‘Independent scientists must take the lead in monitoring AI’

Core principles in this regard are accountability, transparency and independent oversight. In doing so, they call for the immediate establishment of an independent oversight body made up of scientists.

 

Generative AI is a form of artificial intelligence that allows someone to automatically create text, images, audio and other content. Using it has advantages, but there are also disadvantages. There is therefore an urgent need for scientific and societal oversight in the rapidly developing field of artificial intelligence.
First author of the Nature article is Claudi Bockting, professor of clinical psychology of psychiatry at Amsterdam UMC and co-director of the Centre for Urban Mental Health. She argues that "AI tools can flood the internet with misinformation and 'deep fakes' that are indistinguishable from the real thing. Over time, this could erode trust between people, in politicians, institutions and science. Therefore, independent scientists should take the lead in testing and improving the safety and security of generative AI. However, most scientists now have limited ability to develop generative AI tools and cannot evaluate them in advance. They lack access to training data, facilities or funding.' The other authors of the guidelines are: Robert van Rooij and Willem Zuidema, from the Institute for Logic, Language and Computation at the University of Amsterdam; Eva van Dis from Amsterdam UMC; and Johan Bollen from Indiana University.