Organoid Intelligence: Genuine Substrate Migration or Biotech Hype Cycle?
Open questionComputing built from living brain cells is modeled as a possible migration of computation onto biological substrates — but current systems have poor between-session memory, cores that die off, and efficiency claims that are task-specific rather than general. Open until 2030: without stable multi-week memory and scaling past 1M neurons, the thesis collapses to a niche coprocessor — though defense and intelligence investment suggests someone treats it as a strategic hedge, not hype.
The engine's record — word for word
The engine models OI as the Technate's substrate migration strategy. But the divergence: current organoids compute as black boxes spliced onto silicon I/O. Between-session memory is poor. Necrotic cores limit scaling. The 1M-times-efficiency claim (FinalSpark) is task-specific, not general. Falsification: if by 2030, no OI system demonstrates stable multi-week memory retention AND scaling beyond 1M neurons, the substrate migration thesis collapses to a niche coprocessor role. The Jiang test is ambiguous: OI could genuinely disrupt the semiconductor chokepoint OR the same node structure could capture biological compute the same way it captured silicon (patents, platforms, defense contracts). Currently: DARPA and In-Q-Tel investment suggests strategic hedge, not hype. **Mar 24:** OpenAI biometric social network (Face ID / Eye-Scan Orb) = consumer-scale biometric deployment in 2026. Timeline accelerating.
Walk this on the live map →