Peter K. G. Williams

I did my undergraduate work at Harvard. I obtained my PhD with Geoff Bower at the UC Berkeley Department of Astronomy, where I conducted some of the first scientific studies using the Allen Telescope Array.

In 2012 I came to the CfA to work as a postdoc in the group of Prof. Edo Berger, mainly doing multiwavelength time-domain studies of ultracool dwarfs, gravitational-wave event counterparts, and other variable phenomena. From 2018 to 2022 I held a joint appointment as the Innovation Scientist of the CfA and the American Astronomy Society (AAS), and as Director of the WorldWide Telescope project. In 2023–2024 I was the scientific and technical lead for DASCH and prepared DASCH Data Release 7. Since 2025 I have been the Technical Lead for the IAU Minor Planet Center. I was named a 2025 Better Scientific Software Fellow.


Session

11-03
15:39
12min
Pervasive AI as A Second Enclosure Movement
Peter K. G. Williams

What is the relationship between artificial intelligence (AI) and the scientific and scholarly enterprise? I offer an answer to this question that draws on the history and scholarship of the commons. Historically, commons were shared agricultural land, most famously in preindustrial England. The term “commoner” simply refers to a person whose way of life depended upon the commons. In this sense, commoners and the historical English commons no longer exist, but not because of the well-known “tragedy of the commons.” Rather, the enclosure movement systematically converted them into private property of the wealthy and in so doing destroyed the commoners' way of life.

There are still commons today, but the most important ones consist of information, not farmland. They are cultural treasures like open-source software, Wikipedia, and the scholarly literature. From this perspective, scientists and academics are the new commoners, engaging in a distinctive “peer production” process that sustains the information commons. Although the tightening of intellectual property (IP) laws in the twentieth century has been referred to as a “second enclosure movement,” I argue that the rise of pervasive AI — ingesting essentially all existing information commons as input, and flooding them with new outputs — better deserves this moniker. In particular, generative AI interacts with peer production in such a way as to threaten the commons in a way that IP restrictions do not. It requires no extrapolations of the capabilities of AI systems to sketch a future in which the AI industry effectively appropriates the public wealth of the commons for private gain, in the process effectively destroying the commons and the way of life of those who depend upon them.

AI as tool for scientific discovery
Seminar Room 2