Innovation does not happen in a vacuum.

We have arrived at a point in time where Large Language Models (LLMs) can solve many challenges. They are capable of great feats when given the correct tools and context. I myself have enjoyed the benefits of their capabilities but I know it comes at a price.

I never loved the coding aspect of my job. I found it to be mundane when what I really enjoyed was the creativity that went into designing a system or product. The code was just needed to achieve the task. I viewed it in the same way I imagine a construction worker views a hammer or a carpenter views a saw. It was just the tool used to get the job done. Now if that tool has been fantastic and served its purpose I may be happy or even proud of it but usually the craftsmen are proud of their product and not focused so heavily on the tools used. I am happy to relinquish the task of writing code to LLMs. But doing so creates an issue and this issue is not just specific to coding. The issue is how innovation occurs.

Innovation does not happen in a vacuum.

Mathematical discoveries are named after the second person to discover them, because the first was always Euler.

That joke exists because Leonhard Euler kept finishing one problem and noticing the next one before anyone else had language for it.

Take the Basel problem. For years mathematicians wanted the exact sum of one over every square: one plus one over four plus one over nine plus one over sixteen, and so on. People could approximate it. Euler wanted the closed form. He attacked the series with tools that looked reckless by modern standards and got a clean answer: the sum is pi squared over six.

Euler solved Basel and then kept going. Once he could turn a sum over the whole numbers into something exact, he asked what happens if you raise the power: cubes, fourth powers, any power s. You get a function of s. That seed becomes the zeta function. Then he found something stranger still. Because every whole number factors uniquely into primes, the same sum can be rewritten as a product over primes alone. A neat series problem had become a map of the primes. A century later Riemann asked where that map hits zero, and we are still stuck on that question. One finished answer had revealed a much larger unfinished problem.

Passionate and curious people are what drives innovation. Read stories about what it was like working at Bell Labs and it sounds like every scientist's dream. Creative, intelligent, and curious scientists working together. Not just to create a new product but to define whole new spaces of opportunity. The work at Bell Labs has led to inventions such as the transistor, information theory, modern computer vision, Unix, C, the laser, the list goes on. The point is here that these were not employees simply hired to solve a task (or close a ticket). They were empowered to apply their mastery and define new areas. During this period of ideation, development, and discovery researchers will often learn about new topics, start connecting the dots, and then realize there is another problem out there but it is not the one they are focused on right now.

Innovation usually occurs when smart and passionate people solve a problem. For example, the internet was invented to solve a problem. The problem was that researchers at CERN had no efficient way of sharing information with colleagues at universities around the world. Tim Berners-Lee was frustrated about this problem and decided to do something about it. He decided to solve this problem and his solution is the reason you are reading this right now. Tim Berners-Lee could have remained frustrated about the challenges of collaboration but instead he decided to do something about it. He was passionate (whether from frustration or innate curiosity) about solving this issue and ended up revolutionizing not just sharing information but the world itself by inventing the internet.

That is the pattern that matters to me. Tim Berners-Lee stayed with a collaboration problem until the web appeared. Bell Labs gave curious people room to follow ideas into whole new fields. Euler finished one sum and walked straight into a map of the primes. In each case someone stayed with a real friction long enough that a second problem showed up, and that second problem is often what remade the field.

I am happy to hand coding work to LLMs. They are capable tools when you give them the right context. They close tickets, draft code, and can help chase a proof when you already know which question to ask. The harder part of progress is still human. Innovation depends on people who care enough about an open gap to keep looking after the first answer arrives. Tools help you execute. Curious people are the ones who notice that another problem is sitting right next to the one you just solved.

That is why Epilogue exists. These pages are stories of people who kept working a hard question past the first layer. If AI is going to take over more of the execution work, we still need people who stay curious enough to find the next real problem worth solving. Hopefully these stories can inspire the next generation of problems.