Sahil Kulshrestha
Bio
Sahil Kulshrestha is an independent AI alignment researcher and founder of Groundless Alignment (groundless.ai), where he develops what he calls "Live Theory" — a framework for understanding risks and opportunities from moderately capable AI systems deployed at scale as infrastructural technology. His core research agenda critiques the dominant "agentic" frame for AI and argues instead for a co-agentic vision in which AI functions as attentive infrastructure rather than autonomous agents. He received a Long-Term Future Fund grant for the development of a mathematical language for highly adaptive entities with stable commitments rather than stable mechanisms. He contributed to the PIBBSS fellowship 2022 (suggesting and facilitating its deep reading group format) and provided curriculum inputs for the "Key Phenomena in AI Risk" facilitated reading group alongside researchers from FHI, PIBBSS, and MIRI. He has engaged with researchers including Scott Garrabrant and received a $100,000 personal donation from Richard Ngo in 2025 to support his live theory work, with Ngo describing him as having one of the most ambitious and philosophically coherent overall visions for the future of AI among researchers he had consulted.
Links
- Personal Website
- https://groundless.ai/
- -
- Twitter / X
- LessWrong
- sahil-1
Grants
from Long-Term Future Fund
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Mar 23, 2026, 12:47 AM UTC
- Created
- Mar 20, 2026, 2:57 AM UTC