"As it does every time it releases a minor update, the company hailed its latest version as a "significant improvement in general intelligence," calling it the "best model yet for real-world, professional use." In a further display of hubris, OpenAI even went so far as to claim 5.2 is their "first model that performs at or above a human expert level.""
"That version thought for six seconds, and spat out an image with 25 letters, as opposed to the standard English alphabet which uses 26. Already off to a rocky start, 5.1 does an okay job with "A" through "I," but falls off the rails as soon as it gets to "K," which it says "is for Lion." It goes on like this: "O is for jellyfish," while "Q is for penguin" and "R is for snake.""
OpenAI released ChatGPT 5.2 on December 11 and described it as a significant improvement and their first model at or above human expert level. A simple test prompt to generate an alphabetized animal chart exposed major errors. ChatGPT 5.1 produced an image with 25 letters and numerous incorrect letter-animal mappings. ChatGPT 5.2 performed marginally better but still identified only 24 letters, omitting "U" and mislabeling several others. Specific mislabels included K as Lion, O as jellyfish, Q as penguin, R as snake, and Z rendered as "urba" with a turtle. Observers noted these failures alongside OpenAI's planned massive hardware spending.
Read at Futurism
Unable to calculate read time
Collection
[
|
...
]