My name is Nick Merrill. I work on applied AI research at the Forecasting Research Institute. Before that, I directed the Daylight Lab at the U.C. Berkeley Center for Long-Term Cybersecurity (CLTC).
Some past projects I’m proud of, in reverse chronological order:
Internet Atlas (2020-2025), quantitative work on how the internet shapes geopolitics, and vice versa.
MLFailures (2019-), an open-source curriculum on AI fairness and bias. I teach these materials to UC Berkeley students, and to a mix of mid-to-senior-level U.S. congressional aides and Executive Branch officials.
Adversary Personas (2018-2020), which is now used by the U.S. Cybersecurity and Infrastructure Security Agency, Taiwan’s Ministry of Digital Affairs, and Meta. You can read some (independently authored) research on this technique here.
Passthoughts (2013-2018). Think your password! Really: log in using a secret thought. Your thought—and the way you, specifically, think it—authenticate you.
Digital art (2012-2014). Including Lykkemat, and a mind-controlled Reddit reader.
