Welcome to Wikimedia Commons, VPoundstone-WMF!

-- Wikimedia Commons Welcome (talk) 14:19, 30 August 2022 (UTC)Reply

API question

edit

Hi Virginia! At this recent discussion, I suggested that we might want to pursue an arrangement with Google Photos where they would make it easier to share photos from there directly to Commons, just as they currently do to Twitter/Facebook. Donald Trung smartly suggested that I reach out to the WMF to check whether the requisite APIs are in place for such a feature, and you are are listed as the point of contact for the API team, so I wanted to check with you about this issue. Would you be able to offer any insights about this, or to point me to anyone else who could? {{u|Sdkb}}talk 04:41, 15 May 2023 (UTC)Reply

Hello @Sdkb replying here late, but wanted to make sure you received my reply email to this question. VPoundstone-WMF (talk) 15:47, 22 May 2023 (UTC)Reply
@VPoundstone-WMF, yes, I did! Thank you for roping in @MPinchuk (WMF) from the partnerships team, and I look forward to discussing with her how best to proceed with this. {{u|Sdkb}}talk 16:14, 22 May 2023 (UTC)Reply
Hi @Sdkb, thanks for getting in touch & for this suggestion! I do indeed currently work on the Partnerships team, but many years ago, I worked in the Product department at WMF on (among other things) mobile and image uploads, so this topic is close to my heart  . One thing that's important to keep in mind before pursuing any opportunity to bring in a large number of new images into Commons is the impact it may have on the very small number of Commons admins/patrollers. The general public doesn't have much awareness of Commons and, when asked to "donate an image to [Commons/Wikimedia/Wikipedia]," tends to not understand what that means. I speak with some hard-earned experience, because when we introduced an easy way to upload images on mobile web, we inadvertently caused something that the Commons community still refers to as the Selfie-pocalypse. We did a lot of work to try to reduce the number of selfies (and copyrighted images, and other non-encyclopedic/harmful content that was uploaded to Commons as a result), but while we were able to reduce the % of problematic uploads, we were never able to get enough quality contributions to offset the burden we were putting on the Commons community members having to triage/moderate this content. I strongly suspect a "donate image to Wikipedia" feature on Google Photos would run into a similar problem...
That said, I think targeting more aware audiences (e.g., photographers on Flickr who already understand CC licensing) and/or audiences who are engaging in specific activities that indicate they'd be interested & willing to contribute high-quality content (e.g., people who add suggestions/corrections on Google Maps) could be a more productive approach at getting more Commons content & contributors without overloading our moderation community. This is an area we're doing some thinking on for the Future Audiences bucket of the upcoming fiscal year's Product roadmap. I'd be very glad to see you sign up to follow along with Future Audiences & give input on these (and other) ideas for off-platform contribution! MPinchuk (WMF) (talk) 22:46, 22 May 2023 (UTC)Reply
Thanks for those thoughts, @MPinchuk (WMF)! This is the first I'm hearing of the selfie-pocalypse, but it doesn't terribly surprise me — seems like kinda the Commons equivalent of the Article feedback tool on Wikipedia.
For Flickr photographers who understand CC, I'm not sure there's a lot of potential added value, as those who want to release their photos already have, and it's already quite easy for us to import them.
I could envision other possible ways to target more narrowly, though. For instance, with Google Photos, I could envision Google applying a bunch of filters before offering the share-to-Commons feature: Only photos uploaded from the user's phone's camera (not screenshots or direct uploads from their computer), only photos geotagged to a location where Wikipedia is seeking a photo, only photos where Google does not identify a large face. Admittedly, that makes the feature a lot more complicated, so who knows if they'd be willing.
Signed up to follow the work on future audiences!
Cheers, {{u|Sdkb}}talk 23:12, 22 May 2023 (UTC)Reply