A wiring diagram of a fruit fly‘s entire nervous system, painstakingly mapped by neuroscientists over years, is now watching Bitcoin candlestick charts and placing real trades. That is not a metaphor: it is a GitHub repository with an MIT license and $100 in live Coinbase funds behind it.
The project is called Stonkfly, and it signals something worth noting. Connectome-based brain models are leaving neuroscience labs and landing in places their creators never designed them for.
How a Fly “Sees” the Market
Stonkfly converts Bitcoin price charts into neural signals and lets a simulated insect nervous system decide whether to buy, sell, or hold.
Stonkfly runs on MaleCNS v1.0, a complete neural graph of an adult male fruit fly’s central nervous system. The scale is not trivial: 166,700 simulated neurons, 25,582,938 directed connections, and 124,177,617 synaptic contacts, according to The Register.
The system has no body. Its entire world is market data.
Coinbase BTC/USDC candlestick charts feed into the model as RGB images, stimulating thousands of simulated neural signals. A fixed decoding layer reads the resulting spike patterns and translates them into buy, sell, or hold proposals. Those proposals pass through Coinbase AgentKit to place actual orders.
Profit events trigger engineered reinforcement signals in designated dopamine neurons. Per the project documentation, these are artificial signals, not modeled pain or pleasure receptors.
“I gave the fly brain $100 to trade bitcoin. Dopamine neurons are stimulated when the fly makes profit. Neuron activity controls buy/sell decisions and makes trades on coinbase. Will the fly get rich?” according to Wormuth’s post on X.
What Stonkfly Is Not
The project’s own documentation is candid: no profitable learning or live-funded performance has been demonstrated.
The GitHub documentation states plainly: “No profitable learning, strategy improvement, biological replication, or live-funded performance has been demonstrated by this repository’s tests.” The documentation also recommends comparing results against cash or simple exposure baselines, noting that a rising Bitcoin price makes any buyer look competent.
Stonkfly follows Eon Systems’ earlier 2026 demonstration, which used a female fly brain scan (roughly 125,000 neurons, 50 million synapses) to control a physics-simulated body in MuJoCo. That fly walked, groomed, and fed. Stonkfly strips the body out entirely and drops the nervous system directly into a market.
Alex Wormuth (GitHub handle nftechie) previously used the same male CNS connectome to run a DOOM playthrough, a project called DOOMFLY, before pivoting to Bitcoin trading. The underlying connectome data came from large-scale neuroscience mapping efforts including FlyWire and collaborators associated with Google.
The project now has a public website. It tracks a “persistent fly connectome watching candlestick charts” across NVIDIA, Tesla, Apple, Amazon, Microsoft, and Alphabet. It also introduces a community token, STONKFLY, on Robinhood Chain, positioning the experiment somewhere between research artifact and consumer-facing crypto product.
Years of painstaking neuroscience have been compressed into a meme-ready trading bot and a speculative token. Connectome-based models are entering non-biological domains faster than anyone anticipated. The legal and ethical frameworks for what that means do not yet exist.
