ChatGPT continues to knot the brains of teachers around the world. Last warning shot to date, the test carried out by a professor from the prestigious Wharton University, in the United States.
Christian Terwiesch gave the OpenAI chatbot a review of his operations management course, part of the world’s most popular MBA curriculum. Result: marks from B- to B, that is, from 10.5 to 12 out of 20 for the French equivalent.
A very solid grade for the expected level and one that exceeds many students. “Our entire approach in the field of education is challenged by this, and it will only become even more difficult. It is time to rethink it from top to bottom,” Christian Terwiesch alarmed in a university note.
picking up for math
Last week, ChatGPT had already achieved the minimum score to pass the United States Medical Licensing Exam (USMLE), the three-part final exam that allows you to practice medicine in the United States.
But these feats shouldn’t hide ChatGPT’s difficulties in many subjects, including math. “I was blown away by the beauty of the writing: conciseness, choice of words, structure. It was absolutely brilliant,” Christian Terwiesch told the financial times. “But the math was horrible. Language and intuition are good, but he even messed up the relatively simple high school math.”
Still, ChatGPT has one strength: its ability to improve, which will make it much more effective in the future as it learns from its mistakes.
Source: BFM TV
