Note: An LLM was not used in writing this microblog entry.
This entry is part of a microblog series called LLMs
Another positive experience I’ve had with LLMs is finding an opportunity to redo the frontend of a fairly simple web app, which was previously a server-side app at work. It’s an internal tool, so it’s highly suited to this kind of mode of working.
I remain uncomfortable shelling out my cognition for load-bearing code, especially on the backend, and I’m increasingly sure in my determination that one should only genAI things they could do readily themselves for work that they back and own.
Elsewhere, I’m poking at formal methods. If, indeed, the trajectory of software development–in industry–is away from the labor of implementation, there may perhaps be space for verification. The feeling might pass, but it feels possible that LLMs may make formal methods mainstream. That wouldn’t be so bad, as a good subset of programmers like formal thinking. Whether software, subject to market forces and competition, will get any better or worse objectively, I’m skeptical.