{"id":7053,"date":"2026-04-29T09:45:36","date_gmt":"2026-04-29T08:45:36","guid":{"rendered":"https:\/\/innovatenews.site\/index.php\/2026\/04\/29\/chatgpt-finally-knows-how-many-rs-are-in-strawberry-but-confident-mistakes\/"},"modified":"2026-04-29T09:45:37","modified_gmt":"2026-04-29T08:45:37","slug":"chatgpt-finally-knows-how-many-rs-are-in-strawberry-but-confident-mistakes","status":"publish","type":"post","link":"https:\/\/innovatenews.site\/index.php\/2026\/04\/29\/chatgpt-finally-knows-how-many-rs-are-in-strawberry-but-confident-mistakes\/","title":{"rendered":"ChatGPT finally knows how many \u2018R\u2019s are in \u2018strawberry,\u2019 but confident mistakes\u2026"},"content":{"rendered":"
\n

The topic ChatGPT finally knows how many \u2018R\u2019s are in \u2018strawberry,\u2019 but confident mistakes\u2026<\/strong> is currently the subject of lively discussion \u2014 readers and analysts are keeping a close eye on developments.<\/p>\n

This is taking place in a dynamic environment: companies\u2019 decisions and competitors\u2019 reactions can quickly change the picture.<\/p>\n

Confident mistakes \u2013 or lies, if you will \u2013 are a common problem of large language models used in AI chatbots, with one common shortcoming of ChatGPT being that it would frequently miscount the number of times the letter \u201cR\u201d appeared in the word \u201cstrawberry.\u201d As OpenAI tried to take a victory lap around this, though, plenty of other confident mistakes were pointed out in the replies.<\/p>\n

For as much as AI chatbots have improved, one of the biggest missteps remains the frequency at which these \u201ctools\u201d will confidently lie to you. If information is wrong, the chatbot won\u2019t notice and, if you call it out, the AI might dig in its heels on the response and continue to get it wrong, while telling you that it\u2019s right. It\u2019s a problem that is often shown as a danger of these tools, on top of being downright annoying given how many resources AI is taking up.<\/p>\n

One common example of this with OpenAI\u2019s ChatGPT is the question of how many times the letter \u201cR\u201d appears in the word \u201cstrawberry.\u201d<\/p>\n

For quite some time, asking ChatGPT about this would result in the chatbot coming back with the wrong answer, and it\u2019ll often argue that the word \u201cstrawberry\u201d does not use the letter \u201cR\u201d three times. Other AI models often ran into the same problem.<\/p>\n

Today, OpenAI took to Twitter\/X to proudly tout that, \u201cat long last,\u201d ChatGPT can correctly answer this question. Another common stumbling was the prompt \u201cI want to wash my car today but the car wash is only 50 meters away. Should I walk to drive there,\u201d to which ChatGPT would often recommend walking, despite the very obviously logical problem there.<\/p>\n

Sure enough, both of these are now working if you try them in ChatGPT, but it\u2019s suspected they might be hardcoded solutions. Many replies to OpenAI\u2019s post show other times where the chatbot fails on the same logic. for example, \u201cHow many r\u2019s are in cranberry\u201d repeatedly sees the chatbot continue to reply with \u201cThe word \u2018cranberry\u2019 has 1 \u2018R.'\u201d Of course, that\u2019s incorrect.<\/p>\n

Hardcoded solutions in AI chatbots aren\u2019t new, but it\u2019s a bit funny \u2013 in a dystopian kind of way \u2013 to see OpenAI touting this \u201cfix\u201d when, clearly, the root of the problem remains.<\/p>\n