[QGIS-Developer] Tightening AI submission policy?

[Even Rouault]

>
> So far QGIS (and gdal as insiration) are addressing category 1 only.
> This means
>
>    human engages with an LLM to figure out a problem
>    human comes to understand the code and larger issues
>    human writes a patch and commit message
>    human submits that to qgis
>
> is not a problem (not a category 1 problem).

I believe one difficulty is that the purely non-technical / 
non-interaction issues of LLM use will be considered as "political", and 
for good or bad reasons, we generally try to avoid bringing political 
topics on the table when we can because they're perceived as divisive / 
not-safe-for-work topics. The "when we can" is important here. My guts 
tell me that we'd need to be radical on this subject and middle ground 
does not work when we have to face a huge propaganda that is imposed to 
all layers of the society. That said, I've seen recently a call for 
tender from a French public organize where use of generative AI is 
prohibited. That's clearly something to encourage, but I'm not clear on 
the motivation of that restriction. I suspect it has to with the fact 
that generative AI use strongly relies on US companies, so I'm not sure 
how long such restriction if/when Europe gains sovereign AI capabilities.

We are a very social species. If a sufficient subgroup of us manages to 
take the stance that those techs are not acceptable (the list of reasons 
for which they are not is long. that should be easy), others will feel 
more comfortable expressing similar views.

Even

-- 
http://www.spatialys.com
My software is free, but my time generally not.
LLMs contribute to global warming and brain rot