When we use words like "introspection", "hallucination", "understand", "discover", and so on when we're talking about LLMs, we make a dangerous mistake. LLMs have no consciousness, agency, nor self-awareness, and using such terms can make it seem like they do.

(Even "writing code" hits different than "generates code".)

This isn't a pro- or anti-AI comment, it's a truth vs. lying (perhaps to oneself) comment. How we (especially the sellers of trained models) talk about these statistical token generators affects how/when/if we use them and what we expect of them.

Agent-driven development looks fast.

But here’s the problem: code is getting cheap. Confidence isn’t.

I’m seeing teams ship more, test more, generate more — and still lose control over architecture, quality, and long-term maintainability.

This is not a tooling problem. It’s a standards problem.

https://www.the-main-thread.com/p/agentic-sdlc-java-teams-trust-quality-standards

#Java #SoftwareEngineering #AI #AgenticAI #Quarkus #DevOps