An autonomous bicycle powered by a new type of artificial intelligence chip is demonstrated in research published in Nature this week. The chip combines brain-inspired and computer-science-based approaches to artificial intelligence. This hybrid technology may have the potential to improve the capability of these systems to achieve artificial general intelligence - a platform that could, in principle, perform any task that a human is capable of.
There are currently two general approaches towards developing artificial general intelligence: one based on neuroscience, attempting to closely mimic the brain; and a second that is computer-science-oriented, which uses computers to perform machine-learning algorithms. The ultimate goal would be to combine the two, but the systems use distinct and incompatible platforms, thereby limiting the development of artificial general intelligence.
Luping Shi and colleagues have developed an electronic chip that can integrate both approaches. Their hybrid chip has many functional cores that are highly reconfigurable, enabling it to accommodate both machine learning algorithms and brain-inspired circuits. The processing capabilities of this hybrid chip are demonstrated in an unmanned bicycle system. The bicycle responds to voice commands, self-balances, can detect and follow a person and can avoid obstacles. The authors conclude that this work might contribute to the ongoing development of platforms for artificial general intelligence.
Engineering: Earmuffs measure blood alcohol levels through the skinScientific Reports
Physics: Modelling improvements to ride-sharing adoptionNature Communications