Sidechat logo
Download TikTok Instagram X
Sidechat icon
Join communities on Sidechat Download
US Politics icon
US Politics 1w
wrong board but fuck it. i’m in a group project rn with a guy who loves AI and he keeps doing tests on his own and getting results that contradict expected behavior and my data and everytime i explain why he’s wrong he just ignores it and idk what to do
comment

14
upvote 16 downvote
default user profile icon
Anonymous 1w

like I’ve thoroughly explained the technical reasoning on why his data is likely wrong and how it doesn’t make any sense, but he keeps just saying that it has high coherence, and trying to explain away the fact that his results contradict basically all published results on the thing we’re examining and that my data doesn’t. i don’t even know if this is something i bring up with the professor bc his data is demonstrably inaccurate and i can’t afford to lose marks for it

upvote 9 downvote
default user profile icon
Anonymous 1w

oh also here’s a figure from a published paper that does the same kind of testing for similar materials

post
upvote 6 downvote
default user profile icon
Anonymous 1w

and this is his graph

post
upvote 5 downvote
default user profile icon
Anonymous 1w

like this is my graph (with likely incorrect identification of 2 & 3 because testing log got lost, hence my going back to do more testing)

post
upvote 3 downvote
default user profile icon
Anonymous replying to -> OP 1w

email the professor about it politely.

upvote 14 downvote
default user profile icon
Anonymous replying to -> #1 1w

yeah that’s probably what I’ll do. I’m going to the lab to run tests tomorrow and if he insists that his data is correct after i process those then I’ll reach out to the professor

upvote 11 downvote
default user profile icon
Anonymous replying to -> OP 1w

Honestly might be worth going to the lab with him (I assume this is for some class where the tests need to be physically done, and he’s just asking ChatGPT what it predicts would happen)

upvote 11 downvote
default user profile icon
Anonymous replying to -> #2 1w

that’s not the case bc i’ve seen the .txt file exports from his testing, and they’re formatted exactly like actual data exports

upvote 10 downvote
default user profile icon
Anonymous replying to -> #2 1w

and also you can’t go to the lab with him bc he schedules his own testing time and doesn’t tell anyone else about it

upvote 1 downvote
default user profile icon
Anonymous replying to -> OP 1w

lol that’s so weird. And ngl if he found out the format of the export he could probably replicate it

upvote 1 downvote
default user profile icon
Anonymous replying to -> #2 1w

his data isn’t fabricated, it’s too noisy for that unless he’s spending more time fabricating data than it would take to do the actual tests, it’s likely an experiment setup issue

upvote 1 downvote
default user profile icon
Anonymous replying to -> OP 1w

ooooooh I see

upvote 1 downvote
default user profile icon
Anonymous replying to -> #2 1w

like if u look at the graphs i shared his results just make no fucking sense whatsoever

upvote 1 downvote
default user profile icon
Anonymous replying to -> #2 1w

like i am literally showing this mf ISO testing standards and he maintains that his data is meaningful and that Im doing the testing wrong

upvote 5 downvote