ChatGPT, the AI chatbot that gained popularity since its launch in November 2022, has recently come under scrutiny. Studies have shown that ChatGPT’s performance is degrading over time and the new version of the chatbot exhibits odd effects of AI model drift.
A joint study by scholars from Stanford University and the University of California, Berkeley compared ChatGPT’s premium version, GPT-4, to GPT-3.5 and found significant changes in response accuracy across various tasks. The study evaluated ChatGPT’s performance in solving mathematical problems, responding to personal queries,…
Read more on google