Researchers are working on new AI systems that can check their own mathematical answers. This is reported by Cade Metz in the New York Times. The technology is meant to prevent chatbots like ChatGPT from providing incorrect information. Instead, the new systems can generate mathematical evidence to verify their answers.
Silicon Valley startup Harmonic, for example, is developing a chatbot called Aristotle that can solve problems using mathematical logic. Google DeepMind has also unveiled a system called AlphaProof, which won a silver medal at the International Mathematical Olympiad. Experts hope to extend these techniques to other areas, such as computer programming, in the future. The goal is to create AI systems that provide reliable information and no longer produce hallucinations.