Cursor Projects

Developers have released an open source experimental project called Stonkfly, which connects a digital map of a fruit fly brain to real time cryptocurrency trading. The system converts Bitcoin market charts into visual data feeds, stimulating more than 166,000 virtual neurons and 25 million neural connections. Neural outputs from the simulated brain are then translated into automated trading decisions executed through crypto platform tools.

To guide the virtual brain, the system rewards positive financial performance with simulated dopamine hits and penalizes losses with negative chemical signals. However, the repository creator explicitly states that these synthetic rewards have not proven that the digital insect can make profitable trades. The codebase includes strict risk boundaries, such as spending caps and automatic cutoffs, to prevent unexpected financial losses during live testing.

This unconventional experiment highlights how modern software tools allow researchers to connect biological modeling directly with financial automation. While using biological neural maps for stock trading remains mostly an intriguing proof of concept, it raises interesting questions about whether biological neural structures might eventually handle pattern recognition tasks better than conventional artificial intelligence algorithms.