Generative AI: Perspectives from Stanford HAI
The current wave of generative AI is a subset of artificial intelligence that, based on a textual prompt, generates novel content. ChatGPT might write an essay, Midjourney could create beautiful illustrations, or MusicLM could compose a jingle. Most modern generative AI is powered by foundation models, or AI models trained on broad data using self-supervision at scale, then adapted to a wide range of downstream tasks.
The opportunities these models present for our lives, our communities, and our society are vast, as are the risks they pose. While on the one hand, they may seamlessly complement human labor, making us more productive and creative, on the other, they could amplify the bias we already experience or undermine our trust of information.
We believe that interdisciplinary collaboration is essential in ensuring these technologies benefit us all. The following are perspectives from Stanford leaders in medicine, science, engineering, humanities, and the social sciences on how generative AI might affect their fields and our world. Some study the impact of technology on society, others study how to best apply these technologies to advance their field, and others have developed the technical principles of the algorithms that underlie foundation models.