Scientists in China have developed a novel computing architecture that can train advanced artificial intelligence (AI) models while consuming fewer computing resources, potentially leading to the creation of artificial general intelligence (AGI). AGI is a hypothetical system that can reason, contextualize, edit its own code, and understand or learn any intellectual task that a human can.
The new architecture is inspired by the human brain and focuses on “internal complexity” rather than scaling up AI architectures. It builds upon the Hodgkin-Huxley (HH) network model, which simulates neural activity with high accuracy in capturing neuronal spikes. The HH network has rich internal complexity, where each artificial neuron is an HH model that can scale in internal complexity.
The researchers demonstrated this model can handle complex tasks efficiently and reliably, showing that a small model based on this architecture can perform just as well as a much larger conventional model of artificial neurons. This novel approach could eliminate the practical issues of scaling up neural networks and potentially lead to AGI.
While some scientists believe that scaling up neural networks could lead to AGI, others argue that novel architectures or combinations of different computing architectures are needed to achieve this milestone. The development of this brain-inspired architecture brings us closer to realizing AGI, which could have significant implications for various industries and aspects of our lives.
Source: https://www.livescience.com/technology/artificial-intelligence/novel-chinese-computing-architecture-inspired-by-human-brain-can-lead-to-agi-scientists-say