druide67

The finding about removing the 9.8 GB Metal LRU cache for a 38% speedup is the most interesting part. Same lesson as PostgreSQL's advice against application-level buffer pools that compete with the OS page cache : the hardware memory compressor doing 130K decompressions/sec was pure overhead.

Curious about the remaining gap: 5.7 tok/s vs 18.6 theoretical (from SSD bandwidth). Is the ~70% overhead mostly GPU compute on non-expert layers (attention, norm), or is there I/O scheduling room left?

Interesting tool. One thing I've noticed managing rules across Claude Code and Copilot: the same instruction produces very different results depending on the agent. Claude follows multi-step rules well, Copilot tends to ignore anything beyond the first line.

Does Skills Manager handle this at all, or is it purely format/distribution? Seems like the hard problem isn't syncing files — it's that the same "skill" needs different phrasing per agent to actually work.

This resonates. I spent years thinking I enjoyed coding, but what I actually enjoy is designing elegant solutions built on solid architecture. Inventing, innovating, building progressively on strong foundations. The real pleasure is the finished product (is it ever really finished though?) — seeing it's useful and makes people's lives easier, while knowing it's well-built technically. The user doesn't see that part, but we know.

With AI, by always planning first, pushing it to explore alternative technical approaches, making it explain its choices — the creative construction process gets easier. You stay the conductor. Refactoring, new features, testing — all facilitated. Add regular AI-driven audits to catch defects, and of course the expert eye that nothing replaces.

One thing that worries me though: how will junior devs build that expert eye if AI handles the grunt work? Learning through struggle is how most of us developed intuition. That's a real problem for the next generation.

I use Claude Code (CLI) daily for infrastructure work — Docker configs, shell scripts, API development, config management.

What works: I stay in the driver's seat. I own the architecture, make the decisions, validate everything. But I don't need a team to execute — Claude does the implementation. I went from being a solo dev limited by time to running a complex project (multi-agent system, Docker, Synology integration, PHP API) that would normally need 2-3 people.

The key is a good CLAUDE.md file with strict rules, and pushing Claude to think ahead and propose multiple options instead of just doing the first thing that comes to mind. Claude is also surprisingly powerful for audits — security audits, config audits, log analysis.

What doesn't work: it confidently generates plausible-looking code that's subtly wrong. Never trust it on things you can't verify. It also over-engineers everything if you don't rein it in.

The biggest shift: I went from "write code" to "review and direct code." Not sure it's making me a better engineer, but it's making me a more effective one. It extends me.