Coinbase Engineer Connects Virtual Fly Brain to Crypto Trading

A software engineer at Coinbase, Alex Wormuth, has developed an innovative project called Stonkfly, which connects a digital version of a fruit fly's brain to a cryptocurrency trading account. This simulation allows the fly's brain to make trading decisions based on live price charts, specifically for Bitcoin (BTC), Ethereum (ETH), USD Coin (USDC), and Solana (SOL).

The fly's brain operates using a connectome—a detailed wiring map of an adult male fruit fly's nervous system—comprising approximately 166,700 neurons and 25.6 million synaptic connections. The brain processes visual data from a candlestick chart displayed across its two simulated eyes, enabling it to decide whether to buy, sell, or hold based on the colors of the pixels it perceives.

As of now, the fly's trading performance has resulted in a loss of nearly one dollar, with its current holdings valued at $59 in USDC, nearly $5 in BTC, $17 in ETH, and $17 in SOL. The project rewards profitable trades with virtual dopamine, stimulating specific neurons in the fly's brain, while losses trigger a different response.

While the fly's trading activity is limited to paper money, with a starting balance of $100 and restrictions on order sizes and daily attempts, the experiment raises questions about the implications of mapping a brain and its potential to adapt in a trading environment. Wormuth has made the code for Stonkfly available on GitHub, allowing others to explore this unique intersection of neuroscience and cryptocurrency trading.