Can AI Mimic The Human Brain

 A question always arises what the end goal of AI is? With greater computational resources, we have built deeper neural networks which seem to predict almost everything? But the fact remains that we are yet to see a truly generalized AI that can make decisions like humans.  

For this, let's dive into the decision-making center of our body, aka the brain. The brain works magically if we truly observe it. From controlling our internal body processes to making conscious decisions and storing them as memories, it is truly wonderful. What makes it so adept at doing these tasks seamlessly? 

The brain is a vast collection of neurons that seem to be connected with each other. Not only this, it makes new connections depending on your experiences as well as what your body requires. Another factor which we must account for is that it does so using a minimal amount of energy. Simulating the brain as an artificial neural network takes about 10 billion times more energy than our brains do. Deep learning models were visualized on the fact of our brain in which there are layers, with each neuron in the initial layer being connected to all the neurons in the other layer. 

AI has come a long way in deep learning concerning performing specific tasks, sometimes even better than humans. However, the fact remains that a generalized AI is yet to be created that can seemingly learn different domains simultaneously. For example, we have AI performing image processing related to specific domains. However, it is difficult to build a system that can learn everything that it 'sees.'

One key feature of the brain is that it performs processing and memory storage together. This is a key difference. Computers today usually have their logical units separated from their memory units. There is a communication channel between these two units to perform meaningful operations. Another factor to account for is that the brain is somewhat a mixed system, i.e., analog and digital. Computers, on the other hand, purely rely on boolean logic to perform the most basic tasks.  

A common criticism of  'conventional' AI models is that we don't really know what goes on in the inner layers. This is vastly different from our brains. As humans, we can explain our thought processes which led us to take a particular action. We can basically communicate with our 'inner layers' to understand what truly caused the decision to be taken. 

Thus, the question arises whether AI can mimic the human brain. To do so, we'll probably have to reimagine how to design our algorithms and whether we also need to change the way we make computers. There are many ifs and buts, but they can probably be answered if we understand how our brains work. Or maybe the more we learn about our brain, the more it complicates itself?


Comments

Post a Comment

Popular posts from this blog

Time Travel

What if time travel doesn’t necessarily mean travelling back or forward in time but instead viewing a particular point in the universe from two different points at the same time. For example, if there is a star at a distance of 70 million light years from the earth any observation from that star means we are looking at 70 million years into the past i.e. time travel. So instead of moving back in time, we are essentially travelling to another point of reference instead.    The question   remains why haven’t we received any time travelers from the future. Maybe its possible for future time travelers not to be present on the reference point which it is observing at that exact moment as basically that would mean seeing its present. It can only observe the past as its existence is not possible in that time frame.   In a particular relative time its not possible for an individual to be present at two points simultaneously.   However than the question arises...
Read more

Champions Again: A Fan’s Catharsis After India’s 2024 T20 World Cup Triumph

The world is healing. There's background music being continuously played. I feel like dancing and singing my heart out.... Okay, enough of the melodramatic stuff. But hey, this is the stuff dreams are made of. India has won an ICC trophy after 11 years. A World Cup in any format after 13 years and a T20 World Cup after 17 years. I think I deserve a release after all those years of anguish. The last time I saw an Indian team lift a trophy was in the newspaper and highlights of the Champions Trophy. I couldn't see it live as it was already past midnight, and the match did not look like it would begin after the rain. The next day, the newspaper was filled with the team looking dashing in their white coats, highlights showing Dhoni jumping behind the stumps as Ashwin's ball spun away from that bald England spinner. The last time I watched an Indian team lift an ICC trophy live was the 2011 World Cup. It's the year I started watching cricket when my parents thought that at l...
Read more

Neuromorphic Computing

Imagine a scenario in which your computer works like your brain. I mean, instead of operating on boring bits of 1's and 0's, which we all know, your computer functions on spikes much like our "neurons" in the brain. Don't worry these computers are not being built by pieces of our brains(probably a good plot for a post-apocalyptic movie) but instead modelled on the basic phenomena in which our brains work. If this all seems too sci-fi for you, be prepared to be amazed. Although this field may seem to be new, it has been in existence since the 80s. Fun fact, before the usage of digital logic in computer systems, there was significant research going into the development of analog computers. The key difference between analog and digital is that digital basically works on low and high levels, while analog works over a range of values. Anyway, the reason for mentioning this is that analog computing forms a significant part in neuromorphic computing. In fact, neuromorphi...
Read more