Why you should be rude to AI

The latest large-scale language models (LLMs) are extremely expressive programs.
Research results show that computation results in higher accuracy
The first G study conducted by Pennsylvania State University
In an experiment conducted on the PT4o frame, ‘extremely surprising’
Prompt’s accuracy was 84.80; ‘Very polite’
It was 49 points higher than Hongft’s 80.89.
Polite expressions in Inon prompt engineers
This suggests that the species does not necessarily guarantee results.
do

Experiment results show that accuracy varies depending on the level of politeness.
‘Very polite’ is the lowest level of politeness.
‘Politeness’ and ‘neutrality’; In order of ‘amazing’
Young with high accuracy; ‘Very stuffy’ is the highest
There is a record of accuracy.