Why AI makers don't tell their chatbots to answer only what they know

Tech Insider
The logo for Anthropic displayed on a smartphone screen and binary code displayed on a laptop screen behind it.
Anthropic cofounder Jared Kaplan said that AI chatbots that are too "worried" about accuracy can become impractical for users.

Jakub Porzycki/NurPhoto via Getty Images

4 prompts to get your chatbot to fact-check itself and not make things up

Tech Insider
Sundar Pichai
Google launched Bard, its AI chatbot, after ChatGPT went live in November.

REUTERS/Stephen Lam

Some traders say ChatGPT gives them out-of-date information and they waste time fact-checking

Tech Insider
Traders

(Photo by Johannes EISELE / AFP) (Photo by JOHANNES EISELE/AFP via Getty Images)

Subscribe to Fact-checking