Adam Clancy
banner
clancylab.bsky.social
Adam Clancy
@clancylab.bsky.social
Lecturer in Materials Chemistry, Royal Society URF at UCL Chemistry, & Daily Star-Certified "Boffin". Investigating nanomaterial chemistry for energy applications/fun.

Website: www.homepages.ucl.ac.uk/~ucqscla
ORCID: orcid.org/0000-0002-1791-8999
The reply was swift. At 52 seconds, it was the quickest of the day.

"Good question — and you were right to be puzzled. Short answer, my first reply used a wrong assumption"

Thanks ChatGPT. That was basic trig. Cant wait till you do that wrong for someone trusting who cant see you screwed up
November 19, 2025 at 12:47 PM
"The two solutions are
1. Physically reasonable: C-H = 1.045A and Br-C-H = 111.326
2. Mathematical (large, probably non-physical) root:
C-H = 3.023A and Br-C-H = 56.553"

That looked better, and could even be true, but that wasnt what it had said before. "Why is the answer now different?" he typed
November 19, 2025 at 12:47 PM
72s later, the black box spoke back "No" it claimed, "There are two mathematical solutions. One is chemically reasonable, the other is an extraneous (very long bond) solution."

Maybe the experiment or my initial analysis was wrong, thought the scientist, before reading on...
November 19, 2025 at 12:47 PM
"Nice little geometry puzzle — done" said chatGPT after thinking for 1m54s, "Your bond distance is 0.8065A and the angle is 127.626" it claimed.

"Not bloody likely" thought the scientist who knew that a 0.8A C-H bond was way too short. "Is this the only solution?" he asked ChatGPT
November 19, 2025 at 12:47 PM
He had the C-Br bond length and the Br...Br and H...H bond distances and he knew he could just do trig to work out the rest. "But trig is boring" he sighed, and turned to the magic black box of chatGPT to save him having to dig out pen and paper. "Gimmi the C-H length and H-C-Br angle" he demanded
November 19, 2025 at 12:47 PM