Foundation models risk exacerbating ML's ethical challenges

Foundation models, or models capable of generating a range of media, pose risks, according to a new report from Stanford.