https://phabricator.wikimedia.org/T414281#11547106 To quote:
> @Yaron_Koren <https://phabricator.wikimedia.org/p/Yaron_Koren/> The main > problem for communities is WMF supporting an agentic editing project in any > form. If you were to pursue this project on your own, people might disagree > with you; but if WMF is involved (which we would be implicitly just by > facilitating this proposed project), then the student participant would be > facing a much bigger controversy. (If you're not familiar with these > complexities, please see > https://www.wikidata.org/wiki/Q130713537#sitelinks-wikipedia for a > glimpse of the years of problems that LLMs have already created for the > Wikimedia communities.) While we would not be involved directly in that > community conversation, we cannot risk it for a new participant who is here > to learn. Above all, our priority is ensuring a productive and safe > learning environment for the participant. > I’m sorry that we cannot provide you with long discussions but we are a > very small team and are actively managing multiple other projects. This message (imo) makes it clear that the main problem with this task in particular was the presence of agentic AI. https://phabricator.wikimedia.org/T414881 > This is less about "AI" but more about the high effort and risks in > communicating with the Commons community. Such a project would indeed > need a few months of communication beforehand and long-term support > afterwards. Personally I don't think we should put a GSoC attendee > that doesn't know anything about Commons in such a situation. These are/were classification tasks that, for the most part, would help community patrollers (something that the community has previously shown interest in, cc ClueBotNG, Automoderator, ... I swear Commons was also using AI models in patrolling, but I can't find a link to it). There was no generative AI involved (which has been the cause of almost all previous flashpoints with the community). Given that we are still actively deploying classifier models like the revert-risk model in production on enwiki (which is to my knowledge even more against AI), I don't see why these features proposed here were necessarily undesirable, assuming the mentor (as opposed to the GSoC intern/attendee) could reach out to the community and spend a large portion of the time between now and March 16th socializing the changes/gaining the requisite consensus. The way I see it, Pppery's note was a warning/gentle nudge to gain consensus before deployment, not a reason for rejection. Regards, Sohom Datta --- Open-source contributor @Wikimedia On Tue, Feb 3, 2026 at 9:02 AM Thiemo Kreuz via Wikitech-l < [email protected]> wrote: > I feel very sorry for the experience you had. Still the way this mail > is phrased leaves me puzzled. The explanations I can find for each > project sound very different to me: > > [2] https://phabricator.wikimedia.org/T414281#11527150 > Quote: "This work is very complex and too large for the scope of the > GSoC program. Such a large goal would not be setting a contributor up > for success […] We encourage you to consider revisiting this idea with > a smaller project or related work that would be more suitable for a > limited time frame and a less experienced developer." > > This sounds reasonable to me and has nothing to do with "AI". > > [4] https://phabricator.wikimedia.org/T415465#11563889 > "[…] we appreciate that you have taken a different approach on this > proposal, but we have a long list of accepted projects already > (including your T414617) […] we would strongly prefer to avoid having > mentors assigned to more than one project. We will make be sure to > include the guidance of "one project per mentor" in our documentation > in the future." > > Again, not a word about "AI". It sounds like there are indeed problems > with the documentation. But it's said they will fix this. > > Impossible to tell why you didn't get another response. Implying that > "the real reason" must be something else then is not helpful. Please > assume good faith and just ask. > > [3] https://phabricator.wikimedia.org/T414881#11550204 > "[…] the use of AI is controversial in the Wikimedia communities. We > do not have the capacity to sufficiently consult the communities on > this project and we cannot risk putting a student participant in front > of a contentious discussion. Our priority is ensuring a productive and > safe learning environment for the participant." > > This is less about "AI" but more about the high effort and risks in > communicating with the Commons community. Such a project would indeed > need a few months of communication beforehand and long-term support > afterwards. Personally I don't think we should put a GSoC attendee > that doesn't know anything about Commons in such a situation. > > I hope this helps. > > Thiemo > _______________________________________________ > Wikitech-l mailing list -- [email protected] > To unsubscribe send an email to [email protected] > https://lists.wikimedia.org/postorius/lists/wikitech-l.lists.wikimedia.org/
_______________________________________________ Wikitech-l mailing list -- [email protected] To unsubscribe send an email to [email protected] https://lists.wikimedia.org/postorius/lists/wikitech-l.lists.wikimedia.org/
