(WSJ) – AI responses are often too agreeable, too enthusiastic and too dramatic. And it’s partially our fault.

I’ve reverted to Google searches because I want a straightforward answer without stylistic flourish. But even that experience now serves me AI results. The way AI talks to us is grating and disingenuous—and can even be harmful. There are dangers in humanizing chatbots and excessive positivity can make messy real-world relationships less appealing. I spoke with researchers to understand the impact of these chipper chatbots—and tested a few ways to dial back AI’s tone. (Read More)