🙌 Guest Talk: Alex Turner, Former Research Scientist at Google DeepMind
Join Cornell AI Alignment for a talk and Q&A with Alex Turner, an AI alignment researcher and former research scientist on Google DeepMind’s Scalable Alignment team.
Alex has worked on a wide range of problems in AI alignment, including power-seeking behavior, shard theory, mechanistic interpretability, steering vectors, specification gaming, and understanding how learned values and internal representations form in neural networks. He has also mentored alignment researchers through MATS as the lead of Team Shard.
Alex recently left Google DeepMind following the company’s agreement to provide AI systems for classified U.S. government use. In his essay Why I Left Google DeepMind, he describes his efforts inside Google to advocate for stronger restrictions on military and surveillance uses of AI, and explains why Google’s eventual agreement led him to resign.
Alex will give a 30-minute talk, followed by an open Q&A and discussion.
When: Monday, November 16, 4:30-5:30 PM
Location: TBD
Food will be provided!
If you'd like to read some of Alex's work beforehand:
Whether you already work on AI alignment or are interested in how researchers think about the internal motivations and behavior of advanced AI systems, come join us for the conversation!