Ant Stanley @iamstan.dev · 26/07/2026Even when an LLM writes 100% of the code there is a human behind it with a mental model of what they want the program to be. It that context, an LLM can tell you about the code base, but not the theory behind it. 100
Ant Stanley @iamstan.dev · 26/07/2026Also that mastodon post is by Ruth Malan, one of the leading software architects on the planet with more papers behind her than most of us. She's smart, considered, and not one to jump on bandwagons. 200
myydraal @myydraal.bsky.social · 27/07/2026I think the problem is thay assumes a singular vision in a piece of software. Which may be true in some pieces of software, but at places like Amazon the software is a sprawling sea of dependencies, each with their own vision and mutated states of vision overtime due to turn over that it really 140
Joe Fabisevich @mergesort.me · 26/07/2026Hey dude… you're missing the point I was making. I agree with Ruth about what an LLM can and cannot do, but the problem with this as a line of argument is that we are not blindly feeding LLMs codebases, and we can encode understanding into agentic workflows with increasing levels of fidelity. 050
