The Internet is roasting AI models' maths skills again. This time for not being able to count to 200.

The last time the Internet made fun of AI’s trouble with math was when someone discovered none of the models could count the numbers of “r’s” in the word “strawberry.”

OpenAI and Anthropic responded with a public relations campaign to convince everyone their respective models won gold at the Math Olympiad.

I wonder what they’ll cook up for Count-Gate.