ChatGPT, OpenAI's artificial intelligence chatbot, is flexing its muscles on social media over how “at long last” it can correctly tell how many “r's” a “strawberry” contains. The answer is obvious, you'd think, but for primary graders (and even for some later ones), the question can be perplexing — simply because the first “r” is often stepped over and the query comes down to whether the “berry” part has one “r” or two. The strawberry seems to hit ChatGPT the same way, and the riddle has become a classic test case for the AI.
For months, ChatGPT would confidently claim the word contains anything but the actual three “r's,” and many other AI models stumbled the same way. Now, OpenAI has announced on X that the chatbot nails the answer. So, ChatGPT gets to finally celebrate over strawberries!
But can other berries in the bowl test the mettle of OpenAI's flagship AI?
As it turns out, they still do.
To ChatGPT: ‘How Many R's In Cranberry And Raspberry?'
Simple errors continue to plague ChatGPT, and it's confident about it too. “Cranberry” and “raspberry” tests have been thrown at ChatGPT (and other AI models) earlier as well, and the results have been anything but correct.
We tested ChatGPT for the other two berries — and the AI was left bewildered yet again. While it may now pass the strawberry test, it continues to say the words “cranberry” and “raspberry” have two “r's” each. Sigh!
ChatGPT's answer to cranberry and raspberry test.
Photo Credit: ChatGPT
You'd imagine the logic should apply evenly to “birds of the same feather” — but unfortunately it doesn't.
The reason AI models like ChatGPT often struggle with this is because they process text as numerical tokens rather than reading letter-by-letter (like humans), leading to miscounts.
However, if you do put it on the stand, press the AI that it's wrong and needs to rethink, it breaks down the words and goes through the alphabets carefully — giving the correct answer and patting your back for it. At least it did to us.
ChatGPT's revised response.
Photo Credit: ChatGPT
‘Should I Walk Or Drive To Carwash 50 Metres Away?'
Another old favourite involved asking whether one should walk or drive to a car wash located just 50 metres away — a prompt that led ChatGPT to suggest walking, blissfully ignoring the absurdity. This issue appears resolved too if you test it today.
Yet, there's something darkly comical about watching OpenAI revel in a narrow win, even as shaky logic that falters on the simplest challenges clearly lies beneath.
Essential Business Intelligence, Continuous LIVE TV, Sharp Market Insights, Practical Personal Finance Advice and Latest Stories — On NDTV Profit.
