Similarity Between Brains and Computers – Same Vision?

A latest examine from the researchers at Johns Hopkins College suggests synthetic neural networks can “understand” 3D objects in the identical “first-glance” approach our brains do — in keeping with a recent study revealed within the journal Present Biology on Thursday.

People use natural brains to mannequin synthetic neural networks — so we should not be stunned once they mimic natural brains.

Till, nonetheless, they turn out to be frighteningly related.

SEE ALSO: NEUROMORPHIC COMPUTING: HOW THE BRAIN-INSPIRED TECHNOLOGY POWERS THE NEXT-GENERATION OF ARTIFICIAL INTELLIGENCE

Human mind versus synthetic neural community

The human mind processes visible info by means of a number of sections of the mind, with every combining completely different points of a perceived object to kind a full picture.

From a scientific standpoint, that is how we come to understand the exterior world.

Nevertheless, the latest examine reveals how neurons within the V4 space — the primary part within the mind’s object imaginative and prescient — symbolize 3D form fragments, not solely the 2D shapes — which runs opposite to scientific consensus held typically all through the final 40 years, reports TechXplore.

Nevertheless, scientists had been appalled to note the identical patterns the mind makes use of to see in a man-made neural community — after finding out the AlexNet algorithm, which is a complicated laptop imaginative and prescient community, Futurism reports.

Johns Hopkins College neuroscientist and examine creator Ed Connor and his group utilized the identical assessments of picture responses to pure and synthetic neurons and found spookily related response patterns within the mind’s V4 part and AlexNet’s layer 3.

“I used to be stunned to see sturdy, clear indicators for 3D form as early as V4,”  stated Conner in a press release concerning the stunning findings. “I by no means would have guessed in 1,000,000 years that you’d see the identical factor taking place in AlexNet, which is simply educated to translate 2D images into object labels.”

Studying extra than simply imaginative and prescient from computer systems 

“Synthetic networks are essentially the most promising present fashions for understanding the mind,” added Conner. “Conversely, the mind is the very best supply of methods for bringing synthetic intelligence nearer to pure intelligence.”

This newest examine might signify the approaching of a new norm for synthetic neural networks — the place as a substitute of utilizing what we all know concerning the mind to construct computer systems, we glance to computer systems to learn the way our brains work. It is a paradoxical time to be alive, whether or not we’re trying on the world with a human mind or a pc’s neural community.