About this newsletter

This newsletter is where I'll share observations on AI, language, and culture, drawn from my research, reading, and frequent travel to conferences, workshops, and other events around the world.

The news about AI can be overwhelming, and even for people who have been working in this area for a long time, the pace of both industry and research is too much. At the same time, AI is reshaping human culture, and we need to pay attention to these changes.

There are already a lot of AI newsletters focused on models, benchmarks, and product launches. This one is focused on what's happening at the intersection of AI and culture: where AI touches storytelling, poetry, history, online communities, beliefs and values, and more.

My goal is to share frequent, short, and informal updates in the form of "field notes" or "dispatches" from someone who is both a participant in and observer of AI research.

About me

I'm Maria Antoniak, a researcher whose work spans AI, language, and culture. I approach AI as someone who is both a researcher studying language/computation and a reader/writer/citizen of the internet who cares deeply about human creativity, meaning making, and community. My core expertise is in computer and information science, but I also collaborate a lot with people in the humanities and social sciences.

I think that humanities data and theory should guide how we build and evaluate AI, and I also think that AI methods can open new, exciting readings of humanities data at scale.

For example, some of my research has included:

  • Auditing models' pretraining data for poems and examining how model training pipelines break when confronted with poetry's unique data formatting.

  • Examining how people co-write stories with AI, including the personal information they reveal within these stories and the iterative permutations they make to their stories during their editing process.

  • Mapping the narrative structures of people's birth stories, shared in an online community, and the power dynamics described in those stories.

I'm currently an Assistant Professor of Computer Science at the University of Colorado Boulder, where I'm also affiliated with the Department of Information Science and the Boulder NLP Group. I previously worked at the Allen Institute for AI (AI2) in Seattle and the Pioneer Centre for AI in Copenhagen, and I've spent many shorter stints in industry at places like Microsoft Research FATE, Twitter Cortex, and Facebook Core Data Science.

I have a PhD in Information Science from Cornell University, an MS degree in Computational Linguistics from the University of Washington, and a BA in Liberal Studies from the University of Notre Dame.

You can learn more about me and my work on my website.


Why Leaflet?

I have a personal blog on my academic website, but that contains longer form writing on diverse academic topics, without any subscription option. I also post on Bluesky a lot, but I want a place for more focused and slightly longer form content. I don't want to support Substack, and I do want to support the open Atmosphere ecosystem, which Leaflet is built on.