Language-Generating AI Could Revolutionize Science
These algorithms might revolutionize research, but not necessarily for the better, according to Shobita Parthasarathy, an expert in emerging technology governance at the University of Michigan in Ann Arbor. Parthasarathy and colleagues presented a paper on the social consequences of upcoming AI technologies dubbed big language models on April 27 . (LLMs). These can write compelling writing, translate languages, answer inquiries, and even code. Companies like Google, Facebook, and Microsoft are developing them to use in chatbots, search engines, and document summaries. For example, Ought in San Francisco is employing LLMs in research to construct a tool called ‘Elicit’ that uses scientific material to answer inquiries. LLMs are already divisive. They sometimes repeat inaccuracies or bad preconceptions from millions or billions of texts. Researchers are concerned that streams of seemingly authoritative computer-generated language might induce mistrust and misunderstanding. Parthasar...