Equity This Sequoia-backed lab thinks the brain is 'the floor, not the ceiling' for AI
56 snips
Feb 10, 2026 
Aidan Smith, former Neuralink engineer bringing neuroscience-inspired ideas to AI. Benjamin Spector, co-founder focused on creative, first-principles research and commercialization planning. Asher Spector, research-first builder working on data-efficient training. They discuss making models learn more like humans, bets behind a research-first lab, and what radical data efficiency could unlock for new AI capabilities.
AI Snips
Chapters
Books
Transcript
Episode notes
Data Efficiency As The Core Bet
- Flapping Airplanes bets that data efficiency is the key neglected frontier in AI research.
- The team believes solving it will unlock commercially valuable and novel capabilities beyond current scale-driven models.
Human Learning Suggests New Algorithms
- The founders see human learning as a different algorithmic regime from Transformers.
- They aim to explore alternative algorithms inspired by the brain to reach new trade-offs in capabilities.
Brain As Inspiration, Not Blueprint
- The brain serves as an existence proof and inspiration, not a strict blueprint.
- Flapping Airplanes will borrow ideas from biology but adapt them to silicon's different constraints.
