This silicon monster named Darwin Monkey, designed in China, replicates the functioning of an animal brain with unprecedented neuron density. A remarkable leap toward more efficient, adaptable artificial intelligences, and perhaps one day, general AI.
An AI That Works Like A Brain: Darwin Monkey Disrupts Classic Models
You know what? I just came across some information that left me pondering: in China, a team of researchers from Zhejiang University has unveiled an incredible computing system called Darwin Monkey (or “Wukong”), a neuromorphic system mimicking the architecture of a macaque brain with over 2 billion artificial neurons and more than 100 billion synapses. In other words, we are transitioning from a traditional network-based AI to an AI that functions like a living brain.
What impresses me is not just the size of this machine but its ambition. The goal of this project? One day, perhaps, Artificial General Intelligence (AGI), an intelligence capable of perceiving, learning, and deciding in a global context. A significant departure from traditional AIs, which remain highly specialized.
Biologically Inspired Chips for Sobering Yet Stunning Power
Practically speaking, the true value of Darwin Monkey isn’t limited to its impressive numbers. It operates on 960 Darwin 3 neuromorphic chips, each managing approximately 2.35 million impulse neurons. These neurons mimic biological functions: they only activate when information is deemed worthy of being transmitted. This model is known as a “spiking neural network.” The result is a system that is closer to the human brain than any other.
In addition to this remarkable architecture, Darwin Monkey stands out with an electrical consumption of only 2,000 watts while delivering massive computational power. This is significant. We’re talking about a machine capable of competing with the largest supercomputers in the world, while consuming the equivalent of a microwave oven. A striking demonstration of energy efficiency serving performance.
From Research to Our Uses: What Darwin Monkey Truly Promises
For now, it’s not a consumer gadget, but the impacts are already being felt. Darwin Monkey enables the simulation of animal brains (zebrafish, mice, macaques) and serves as a research platform for neuroscience. Additionally, it can perform advanced cognitive tasks: content generation, logical reasoning, and mathematical problem-solving, thanks to an architecture that supports autonomous learning.
Thus, for the curious general public, this means that tomorrow’s intelligent systems (home devices, vehicles, robots) could be much more responsive, adaptive, and above all, less energy-hungry. The integration of a more flexible and less rigid intelligence also paves the way for applications that seemed unimaginable ten years ago. An AI that adapts, learns in real-time, and consumes little: that might just be the future.
This Supercomputer Paves The Way For AGI, But Raises Essential Questions
That said, even with 2 billion neurons, Darwin Monkey doesn’t possess human intelligence yet. Researchers acknowledge it as an important step but not an endpoint. Can it truly transition from one domain to another without being retrained? AGI remains a hypothetical aspiration. For now, it should be viewed as a living laboratory for the AI of the future.
So, how do we prepare for this revolution? First, by taking a closer interest in neuromorphic architectures. Next, by anticipating their impact on our daily lives. And above all, by maintaining a critical perspective. Because even the most impressive artificial brain is still a tool. And it is up to us, humans, to make good use of it.
