Science Articles

WHAT ARTIFICIAL BRAINS CAN TEACH ABOUT THE LEARNING OF OUR REAL BRAINS

The human mind is the most difficult thing to study. Brain is complex, mysterious with multiple roles to play. Every individual has varied thinking pattern with little to no similarity. The study of human brain is the most complex field of science which has no boundaries whatsoever. Thus, the scientists have resorted to using artificial intelligence for understanding how the real brain operates.

The human mind is the most difficult thing to study. Brain is complex, mysterious with multiple roles to play. Every individual has varied thinking pattern with little to no similarity. The study of human brain is the most complex field of science which has no boundaries whatsoever. Thus, the scientists have resorted to using artificial intelligence for understanding how the real brain operates.

THE INTERVIEW AND THE CRUX OF WHAT WAS SAID

Psychologists have employed the use of artificial intelligence to understand the working pattern of our brains. Marco Zorzi is a psychologist from the University of Padua, Italy who first employed the use of artificial neural networks for understanding the connection between recognition of written alphabets and visual cortex. His interview gave an insight about the work done on the subject.

THE OBSERVATIONS

Ø  Neural recycling was used by his team for studying letter perception by the brain. The team observed that the artificial brain showed signs of letter recognition when trained. They used the natural images in comparison to the one that weren’t. Recycling helps in recognition.

Ø  The team used unlabeled images of a letter to show to the artificial neural network which was already trained with natural images. The process was called generative owing to the fact that patterns from top to down were generated. Artificial intelligence showed that it recollected the data for understanding the new sensory information. Sometime later, labels were put on the letter by algorithms while the scientists guided it when it went wrong.

Ø  The supervised learning method helped in understanding the feedback connections of the brain. The intrinsic brain activity forms images first. When it is combined with the sensory activity, the feedback helps in interpreting the input and revealing the output. For example, if in a sentence, some word is hidden or partially erased, the brain automatically fills in the data of what it expects to see.

Ø  The artificial intelligence network that was trained by using natural images showed more response towards real letter images over the made up images.

Ø  Human cognition is another thing where deep unsupervised learning can help in understanding the visual output in a network of neurons. It is easy to understand things with unsupervised learning.

Ø  The scientists also found that there is no evidence that dyslexia happens due to one deficit in the brain.

THE RESULTS

The observations and further experiments would help in studying the working of the human brain which would enable the scientists to explore more learning solutions for people with difficulty.