This month, we visited 13 US universities to connect with AI researchers, discuss AGI progress, and inspire more people to participate in ARC Prize.
With over 1,500 students and professors from top US AI programs, we're proud of the impact the tour made.
We want to share a few tour highlights and give thanks to everyone who helped make this event series happen.
We're hosting a virtual event tomorrow to make sure everyone around the world has access to the content we shared while on tour.
Come hear directly from François Chollet, Creator of ARC-AGI, about technical approaches that might defeat ARC and win you the Grand Prize!
And our encore appearance at the University of Washington takes place October 30.
Todd Gureckis (Professor of Psychology) and Brenden Lake (Associate Professor of Psychology and Data Science) have co-authored 2 papers studying how humans interact with ARC-AGI. They're the directors of the NYU Minds, Brains, and Machines initiative designed to understand intelligence through psychology, neural science, philosophy, linguistics, data science, and AI.
Along with Wai Keen Vong (Research Scientist, NYU Center for Data Science) and Solim LeGris (Doctoral Researcher, NYU Computation and Cognition Lab & Human and Machine Learning Lab), we are excited to collaborate on upcoming improvements to the ARC-AGI benchmark.
Josh Tenenbaum (Professor, Department of Brain and Cognitive Sciences) is focused on understanding, "that most elusive aspect of human intelligence: our ability to learn so much about the world, so rapidly and flexibly."
His lab has likely the highest density group of researchers working on program synthesis in the world. Some are directly or indirectly working on ARC-AGI.
The lab studies an intriguing diversity of research topics.
Guy Van den Broeck (Professor of Computer Science and Samueli Fellow) directs the Statistical and Relational Artificial Intelligence (StarAI) lab at UCLA. The lab researches Machine Learning, Knowledge Representation and Reasoning, and AI.
We're fans of Guy's notable 2022 paper "On the Paradox of Learning to Reason from Data", which explained how a model can attain near-perfect accuracy on in-distribution test examples while failing to generalize to other data distributions over the exact same problem space. Plus, fun fact: he appeared on comedian Daniel Tosh's podcast.
The lecture hall where we spoke was right across from the room where the internet was born!
David Haber (General Partner) and team hosted us at the beautiful Andreessen Horowitz NYC office. We spoke with 75 amazing AI researchers and startup founders including Zenna Tavares (Basis), Aymeric Zhuo (Agemo), and Marco Mascorro (Fellow AI, a16z), and so many more.
We shared the story behind ARC Prize and learnings from our university tour. This event reinforced our belief that ARC-AGI is a powerful guiding force for anyone working to build AGI.
Thank you to the following people and groups that helped make our idea to visit over a dozen universities a reality. We're grateful for their time and support and we look forward to more collaborations in the future.
Don't forget to register for the virtual event tomorrow!