Ransalu Senanayake

LLMs Add Safety Risks To Physical AI

Robotics and artificial intelligence, or AI, advances are expanding possibilities to boost the beneficial capabilities of science and engineering and the technologies they help to create. At the same time, experts warn that ways are needed to prevent accidents caused by  machines equipped with AI from developing built-in biases, especially if those biases impact humans negatively. Those challenges are being addressed by Ransalu Senanayake (pictured), an assistant professor in the School of Computing and Augmented Intelligence, part of the Fulton Schools, and director of ASU’s Laboratory for Learning Evaluation and Naturalization of Systems. He says such smart technologies can be vulnerable to a kind of hallucination we must learn to guard ourselves against.