Re:LLMs are still surprisingly bad at some simple tasks
by ozgung on Sunday September 21, 2025 @09:43AM
(#45322670)
Correct. This actually falsifies OP's argument. Compared to OP's list from 2 years ago [1] ChatGPT omits ".search" but it says it's not a TLD anymore. GPT also finds 2 near misses, picture(s) and code(s). It does this in 10 minutes with 33 reasoning steps. It verifies them and provides citations in this time. Also checks OpenAI policy documents for some reason.
[1] https://shkspr.mobi/blog/2023/09/false-friends-html-elements...
Re:LLMs are still surprisingly bad at some simple tasks
by hobofan on Sunday September 21, 2025 @12:31PM
(#45324193)
> ChatGPT omits ".search" but it says it's not a TLD anymore
Not sure where you got that information from (can't find it in any of the 3 ChatGPT logs here), and I'm pretty sure that it's false.
It's still a part of the official IANA list[0] that it referenced in my chat log , and from what I can tell there has been no delisting of that TLD. (It's always been a niche Google-only TLD, though).
From all indications it doesn't pick up `search` because it doesn't recognize it as an HTML element.
Re:LLMs are still surprisingly bad at some simple tasks
by simianwords on Sunday September 21, 2025 @09:22AM
(#45322486)
https://chatgpt.com/s/t_68cffbc05ef48191996ffbaa3c6e55a7 same with non pro.
Re:LLMs are still surprisingly bad at some simple tasks
by hobofan on Sunday September 21, 2025 @09:28AM
(#45322533)
Non-pro: https://chatgpt.com/share/68cffb5c-fd14-8005-b175-ab77d1bf58...
It's consistently missing `search` for all of us.
Tried on ChatGPT, seems fine.