At the same time as what I am doing now, it is difficult to discover genuinely valuable problems when I approach things by thinking about technology first or by making technology for technology’s sake. Let’s use a hammer as an example.
If I start with the purpose of solving a problem around me, I first think about where the hammer will be used. Because there are concrete uses, such as breaking down a door or driving a nail into a wall, it is relatively easy to make a hammer suited to that purpose.
But when I think about technology first, the approach changes. I begin with an engineering understanding of the hammer’s materials and mechanics, its structure and shape, and then think about how to improve the hammer head and make it more efficient. In this process, it is difficult to look at the hammer’s necessity and the user’s actual use from a broader perspective. Instead, I can become an expert who cuts the hammer’s steel better than anyone and precisely shapes its wooden handle. And those skills are also easy to sell to producers who want to use hammers for many purposes.
But there is a problem.
To make a hammer better suited to a producer’s use, imagination alone is not enough. I have to become a producer myself. And to become a good producer, I ultimately have to start by solving problems around me. In other words, my desire to become a technical expert and build infrastructure with a technical moat can remain nothing more than an attraction to technology that looks fun and impressive unless I become a producer myself.
Of course, technical problems can also be found while shaping a hammer head. Even without considering the hammer’s overall use, it is possible to create a solution that optimizes a specific problem discovered while cutting the steel. But until now, I had the vague goal and desire to dig into technology first and turn it into infrastructure.
Where Does Deep-Tech Infrastructure Begin? After researching countless cases, I found that many deep-tech infrastructure founders did not try to build the underlying technology itself from the beginning. They first made a hammer and sold it to consumers, or discovered problems while becoming producers themselves. They approached deep tech based on the technical bottlenecks and questions they encountered along the way. Even the future insight that an era would come when a dedicated GPU for Transformers would be needed, as in Etched, can ultimately be understood as a technical problem and vision of the future discovered while cutting the steel of a hammer head. In other words, a grand technical vision does not necessarily begin with technology in the abstract. It can begin with a problem discovered while making and using something real.
Technology Learned at School and Technology That Creates Real Value School is a place where we learn the technical elements that make up a hammer deeply and systematically, rather than thinking first about what the hammer is for. We learn the material science of the hammer and study academically and deeply how to shape the hammer head precisely and make the handle well. This is because entrepreneurs and capitalists who define the hammer’s use or build the pipeline that produces it do not need to understand every technical element themselves. They can hire people with deep expertise in each component and technology. In this structure, schools focus on training experts who can understand and improve specific technologies deeply. That is why it is difficult to create something new and materially valuable from what is learned at school alone. In the end, I have to think about and experience for myself where technology is actually needed, how it is used, and what problems it can solve.
The Conclusion from Two Weeks of Exploring Technology For the past two weeks, I explored by setting the technology I wanted to work on as an anchor first, then looking for its use or making technology for technology’s sake. I chose a technology I liked first and then thought about what problems could be solved with it. But looking back, this was an approach that made it difficult to discover good ideas. I have always liked technical challenges that other people cannot easily do, like black magic or a feat of strength. But during those two weeks, by using technology as the anchor, I was thinking about doing a feat of strength for the sake of doing one. Even though there was a faster and better way that did not require such a feat. That does not mean the past two weeks were meaningless. I actually think they were meaningful. I wanted to set a topic with technology as the anchor, and at the same time I had a vague curiosity that made me want to verify for myself why that approach did not work well. By exploring various cases and thinking deeply, I reached a conclusion that I could verify and accept for myself, rather than simply accepting someone else’s advice. That is the biggest benefit I gained from those two weeks.
A Shift in Thinking: Make What Is Needed Before Difficult Technology Once this question was resolved, my thinking framework began to change. Until then, I had been overly absorbed in the idea of deep tech.
I thought that because AI had greatly increased productivity, I should now take on more difficult and complex problems. I also carried the vague assumption that it was natural to spend six months going deep on one difficult technology. But wanting to do deep tech and build difficult technology does not mean I must challenge the difficult thing from the beginning. There is no need to study the structure of a hammer head or develop a better way to cut steel when I do not even know where the hammer will be used. Before that, I need to think carefully about what use and need exist, and whether a hammer is needed at all.
Then it is more important to quickly make hammers for different uses in two-week or one-month cycles and hear the reactions and feedback of real users. Only by making something, using it, and offering it to other people can I learn what is needed and what is lacking. If I discover a problem that is difficult to solve with existing technology, it will not be too late to study the structure of the hammer head or dig deeply into better ways to cut steel. Technology that starts this way has a clear problem to solve, a real use, and a clear reason to build technical expertise and a moat.
This is not about giving up on deep tech. It is about changing the order in which I reach it. Instead of choosing the technology first and forcing myself to find its use, I should make something that is actually needed, discover technical problems, and build deep expertise and infrastructure while solving them. I do not need to choose a grand technical vision or a difficult problem right now. I can make something small but useful quickly, and use that experience to discover the next problem.
Let’s think flexibly. Let’s find a faster way.