My hypothesis is that
there is a new opportunity for Extensible Software on the web. LLMs radically lower the cost of authoring extensions, and modern sandbox primitives lower the deployment cost and provide good security boundaries. We can build our app as a solid, accountable core, and allow users to safely extend it in many directions by having LLMs fill in the missing pieces.We can give our users super powers.
— [Jeremy Morrell](https://jeremymorrell.dev/blog/extensible-software-in-the-age-of-llms/), Extensible Software in the age of LLMs
Tags: [sandboxing](https://simonwillison.net/tags/sandboxing), [llms](https://simonwillison.net/tags/llms), [ai](https://simonwillison.net/tags/ai), [generative-ai](https://simonwillison.net/tags/generative-ai)
source & further reading
simonwillison.net — original article
Conceptual integrity and counting lines of code
Qwen 3.8 27B scores 52 on the Artificial Analysis Intelligence Index
Qwen 3.8 27B is excellent, but it defaults to wildly overthinking things