How people trust this software for decision-making is beyond me…
https://chat.openai.com/share/534ac2bd-332c-43d5-a26c-4b1bd8050c38
https://chat.openai.com/share/534ac2bd-332c-43d5-a26c-4b1bd8050c38
@axbom Have you heard of the guy that convinced ChatGPT that 2+2=5? Took them some tries, but eventually they did it.
It's a correct assumption that ChatGPT is inherently flawed, but I guess that it is more related to our flaw for destruction.
Follow
@mctwist @axbom more such examples in this paper:
"GPT-4 Can't Reason"
https://arxiv.org/abs/2308.03762
