Post

Simon Willison
@simonwillison.net

At the risk of starting the flame war to end all flame wars... Modern LLMs (GPT-5.1, Claude 4.5, Gemini 3) produce excellent code and can be a significant productivity boost to software engineers who take the time to learn how to effectively apply them - especially if used with coding agent tools

Nick Williams
@wicky.nillia.ms

I was very skeptical of their utility for coding beyond the "fancier autocomplete' provided by early copilot. But the last 9 months or so I challenged myself to be an informed hater and earnestly started using LLMs/agents for more tasks

Nick Williams
@wicky.nillia.ms

I have to admit I was wrong. They are useful for coding, and can be a real productivity booster. There's definitely an intuition you need to gain to know when/how to utilise them. I think that initial learning phase sours people. It definitely did for me

Simon Willison
@simonwillison.net

The initial learning phase was also a whole lot worse just 12 months ago - it took a lot more effort to get useful, productive results out of them prior to the coding agents that could actively execute and fix bugs in their own code

0 reposts12 likes

Nick Williams
@wicky.nillia.ms

Yeah I had tried a few times. And had very mixed success even in self contained tasks. Now I've managed to wrangle it for valuable product work, versus say the odd bash script or custom ESLint rule (they are valuable too! But they're just ancillary tasks)