the code that gets deployed, regardless of who—or what—wrote it.

What stood out to me was the idea that AI can dramatically increase output without guaranteeing understanding. The bottleneck is no longer producing code or content—it is developing the judgment necessary to evaluate, direct, and improve what AI generates. If we outsource all of our thinking to AI, we may become more productive while simultaneously becoming less capable.

This has become a normal daily occurrence for many of us.

Faced with this disappointment, I started to question whether this must be the working style of the future. As Viktor Frankl once wrote, "When we are no longer able to change a situation, we are challenged to change ourselves." This era of AI is exactly that kind of situation—how we grow within it is the only thing left in our control.

Therefore, what I do now is after AI delivers a result, I try to ask it more questions, treating it like a real lecturer. I ask why it chose a specific approach. Fortunately, the silver lining is that I have a lot of raw, hands-on experience from the past. Since this code uses Next.js and React, I am incredibly familiar with how it works.

My extensive hands-on coding in 2024 gave me the technical foundation I need to actually supervise and guide these AI tools today.

This foundational experience means I can push myself to a more senior level. I now have a fleet of junior developers (AI agents) helping me build code, which allows me to inject my own critical thoughts and practical experience to make the system more efficient and scalable.

This is exactly what I want to achieve in this era. It is no longer just about gaining hands-on programming experience, but about developing soft skills and leadership. Because more and more manual development will be replaced by AI agents, we must shift our focus to higher-level architecture and strategy.

To summarize, AI can absolutely deprive us of our ability to learn new things and process information deeply. However, it also presents a rare opportunity to overhaul our workflow, elevate our skills to a higher level, and consciously decide what we truly want to build.