I find it very dishonest that they design their donation banners in a way to paint themselves as people who are in a dire, needy situation when in reality they could run that high-traffic website for several more decades with their current capital. This article claims it could even be 75 years.
In case the rich managers will actually run down the website, I think another website like it would pop up really quick. The Wikipedia software is open-source and there are many other Wikis already. Creating another "general" instance seems trivial. People who do the work would not lose a permanent paid employment because they don't have one, they would just switch over to the next instance. Wikipedias data (and many other Wikis) are already being backed up by third parties. The knowledge will not be lost.
Oftentimes, doing a job for the "good will" really doesn't pay off. Being an admin or moderator can take a huge mental toll. That's one reason we are not on reddit anymore: They also don't pay their moderators while the CEO is a rich asshole. Another current example is the reason of the shutting down of the lemm.ee instance (explained in their stickied post). A lot of workers in non-digital jobs get abused in the same way: No or low salary but too much working hours. They, too, are expected that "doing something good" would somehow prevent the excessive mental and physical stress they experience. Examples are hospital and nursing staff and animal welfare and shelter staff.
I also think that it isn't 100% good or bad. AI can be helpful and supporting if you human-check the results. It can also be wrong, misleading and copyright infringing. Furthermore, forcing features onto users which they don't like is annoying, especially if their data gets abused for it.
Differentiated thinking is important for this topic.
It might be there because there is a lot of data associated with the steam account, especially the community part of it, e.g.:
- Recorded playtimes
- Times and dates when you are regularly logged in
- Possession of games which are precisely tagged by genre/interests/etc.
- On which time and date you spent how much money (participation in sales in the steam store)
- Timestamped posts and comments in groups based on various interests etc.
- Curators/devs/publishers you follow
- Your game wishlist
- Connection and interaction with other steam accounts (friends list, chat, trades, gifts)
All this can be used to create a very detailed behaviour profile and accurately deduce the social status of the real person who uses the account. Maybe the data isn't misused and it's just there so the features can actually exist.
Personally, I doubt Valve actually does this as expansive and invasive as other big tech companies. I'm pretty sure they at least aggregate anonymised data to measure how e.g. their sales perform, which game to promote on the store front page etc.
But we can't be sure because it's not public.