Since joining the company, I've read an enormous number of tech articles—IT, economics, programming, you name it. Using a Chrome browser plugin called Surfit, I was able to discover many useful articles. Through countless bookmarked company blogs and individual bloggers who write excellent content, I spent time continuously reading their thoughts and reflecting on my own.

When I do a rough calculation, I've read well over 2,000 articles. On heavy reading days, I'd go through 5 to 10 articles. Conservatively averaging 3 per day, that's 5 days a week (×5), 4 weeks a month (×4), 12 months (×12)—about 720 articles per year. Having been at the company for roughly 2 years and 9 months, that comes out to approximately 2,000 articles. Here's what I've learned from this experience.

Speed Reading and Patterned Information Digestion

When you read a lot of articles, you naturally develop your own way of digesting content. Unless it's a purely personal blog, professional article platforms and company tech blogs tend to craft content with a clear narrative structure for their readers. So you mostly encounter well-organized, polished writing.

Certain paragraphs contain the key points. I'd identify and prioritize understanding those paragraphs first, then gradually expand to surrounding paragraphs when deeper understanding was needed. This approach not only allows you to read articles quickly but also helps you summarize content by importance.

Over time, your reading speed and overall comprehension improve. I believe this is incredibly helpful for quickly digesting the outputs that AI produces in our current era. In fact, I've found myself unconsciously applying this same pattern to long responses from ChatGPT or Claude—identifying the core paragraphs first, then expanding into details as needed.

Unconscious Competence in Practice

Through this process, I naturally developed an awareness of domestic and international trends in both business and technology. If you combine this broad awareness with the ability to quickly absorb deep theory and information through AI, I believe you get close to achieving what's called unconscious competence.

When I'm given a completely unfamiliar task in a frontend development project, my thought process goes like this: First, I review what can be leveraged from previous experience. At this stage, I quickly scan whether past projects had similar domains or technical requirements, what approaches were used, and what problems and solutions were encountered.

Next, for entirely new aspects, I use AI to summarize and concretize the goals that need to be achieved. What's important here is the process of transforming vague requirements into clear, measurable objectives. For example, if someone asks to "improve user experience," I use AI conversations to specify exactly which metrics need improvement, what the current bottlenecks are, and what results are expected after the improvements.

Once the goals are concrete, I check whether they can be merged with previous experience. In this step, I evaluate how existing architectures, component structures, and state management patterns can be applied to new requirements. Sometimes existing approaches can be directly extended, but other times I realize a completely new approach is needed.

Then I proceed with the actual design, considering detailed layers like scalability, optimization, and efficiency. Here, I don't just solve the immediate problem—I comprehensively consider how flexibly we can respond when similar requirements arise in the future, whether there are potential performance bottlenecks, and how complex it would be for other team members to maintain.

At this point, I use AI to review whether I've missed anything or if better alternatives exist. AI can objectively identify design flaws or suggest libraries and patterns I wasn't aware of. But the important thing is not blindly accepting AI suggestions—it's filtering whether they truly fit our project's context.

To optimize development speed, I leverage AI as much as possible. Repetitive code writing, test case generation, documentation—I boost efficiency with AI assistance. But again, the key isn't using AI-generated code as-is; it's the process of modifying and optimizing it to match our project's coding conventions and architecture patterns.

Finally, I review any questionable points about architecture or language theory and philosophy, reducing complexity to make it easy for team members to understand. No matter how technically brilliant a solution is, if team members can't understand it, it becomes poison in the long run. So I invest a lot of time in expressing complex concepts simply and creating clean interfaces through appropriate abstraction.

This entire process flows naturally and unconsciously without any blocks, which I believe demonstrates automated execution capability and proves the efficient work capacity that humans can currently achieve. At first, each step required conscious deliberation, but now the process seems to operate almost unconsciously.

Additionally, I believe that frequent exposure to influential and highly resonant articles significantly improves intuitive judgment in the right direction. For example, when encountering a new technology or library, I can quickly determine whether it's just a passing trend or a genuinely valuable tool. I think this is because I've indirectly experienced countless success and failure cases through tech articles.

Because I repeatedly review the latest technologies and move toward applying them stably, this is different from simple standardization. In particular, I hold the belief that technology-related standardization leads closer to becoming legacy.

What I mean by simple standardization is, for example, stubbornly clinging to once-popular technologies like jQuery. Of course, such technologies can still be useful in certain situations, but in most cases, better alternatives exist. So I always strive to find balance and stability between the latest trends and legacy technologies.

The Need for Review and Knowledge Accumulation

If we think of article information as food, while nutrition is absorbed during digestion, beyond a certain amount, most of it passes through without being absorbed. In reality, even when you read a lot of information, most of it fades from memory over time. Even the absorbed nutrients disappear within a day or two, so I decided I should record the articles that left meaningful nutritional value, compare them with my own thoughts, and write posts that could share those insights.

In practice, these days when I read an article that leaves an impression, I try to at least leave a brief note. I record the core insights from the article, how they can be applied to my situation, and how they differ from what I already knew. As these records accumulate, they form my own knowledge base that I can quickly reference when similar problems arise later.

What you can trust is visible results and recorded history. No matter how much you've learned, if it's not documented, it's hard to verify. That's why I've been trying to organize what I learn on my blog or Notion, and whenever possible, I try to also document the experience of applying it to actual projects.

Feelings of Helplessness and Skepticism, and Attempts to Overcome Them

It wasn't all advantages. Continuous exposure to articles about a rapidly changing world leads to significant anxiety about living in such turbulent times. This can become the root of negative emotions and thoughts like helplessness and skepticism, which sometimes made things really hard for me.

Being someone with a vivid imagination, I sometimes felt a vague anxiety about whether a fragile human like me could solve increasingly fewer problems in the exponentially growing AI world, along with helplessness and skepticism about a reality where I didn't know how to respond. Especially as AI became capable of coding, I often wondered, "Will the profession of developer even exist in a few years?"

The impact was bigger than I expected. When I encountered myself becoming cynical and pessimistic, I sometimes felt truly weak. Realizing that someone who always considered himself positive and driven was being intimidated by the pace of technological advancement was genuinely disconcerting.

However, since this wasn't benefiting my life, I asked the very AI that gave me anxiety for solutions to overcome it. It's ironic to ask the source of your anxiety how to resolve it, but I think there's no more unbiased advisor than AI for getting objective counsel.

Instead of spending time worrying, immersing myself in something and achieving a sense of accomplishment was enormously helpful. Reading books by authoritative figures who promote a positive vision of the future also helped greatly. In fact, the sense of accomplishment from being absorbed in development—completing a new feature or solving a complex bug—overwhelms vague anxiety. As these experiences accumulate, I gained the conviction that "there's still so much that humans need to do."

The book I'm currently reading is "Ray Kurzweil - The Singularity Is Nearer," which I'm really enjoying. The author envisions a positive future, which gives me a sense of peace. Kurzweil views technological advancement not as something to fear but as a tool for expanding human capabilities. This perspective has greatly alleviated my anxiety. I plan to write a review of this book later.

Being Half a Step Ahead on Information and Its Practical Application

Being able to acquire technology-related information slightly faster than others does give you certain advantages. I invest small amounts in stocks, and being half a step ahead on information has occasionally yielded small profits when I combine various trends in my investment decisions. Though so far I haven't suffered major losses, I invest so conservatively that it's honestly not much different from anyone else, so it's hard to call it a clear advantage.

For example, by understanding AI technology trends first, I can try to predict movements in related stocks, or anticipate price changes in companies when a new or emerging technology starts gaining attention. Of course, this isn't always accurate, and since investment is determined by multiple complex factors, there are limits to relying solely on technical information.

But I think there's value in at least being able to approach investment with some reasoning rather than completely blind. I'll explain this in more detail in a stock-related post once I've accumulated more data.

Conclusion

What I gained from 2,000 articles wasn't just information—it was a transformation in the way I think. Even in a rapidly changing environment, I was able to develop the ability to identify the essence and flexibly apply existing knowledge to new situations. I want to continue this learning approach while building deeper technical expertise.