Ed Kroc

And your quarterback ...

Professors having 50 papers per year is exactly the problem. That’s almost a paper a week. The system is broken because we end up with a few professors clogging all the resources and essentially overseeing paper mills.

Regularization techniques prevent machine learning models from overfitting—memorizing noise instead of learning patterns. Methods like L1 (Lasso) and L2 (Ridge) add a penalty to the loss function, shrinking weights to simplify the model. This ensures the AI generalizes well to new, unseen data. In real life, this robustness is critical for autonomous driving to ignore sensor noise and for medical AI to make reliable diagnoses across different hospitals.

OK am I nuts? I've heard people say Aella is a legit "data scientist" but I've read this Aella "data analysis" like 5 times and... It seems like the most incorrect data analysis I've ever seen. But Yudkowsky RT'd it with a glowing endorsement and he knows math, so I really need someone to pinch me right now. Like either I'm crazy, or Aella is so gaslit by her simps that all of their brains have gone mushy together. Can someone (who is not an Aella simp) please look real quick and pinch me? I tried asking her in X replies but that did not really help. Her analysis responds to a report by Eric Kaufmann, who presents repeated cross-sectional surveys of college students over a few years (not panel data, different cohorts). Kaufmann's data suggest nonbinary identification among university students increased from 2019 to 2023, then decreased or at least stopped increasing. Aella presents data from her own survey, which is treated as one point in time (submissions have come in over a few years but she pools it into one period for her analysis), in other words, there is no time dimension in her analysis. She presents a graph showing her measure of trans by age and declares, "I don't find a dip. Weird." Am I crazy or... Is she confusing Age with Time? Is that what she's doing?! Like 4 times in a row in multiple analyses? And she does nothing else? If so, this would be the most deeply naive mistake you could ever possibly find in a blog post from someone who identifies as a researcher. Like this would be legendary, the "chart crime" of the decade. If you don't know, there is a difference between AGE (lifecyle), TIME aka PERIOD (calendar years), and COHORT (birth year). Kaufmann graphs a kind of period change. This is what everyone means if they talk about change over time, something increasing or decreasing over time... Like tons of variables just go down over the lifecycle lol. That's all her graphs seem to show. You're more likely to be trans when you're 20 than when you're 70 lol. Duh? These graphs cannot show any decline (or increase), in anything over time, in any way even slightly comparable to the analysis she is attempting to question or challenge or correct. Like her graphs are infinitely structurally incapable of saying anything whatsoever about the object of her analysis. Unless, again, I'm crazy. After this post, she's tweeting how bad professional researchers are and how good she is compared to them. Eliezer Yudkowsky RT'd this post saying "one of Earth's top scientists on sex and gender has published her latest work." So can someone please someone tell me: Am I an idiot, or are Aella and Yudkowsky both idiots?