Coinbase software engineer Alex Wormuth has developed Stonkfly, an open-source experiment that uses the complete mapped nervous system of an adult male fruit fly to make Bitcoin trading decisions. Built on the MaleCNS v1.0 connectome, the simulation models 166,700 neurons and roughly 25.6 million synaptic connections. It converts live BTC-USDC market data from Coinbase into RGB visual input for thousands of simulated sensory neurons, with neural activity producing buy, sell or hold signals. Profitable trades activate 15 PAM11 dopamine neurons, while losses and trading fees activate two PPL101 neurons associated with aversive responses. Stonkfly began with $100 in paper trading and generated a $1 profit after one day, but Wormuth said the result likely reflected market movements and random neural responses rather than reliable learning. The project follows Wormuth's earlier DOOMFLY simulation and does not reproduce important biological processes such as chemical neurotransmitters, gene expression or broader neuromodulation.