🤖 AI Summary
Swiss startup FinalSpark and other labs are developing "wetware" biocomputers by growing human neuron clusters (organoids) from stem cells and interfacing them with electrodes to send and record electrical signals. In the lab demonstration, organoids derived from donated skin‑cell stem lines respond to simple keyboard-triggered stimulations with EEG‑like activity, showing early forms of input/output and plasticity that researchers hope to harness for learning tasks. Teams including Cortical Labs and Johns Hopkins have already shown related feats (artificial neurons playing Pong; mini‑brains for drug testing), and proponents argue biocomputing could one day perform certain AI workloads with far lower energy costs than silicon.
Significant technical and ethical hurdles remain: organoids lack vascularization so sustaining them is difficult (FinalSpark keeps them alive for up to ~4 months), and researchers observe unexplained bursts of activity before organoid death. Experts caution wetware is unlikely to replace silicon broadly but could complement it in niches such as neuromorphic research, low‑power inference, and improved disease models that reduce animal testing. The work raises scientific, practical and philosophical questions about using living tissue as computational substrates, while offering a provocative new avenue for energy‑efficient, biologically inspired AI.
Loading comments...
login to comment
loading comments...
no comments yet