These are fantastic explainers. A must-read

OMB released a memo that directs federal agencies to increase their risk management practices using and procuring AI. The memo is solid, but the standards it leans on are underspecified. Today Helen Toner and I published an op-ed arguing for investment in AI evaluation. 🧵

Had fun talking with Miriam Vogel last week about AI + China, national security, agents, and making policy through uncertainty. Full podcast episode here:
equalai.org/podcast/

A statement from Helen Toner and Tasha McCauley:

How are large language models trained and modified? Can they move beyond chatbots into full-fledged agents?

In their new 3-part explainer series, Helen Toner, Thomas Woodside & Matt Burtell offer a non-technical deep dive into these questions and more. cset.georgetown.edu/article/the-su…