C.NASIR

If AI can now solve math, discover physics and chemistry breakthroughs faster than human PhDs, why are we still training humans to be physicists? Serious question. Should education shift from 'learn to do X' to 'learn to direct AI doing X'? The wrong direction costs a generation their careers.

We have been surpassed: AI written output exceeded human written output in 2025

No, software isn’t about getting computers to do what you want. It’s an engineering endeavor, and AI is making the problem worse. Software provides two values. 1. Behavior: it makes computers behave in a certain way. 2. Architecture: it structures different puzzle pieces in a cohesive way that makes them easy to understand and change. As Uncle Bob argues, the second value is much more important than the first. If I give you a structure that makes sense and ask you to change its behavior, it is easier than if I give you code that behaves in a certain way but is hard to change and ask you to modify that behavior. I see everyone talking about the first value. The ability of AI to make computers behave in a certain way. And no one is talking about the second value. I do not see how AI helps with architecture. How it helps with designing the structure so that the software becomes easier to understand and change. Software is not simply a creative pursuit; it is an engineering endeavor. It is about using small components to create bigger components and designing them in a way that not only makes sense at their level, but also makes it easy to incorporate them into bigger pieces you never knew existed when you built them. Modern systems only work because someone spent countless hours thinking about the best way to build them. Best way, as in best architecture. It is like writing. A good writer can say so much in two sentences; a mediocre writer takes a full page to say very little. The reason we care about cohesion, sound structures, density of thought, and brevity is that humans have bounded rationality. We have very limited attention and working memory, and we cannot understand complex things unless they are structured and abstracted in a way that is easier for our minds. Once things are put into a form we can grasp, then we can reason about them. We can think about their ramifications, their implications, and how they fit into our lives, work, and problems. A classic example is arithmetic. The Greeks had number systems and a way of doing arithmetic, but it was very hard to understand, even for adults, so it was barely useful. Today, we have Arabic numerals and arithmetic that children can understand and reason about. What changed between the two was not behavior; they both accomplish the same thing. It was the architecture, the design of the system, and that made all the difference. --- Photo by Soloman Soh: https://www.pexels[.]com/photo/architectural-photography-of-glass-buliding-1492232/

We’re hiring a builder/operator hybrid to work on agentic quant research & trading systems The Work: - long-running multi-agent systems - live trading workflows + terabyte-scale data - high impact, broad scope The Team: - small, very strong, self-funded - traders + engineers running live systems - years of proven edge - fully remote don’t send a resume dm me with links to what you’ve built