Google researchers help develop the first complete digital map of an adult male fruit fly's brain, which modders programmed to play Doom, trade BTC, and more

If you are interested in AI agents, you can now use a fruit fly's digital brain to perform complex tasks.

Google Research partnered with HHMI Janelia Research Campus to create MaleCNS v1.0, the first complete digital map of an adult male fruit fly's brain. This connectome, detailing over 166,000 neurons, follows the release of the female fruit fly brain map in 2024. The digital model has gained viral attention as modders programmed it to perform various tasks, including playing video games like Minecraft and Doom, trading Bitcoin, and even navigating simulated environments. Researchers aim to use these detailed neural maps for breakthroughs in neuroscience, pharmacology, biotechnology, and artificial intelligence. The creation of both male and female connectomes also enables direct comparison of neural wiring differences between sexes in the model organism. While the digital brain is not sentient, its successful implementation in diverse virtual tasks has sparked widespread online experimentation and debate.

The successful integration of the fruit fly connectome into diverse applications highlights the increasing feasibility of simulating and controlling complex biological systems. This development suggests a future where digital twins of organisms could accelerate research and enable novel applications across various scientific fields. The viral adoption by modders also underscores a growing trend of citizen science and open-source creativity in pushing the boundaries of what's possible with advanced biological data.

The creation of comprehensive connectomes, like MaleCNS v1.0, is a significant step towards understanding the fundamental principles of neural computation. For AI development, this offers a pathway to explore biologically inspired architectures that might surpass current deep learning models. As these models become more detailed and accessible, they could lead to more sophisticated AI agents and a deeper understanding of intelligence itself, potentially impacting fields from medicine to robotics.