A series of tweets from various users have reported unusual behavior from ChatGPT, indicating a widespread issue affecting the AI's responses. Users have described the AI's outputs as ranging from nonsensical and erratic to unexpectedly switching languages, including speaking Spanish, and metaphorically 'casting spells' on people. The phenomenon has sparked a mixture of amusement and concern among the user base, with some highlighting the potential for both humorous and unsettling interactions. The situation has led to discussions about the reliability and unpredictability of AI, as well as the importance of human-AI collaboration in developing robust systems. Notably, a case study was mentioned where ChatGPT, in collaboration with humans, developed a Monte Carlo simulation without any human-written code, showcasing the AI's potential when properly managed. The incident has attracted attention from both users and observers, with some calling for a closer examination of the AI's current state and its implications for future use.