Ai Study Evaluates Gpt-3 Using Cognitive Psychology

Researchers from the Max Planck Institute for Biological Cybernetics recently published a study in the Proceedings of the National Academy of Sciences of the United States of America (PNAS) analyzing the general intelligence of large language model (LLM) GPT-3. The study found that GPT-3 solved vignette-based tasks similarly or better than human subjects and it made decent decisions from descriptions. AI is making daily headlines due to the wildly popular conversational chatbot ChatGPT by OpenAI, which reached 1 million users in only five days.