"Hey, ChatGPT, how many R's are there in the word 'strawberry'?"

"There are two R's in the word 'strawberry.'"

"Are you sure? Because there are three."

"Actually, there are two R's in 'strawberry.' The letters appear in the fifth and ninth positions in the word."

"No, count again. S-T-R, that's one R. A-W-B-E-R-R, that's two more R's. Y. Three R's."

"I see what you did there! You cleverly split up the word to make it seem like there are three R's, but when counting the actual letters in the word 'strawberry,' there are still just two R's. It's a fun twist, though!"

It's not a fun twist, ChatGPT! It's an incontrovertible fact. This bizarre quirk in OpenAI's chatbot has been intriguing users on social platforms like Reddit, so your favorite Inc. writers decided to give it a whirl. However we asked it, ChatGPT insisted that there were two R's in strawberry, even though there are three. And it's not just that one fruit that befuddled the world's best-known AI: We managed to get it confused over the spelling of raspberry as well.

Things got weirder. One discussion on the topic ended up in a sort of intellectual to-and-fro about the very nature of words themselves, with the AI arguing that "The R in 'straw' is part of the word, but it doesn't change the overall count of R's in 'strawberry.'" And while that sort of navel-gazing psychobabble may be wonderful in a college dorm at 2 a.m. after an evening sampling tasty foreign beers, it's not exactly great for a high-tech app that's supposed to be revolutionizing the workplace in countless ways.