generated code by unethical LLM is a tangible and good reason, better than no be tracked on the origin because at the end you can circunvent the issue with another tools.

There is no better option than ublock origin when it comes to adblocking extensions. I genuinely believe this is a psyop at this point, likely purveyed by the likes of Google (Look at them pointing out that Searxng is bad!). You will be full on releasing all of your data to them without anything to stymie the flow. Again, there are no other fucking options. Your options are, what, running very out of date software from pre-2019? This alone has glaring security risks.

What are you planning to do? Genuinely, this list is just a list of complaints. There are next to no real solutions on the list. In fact, the only solution provided by this list for browsing the internet, a basic task is to run BSD with fucking NetSurf because Linux is now out of the question. Netsurf is essentially deprecated, it hasn't been updated since 2023. It doesn't even containerize itself! It barely has tabs! It doesn't have adblock! To adblock you would have to do something like a pihole which checks notes would also use the same lists generated by LLMs from uBlock Origin. Even a hosts file on BSD would still require this. To securely run it, you would need to have BSD, find a non-ai VM software compatible with BSD (of which none are listed) and load NetSurf in the VM. Even then, your setup will be so ridiculously unusual that you would be instantly tracked by any and all services. You would also likely need to clear the VM each time you run it and avoid opening more than one window for NetSurf to avoid cross process contamination from how dead Netsurf's security development is. And this is ignoring that you need to follow the list to a T for the nebulous benefit of avoiding AI development, many programming languages and filesystems are dead in the water. I don't even know if its possible to assemble the BSD voltron necessary to be fully anti-AI, the most reasonable option is to make yourself a giant target by running very out of date programs from the pre-AI era and on old hardware.

It's also rich that this list is hosted on Codeberg which almost certainly hosts projects and filesystems that are not ordained by the list. And of course, you posted this on Lemmy, which is also considered to be slop.

This "ethics" is pure liberalism and idealism. You do not live in reality!

I'll grow suspicious about the use of LLM if Hurd is suddenly able to do anything practical this year

At least Haiku is actually daily drivable if your hardware is supported.

Just a quick warning, there's AI code in BSD too,
Examples from OPs link:
OpenBSD - https://github.com/openbsd/src/commit/9c2b8e445a0bdfafdd6148b1760f00aa5429627b
FreeBSD - https://github.com/freebsd/freebsd-src/commit/1644e2ffd2640fa3e2c191ceaf048a5fc8399493

TempleOS.

I was really wondering what's the future of coding as a craft given the impact of LLMs. My opinion on LLMs is irrelevant to that.

Artisanal woodworking still exist even if we have IKEA stores, but you don't go to them for the same reasons.

(PS : I don't like your comment neither and find it rather aggressive)