Brain-inspired Intelligence

by+Li+Zhuoxi

Brain-inspired intelligence refers to a next-generation artificial intelligence （AI） system modeled to mimic human brain behavior using neural structures and an information management scheme. Its a revised AI system simulating human thinking through the study of the human brain.

Intelligentization has become the goal for the next step of the technological progress， and the human brain has become the holy grail of AI. The development of brain-inspired intelligence could help humans learn how the brain handles information， which could in turn inspire drastic improvement in a wide variety of fields.

China established its first statelevel brain-inspired intelligence engineering laboratory， the National Engineering Laboratory for Braininspired Intelligence Technology and Application （NELBIITA）， in Hefei， capital of Anhui Province， through joint efforts from Fudan University， Shenyang Automation Research Institute under the Chinese Academy of Sciences （CAS）， the Micro-electronics Research Institute under the CAS and Baidu Inc.， under the direction of the University of Science and Technology of China（USTC）.

The laboratory specializes in research on the human brain mechanism and developing AI technology， using the human brain mechanism as a reference point to improve emerging industries such as braininspired cranial nerve chips and brain-inspired intelligent robots.

According to Wu Feng， director of the NELBIITA and executive dean of the School of Information Science and Technology under the USTC， the development of braininspired intelligence is hindered by three bottlenecks： uncertainty about the brain mechanism， inaccurate models and methods of brain- inspired calculation and restraints of computing fabrics and capabilities. The bottlenecks could be overcome via the following six steps：

A. Conduct big data analysis on the human brain to grasp its operational laws.

B. Help brain-inspired intelligence learn the basic functions of the human brain by imitating its visual perception， auditory perception and language processing to ultimately make intelligent decisions similarly to the human brain.

C. Insert chips into braininspired intelligence and produce them in batches that can be widely applied in various sectors.

D. Erect a brain-inspired computing platform for large-scale concurrent computation.

E. Establish a brand-new computing system for human brain computer calculation based on quantum technology.

F. Develop brain-inspired intelligent robots with independent abilities.