Can We Use Algorithms to Understand How Our Brain Functions?

                                                                                                                               

Mark Albert presented information to Loyola’s neuroscience community explaining how the use of technology can be beneficial in a wide variety of fields, including neuroscience. He mentioned a paper called, Could a Neuroscientist Understand a Microprocessor?, by Eric Jonas and Konrad Paul Kording. In their laboratories these researchers took on the role of using a microprocessor in order to better understand the human brain by taking another complex system, a microprocessor, and comparing it to the most complex system, the human brain. There are still countless aspects of the brain’s circuitry that we don’t understand. The model organism that was used in their experiments was ones that we don’t commonly think of as being used in a science experiment- a microprocessor. They used the concept of video games as the overall model organism and three different video games we all might remember playing as the “behaviors” of the model organism. The three video games that were used in their study were Donkey Kong (1981, Space) Invaders (1978), and Pitfall (1981). It is hard to imagine any similarities between a human brain and a computer game, but in reality they share many similarities. The brain involves complex circuitry connecting neurons in order to drive thoughts and behaviors; the computer games do the exact same thing by forming circuits between each computer unit. The major difference is that we know and are able to recreate a computer’s circuits.

Science magazine recently published a fascinating interview between Mathew Hutson and Marco Zorzi, a psychologist at the University of Padua in Italy. Marco Zorzi and his colleagues used artificial intelligence neural networks to find out more about how the brain does certain tasks, such as letter learning. They trained the model using natural images, such as landscapes. The model then uses these images to form the basic vocabulary that is needed to learn about the various letter shapes and form better letter recognition. They chose this because there is a hypothesis about letters and language learning that occurs in humans.  The hypothesis is that the letters of a writing system match the visual environment of which that specific writing system originated. These researchers were manipulating artificial intelligence models in order to get more information about the way our brains work and are able to recognize letters. This is only one possibility out of the many that are being explored to gain a greater understanding of an extremely complex circuit network that is out brain. 

While much of the brain’s circuits still remain unknown, using a microprocessor’s system and artificial intelligence models are both steps in understanding the intricate neuronal circuits in our very own brains.

Works Cited: 

Jonas E, Kording KP (2017) Could a Neuroscientist Understand a Microprocessor? PLoS Compute Biol 13(1):e1005268.doi:10.1371/journal.pcbj. 1005268 

Hutson, M. "What artificial brains can tell us about the way our real brains learn". Science. 29 September 2017. Web 17 Oct. 2017. 

http://www.sciencemag.org/news/2017/09/what-artificial-brains-can-teach-us-about-how-our-real-brains-learn?utm_campaign=news_daily_2017-10-02&et_rid=338037551&et_cid=1577514 

Image:

Escobar, Matt. "Artificial intelligence: Understanding how machines learn" Robohub. 

http://robohub.org/artificial-intelligence-understanding-how-machines-learn/