On Tue, Feb 24, 2026 at 01:54:38AM +0900, Simon Richter wrote: > No, I expect the people performing the simple tasks with AI assistance will > no longer gain the necessary understanding to graduate from there and take > on more complex tasks over time... > > ... I personally detest it, because it's taking > over precisely the fun aspects of coding and leaves me with the kind of > drudgework that I used to raise my consulting fees for....
There are many different ways that people are experimenting with using AI. Here's a concrete example of how some projects which feed into Debian are using AI/LLM's today, as opposed to the sort of hyothetical strawman use cases that have been thrown around in many an AI-related discussion: http://lore.kernel.org/r/[email protected] The TL;DR is that this is a 61 patch series that was mostly generated using an LLM, but which was then checked and verified by a senior kernel developer. The patch series converted the type of the inode number found in the inode structure from unsighed long to u64. The changes are composed mostly of changing the type in function parameters, temporary variables on the stack, printf format strings, trace points, etc. This could have been done using a series of global search and replace, plus trial builds, and then fixing all of the build warnings that would come up. But it would have taken a human more wall clock time, and the work done by the LLM could fairly be categorized as being "drudgery". I certainly don't consider an example of the "fun aspects of coding"! It's also the sort of thing which would be very hard for someone to detect that an LLM was involved unless the developer discloses that fact (which is the Linux kernel policy). A 61 patch series to the Linux kernel is certainly not "trivial". But at the same time, in terms of how AI is being used, at least in this example, it's simultaneously much less impressive than what some of the AI CEO's have been putting out in their marketing materials, and perhaps, less threatening than how some of the folks who fear AI have been promulgating. Cheers, - Ted
