Used to, but then no one wanted to federate with them because it made moderation harder, so they got rid of it.
Honestly, if my career didn't rely on it, I would entirely skip the publishing part of science. It takes so much more work to communicate your findings than to make them. Of course, that doesn't help the cause since no one else will be able to build off of my work if I don't share it.
Depends if you count undergrad. One that comes to mind is the RWKV paper.
I don't need no bias. I need it to be biased towards the views of my fellow Canadians. I don't care what our neighbours down south think of the state of our home.
purpose-built ML models that are amazing and useful. And almost no one [...] is using them
They are in fact everywhere. We just don't advertise their presence because no one cares. This stuff also goes by different names in different fields because people keep reinventing the wheel.
Starlight tour, but for spiders.
I've mostly heard it used in the sense that there exists a million ways to accomplish this and your way is as good as any other, or to convey the validity of personal preferences.
It applies to physical goods too. Some companies out there will build a reputation of making high quality products for their price, then coast off that reputation while dropping quality and cutting costs.
Truth teller: "He'll point you towards the door that leads to certain death"
AI is such a broad term that I'm sure they have something that qualifies.
Depending on the gym, some are a lot more third-spacey than others. I've been to a smaller gym where people just hang around after their workouts to socialize, with occasional impromptu dinner outings when the gym closed for the night. I miss that place. You still meet people at bigger commercial gyms, but it's not the same.
Context switches