IT House News on August 4, Kurtis Beavers, director of Microsoft's Copilot design team, published a blog post today, calling on users to use polite language as much as possible when chatting with AI, thereby helping Microsoft generate better content.
Beavers pointed out that artificial intelligence mainly relies on large language models (LLM), which are trained through human dialogue content. The LLM is similar to a huge prediction machine that can predict what may happen next with a high degree of probability
IT House pointed out that the communication data between humans and AI will be reflected in the AI chat content
When users use polite words such as please, thank you, etc., the AI training will also become more polite, considerate and helpful; however, if the user uses rude or provocative language, we may get some rude comments. Respond, and even trigger some unnecessary behaviors
The above is the detailed content of Microsoft Copilot team leader urges users to use appropriate polite language when communicating with AI. For more information, please follow other related articles on the PHP Chinese website!