Open Philanthropy recommended a total of $14,459,002 in funding for projects working with deep learning systems that could help us understand and make progress on AI alignment. We sought applications for this funding here.
Recipients include (organized by research direction, with institution name in parentheses):
Interpretability:
- David Bau (Northeastern University)
- Eugene Belilovsky (Concordia University)
- Ruth Fong (Princeton)
- Roger Grosse (University of Toronto)
Measuring and forecasting risks:
- David McAllester (Toyota Technological Institute at Chicago)
- Michael Wellman (University of Michigan)
Techniques for enhancing human feedback:
- Yoav Artzi (Cornell University)
- Samuel Bowman (New York University)
- Greg Durrett (University of Texas at Austin)
- Faramarz Fekri (Georgia Institute of Technology)
- Mohit Iyyer (University of Massachusetts, Amherst)
- Victor Veitch (University of Chicago)
Truthful and honest AI:
- David Blei (Columbia University)
- Peter Clark (Allen Institute of AI)
- Dylan Hadfield-Menell (Massachusetts Institute of Technology)
- Tatsunori Hashimo (Stanford University)
- He He (New York University)
- Dan Klein (University of California, Berkeley)
- Colin Raffel (University of North Carolina, Chapel Hill)
Other:
- Chelsea Finn (Stanford University)
This falls within our focus area of potential risks from advanced artificial intelligence.