COPTR: Add use of AI to infobox?

[ross-spencer]

Sorry for the spicy one this morning, but should we add use of AI, either by way of supporting connectors (thus supporting the ecosystem) or having being used to aid development using AI to the info box?

Being on the fence as much as possible, AI concerns have been around:

• Ethics,
• Environment,
• Sustainability,

And from a software development and preservation perspective it isn't clear what pathway to long-term availability and usability of code that AI provides.

Similar discussions

I haven't been collecting many critiques from the community, but one from FITS: harvard-lts/fits#421

And a recent one on the Code4Lib list that provides a good breakdown and useful discussion points: https://lists.clir.org/cgi-bin/wa?A2=ind2607&L=CODE4LIB&P=13529

Suggesion

Add the data to the infoboxes and basic entry input form. Maybe just a boolean for yes/no AI. But this could be split into different uses, e.g. AI for maintenance, AI connectivity, and so on.

I'm not 100% sure how the infobox works to find information but as a browser seeing this information would be helpful. If there are enhanced search opportunities e.g. via the wiki's semantic capabilities then that's also going to be helpful.

[anjackson]

Thanks Ross. This seems like a reasonable request, although I'm not sure how to define the flag.

It's a complicated time, and a blanket AI? yes/no feels very coarse. Not only because it conflates e.g. machine learning with large language models, but also because I'd like to cater for the case where a project goes AI and then decides it's no longer a good idea. I don't want say projects (or by implication, people) are forever tainted by having used AI in the past (admittedly, legally, this may be the case). Maybe a flag indicating whether a project currently has a no-AI policy? (no-LLM policy? no-agentic-development policy?!)

Also noting openpreserve/odf-validator#293 as a related discussion.