🚨The Race to Simulate a Brain Has Started
You may have already seen the news about Eon Systems running a fruit fly brain in a simulation and if you did, you probably just assumed it was just another AI headline, but it isn't.
What's happening right now is the beginning of a very real scientific race. The goal isn't just about creating better AI models. The real goal is building the first fully functional digital brain.
Eon Systems is one of the new players pushing this idea forward. Their recent experiment used the complete wiring map of a fruit fly brain which according to them contained about 140,000 neurons and tens of millions of synapses, and connected it to a simulated body. The digital brain receives sensory signals, processes them through the neural network, and sends motor commands back to the body. In other words, behavior emerges from the biological wiring itself. Just a neural circuit running in software.
But Eon isn’t alone because behind the scenes there are really three camps all moving toward the same destination.
The first is the startup approach, represented by Eon Systems. Their strategy is to map brains, simulate them, and scale up from simple organisms to more complex ones.
The second group is the academic connectomics community. Projects like FlyWire have spent years reconstructing the entire neural wiring of the fruit fly brain using electron microscopy and AI tracing. Their work produced the massive datasets that make simulations like Eon's possible.
Then there's the heavyweight, DeepMind and Google's neuroscience teams. They're not advertising 'brain uploads,' but their research into neural circuits, biological computation, and large scale neural simulations is clearly moving in the same direction. Instead of copying a brain neuron by neuron, they're trying to extract the computational principles that biology uses and build AI systems around them.
A biological brain runs on about 20 watts of power. Modern AI systems require entire GPU clusters consuming megawatts. If researchers figure out how brains achieve that efficiency, the entire AI industry would change overnight.
Medicine would change at the same time. A working brain simulation could allow scientists to test treatments for neurological diseases like Alzheimer's or Parkinson's inside a digital model before ever touching a patient.
Even robotics could benefit. Animals control bodies in ways robots still struggle to replicate. Digital nervous systems could give machines the same adaptive behavior biology evolved over hundreds of millions of years.
Mapping a brain requires slicing it into microscopic layers and scanning every connection with electron microscopes. The fruit fly connectome already took years of work and produced massive datasets. A mouse brain would be hundreds of times larger. A human brain would contain roughly 86 billion neurons and hundreds of trillions of synapses.
That's an engineering challenge on a completely different scale. Still, the direction we are heading is becoming clear. First the worm nervous system was mapped, then the fruit fly. Next will most likely be zebrafish and small mammals.
Each step pushes the boundary of what can be reconstructed and simulated.
The fruit fly experiment might look small, but a biological brain blueprint running inside a digital environment and generating behavior is huge.
For the first time, researchers are starting to move from studying brains piece by piece to rebuilding them as complete systems.
#Neuroscience #AI #BrainSimulation #ArtificialIntelligence #FutureTech
Source:
eon.systems