In recent years, AI hallucination has been criticized for its role in chatbots, but surprisingly, it has proven to have a unique value in scientific research. A team of researchers at the University of Washington, led by David Baker, has leveraged AI hallucination to design new proteins, resulting in a $1.5 million Nobel Prize in Chemistry in 2023. Their laboratory has designed over 10 million new proteins that do not exist in nature, with approximately 100 patents granted, including applications for cancer treatment and other medical uses.
Meanwhile, a team at California Institute of Technology, led by Anima Anandkumar, has successfully employed AI hallucination to create a new type of tubing that significantly reduces bacterial contamination. DeepMind's Science Director, Pushmeet Kohli, notes that AI has demonstrated remarkable creative potential, enabling scientists to explore novel research avenues.
Scientists emphasize that AI hallucination in the scientific community is rooted in natural and scientific facts, distinct from chatbot-generated hallucinations based on ambiguous internet data. They firmly believe that AI-driven creativity will continue to pave the way for groundbreaking discoveries in fields such as energy harvesting and disease treatment.
Read more at: [URL to New York Times article]