It’s hard for me how he recognizes all the (social) issues with LLMs and still comes to the conclusion that he isn’t in part responsible for them.
It’s hard for me how he recognizes all the (social) issues with LLMs and still comes to the conclusion that he isn’t in part responsible for them.
It’s not possible to avoid llm use in volunteer open source development if you prioritize software quality and security.
Whether we want it or not, people looking for bugs are using llms. They are finding phenomenal numbers of bugs. Implementing the fixes has created a huge workload.
Whether we want it or not, the Wayland people have decided that a new, piecemeal system for graphical desktops that relies on many disparate implementations of protocols is what we will use going forward. This has created a tremendous workload because there are now many different channels through which bugs and updates must be handled.
In this situation a volunteer development community must rely on every work saving tool they can just to keep up. For the person in the trenches there is no alternative.
If you prioritize quality as measured by the extent to which your software supports the various Wayland doohickies or security as measured by active exploitable bugs, you have to use llms.
These conditions will eventually remove the structure of incentives for open source development.
None of the above is a value judgement.