I always wonder what kind of businesses I’m unaware of because the founder chose to stay off the popular culture. It feels like most companies are either backed by VCs or publicly known. I don’t think I want to create a startup. I want to form a boring company that delivers. But I don’t know in which form or what...
The difference between spec-first and plan-to-implementation is the same difference between what to build and how to build. Initial analysis with AI and specs should focus on what to build, not how to build. In the engineering world, two things are really close. That's why product thinking is required in agentic...
AI produces the average of the internet. Which knowledge are you relying on? What is your information source? The AI code generation doesn’t follow team conventions or tacit knowledge unless these are explicitly stated in the AI agent’s context. Reply via email | Reply via Mastodon | Comment
SpaceX’s IPO is restricted to Chinese and Hong Kong investors. There are similar trends in chips and AI technology, too (the Fable 5 LLM model is restricted to the US by the US government because of its power). SpaceX is not the main thing. The problem stems from gaining access and influence over a country’s defence...
Software engineering as we used to know it is dead. AI condensed the code-writing step of the process from days to minutes. Engineers found themselves having many tasks on their plates, running many agents in parallel. They are burning out tokens and themselves at the same speed. However, I argue that the approach...
Many people are struggling to understand what they think about AI and how to approach it. Perhaps the bigger problem is that the technology is not fully deterministic but probabilistic. I quite like the framing of the idea Dan presents: AI's biggest difference is that it cannot care about anything. I certainly can...
Many people are struggling to understand what they think about AI and how to approach it. Perhaps the bigger problem is that the technology is not fully deterministic but probabilistic. I quite like the framing of the idea Dan presents: AI's biggest difference is that it cannot care about anything. I certainly can...
Recently, I learned what “disagree and commit” truly means, a concept everyone struggles with a lot. Committing to an idea that you disagree with is difficult. Or, let’s say it was, until today. The idea is that you may disagree with whatever life throws at you. You accept it as it is, but you have to commit to...
I never understood what people meant by “strategic thinking” or “being strategic.” I couldn’t accept the feedback when I heard “you need to think more strategically” as well. Nobody could explain what that “strategic” word meant. I finally understood the real meaning when I tried to build an engineering strategy...
Managers Returning to "Coding" with AI We have been talking with other engineering leaders around me about the impact of AI in our profession. Although nobody knows what will happen, I kept thinking about where I'm standing and what others are doing today. There is certainly a trend among managers to go back to...
The tighter the connection between business strategy and technical design, the easier for a software engineer to understand and implement. The more apparent this knowledge is, the easier for the AI to pick up and apply the principles and engineer to guide it correctly. Chapter 1. Analyzing Business Domains While...
Mediations is a newsletter I write for software engineers and leaders but I never focus solely on software engineering. This is a newsletter for humans trying to mediate the challenges life presents, so the topics are as wide as it gets. In the last few newsletters, I talked a lot about AI. While it's both bothering...
