« first day (4804 days earlier)      last day (227 days later) » 

1:32 AM
@SextusEmpiricus I want to have a statement that articulates the hesitancy to blindly rely on any statistical content from ChatGPT and subsequently basing your post on that. A cautionary tone, to be precise.
 
 
10 hours later…
11:05 AM
@User1865345 I think that would be a very good addition to precisely that Meta thread. Better to have comment templates like that in one place, rather than spread over multiple Meta threads. Do you want to take a swing at it? (Just please write about "ChatGPT and similar Large Language Models", because who knows what the main LLM in five years will be.) Thanks!
 
11:30 AM
The joys of working as one of few statisticians in a statistics-affine software company: someone has a task, reads Wikipedia, thinks "PCA sounds right, I can make this work", later leaves the company, yet later the customer asks questions, and the statistician learns what the software does. No, PCA was not the right tool for the job.
I think I will get a t-shirt made: "Please don't learn statistics from Wikipedia." Below that, in smaller print: "Would you learn brain surgery from Wikipedia?"
 
 
2 hours later…
1:31 PM
@StephanKolassa okay. I will do that at night.
@StephanKolassa I will buy that.
 
 
2 hours later…
3:07 PM
Another example of overhyped hnq:
5
Q: Conditional variance notation

RomanWhy does the formula ${Var} (Y|X)= {E} {\Big (}{\big (}Y- {E} (Y\mid X){\big )}^{2}\mid X{\Big )}$ condition $EY$, but not $Y$, on $X$?

 

« first day (4804 days earlier)      last day (227 days later) »