This in particular: "At this very moment, thousands of Uyghurs are put into concentration camps based on computer vision algorithms invented by this community, and nobody seems even remotely to care." Yep. Priorities people.
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Meh. The lack of focus on making tools and integrating findings is all that matters. Gotta get those "we beat the benchmark" papers reinterpreted as: Great! Integrate it into the toolset I use or I don't care. It's happening, but slowly.
-
what makes you believe this? from a holistic view, don’t we want a healthy, supportive community to advance the field and its people? i’m not sure if you intended to, but when i hear “all that matters is building tools” it feels a bit disingenuous
- Show replies
New conversation -
-
-
I imagine that several of the points are a problem in academia as a whole.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This phenomenon is not new. Boltzmann was a victim of professional harassment and after suffering from bouts of depression took his own life. One of the greatest minds in history gone because of the toxicity of the scientific community.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This is absolutely damning: "Tuning hyperparameters on the test set seem to be the standard practice nowadays. Papers that do not beat the current state-of-the-art method have a zero chance of getting accepted at a good conference."
-
"As a result, hyperparameters get tuned and subtle tricks implemented to observe a gain in performance where there isn't any." - so, if this is true, then the state of our field is pure garbage, with milestones being fake achievements to a large extent.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.