https://benn.substack.com/p/should-we-be-nice-to-chatgpt

Should we be nice to ChatGPT?

14 Jul 2023 Digital analytics 475

Benn Stancil completed an awesome and very detailed experiment with ChatGPT to find out if the responses the LLM provides vary depending on the kindness level of the requests!

Spoiler alert: Nice guys finish last!

When asked to do something in a neutral tone, ChatGPT correctly completed the task 73.4 percent of the time. When asked to do the same things in a nice or mean tone, it only completed the task correctly 70 percent of the time. This pattern held across different temperatures as well, with ChatGPT scoring the highest when the temperature was set to zero and the question was asked with a neutral tone.

Benn Stancil

I totally recommend his substack!

Panagiotis

Written By

Panagiotis (pronounced Panayotis) is a passionate G(r)eek with experience in digital analytics projects and website implementation. Fan of clear and effective processes, automation of tasks and problem-solving technical hacks. Hands-on experience with projects ranging from small to enterprise-level companies, starting from the communication with the customers and ending with the transformation of business requirements to the final deliverable.