October 29, 2024

Zymy Le

12th Grade

Fountain Valley High School

How Are Artificial Intelligence and the Human Brain Connected? 

When thinking about the background of AI and the depth of its development and inspiration, the human body and brain take credit for just about all of it. Underneath the apparent difference between humans and AI, there are plenty of characteristics between both parties that are alike. Since the early 1900s, AI has been a topic of discussion amongst scientists and researchers, who looked towards nature for their resources. 

Neural Networks and the Components of AI’s “Brain”

Just as humans have their component of life being cells, AI has inspired networks which are called ‘units’, compared to the notable cells of a human’s brain. Brain cells of humans allow for depth and complexity in providing and storing information, just as AI’s ‘units’ have been inspired to collect an input and chuck out an output. AI also has the built in features of quick understanding and retaining information, which is in turn, inspired by the human’s brain ability to adapt and respond to various situations. 

However, ‘units’ are not the only factors inspired by humans. Various principles of the brain that AI has taken inspiration from are parallel processing, attention mechanisms, reinforcement learning, recurrent feedback, and many more. AI is taught to adapt these features that involve breaking up bigger tasks, focusing on specific terms, trial and error, and seeing what makes sense to output and what doesn’t. 

A human brain has three main sections, the dendrite, axon, and the soma. Each of which allows a person to collect signals, send the signal, process it, and finally finish the task. These factors contribute to AI’s most significant field of creation, neural networks. Just as humans have three main sections of the brain, AI mimics the structure of this system and displays its algorithms through this architecture. 

The Creation of AI

The main goal in creating a successful AI system is to particularly mimic the natural procedures of basic neural networks in the human brain. The first official model of a neural network originated from 18-year-old Walter Pitts and a proposed theory from McCulloch, both scientists interested in physiology and biological mechanisms. They both came up with “The Perceptron” which scaled a singular artificial neuron into a network of neurons. However, the technological system had some flaws and limitations, which is where the “Back Propagation" by Paul Werbos was introduced, which was a system designed to copy the human brain's strengths and processes. Werbos changed the game for the future and journey of AI, with his implementation of deep learning, which is present in a majority of AI systems today. 

Scientists at Stanford even say that “AI is eventually going to be able to do everything that humans can, and this will happen faster than we think” (Soltesz). Professors and researchers at Stanford worked on the focus of nerve processes in correlation to AI’s guided behavior and elaborate on circuits in the human brain and how they are able to be applied to structures of AI’s perception and sense of self. 

As artificial intelligence is getting more and more alike to nature’s biological background, some fear that it may become conscious as a biological concept. After all, for something to be created by the human brain itself, it must have comparing factors.