Artificial Intelligence (AI) is one of the most disruptive technologies the 21st century has to offer. The future of business can be completely transformed with the application of AI. At its core, AI is the ability for machines to think for themselves when given a task, allowing the machine to handle its task independently of human input other than for the initial task.
In this article, we will draw a comparison between artificial intelligence and human intelligence.
Most people have used Google Assistant, Siri, Cortana or Bixby at some point in your life. What are these? These are our digital assistants. They understand our voice. They can answer our queries. If you ask “Who won the 2011 World Cup final?”. It will answer “India”. If you ask it to call your mum, it can go through your call logs and call your mum. How are all these possible? All of these are a few applications of Artificial Intelligence.
In formal terms, AI is the ability of computers to learn and think. It is all about emulating the same process that our brain uses to learn and think into the computer.
How Artificial Intelligence Works?
Let us try to understand this with an example of predicting the weather for 2020.
First, we have to provide the computer with a lot of data. Let us suppose we feed data from 2006 to 2019 into the computer.
Now we will use the 80:20 rule. 80% of the data will be used by the computer to develop algorithms and learning. The remaining 20% will be used for testing.
Then we feed the entire 80% of the data into the computer. The computer analyzes the data and looks for trends and patterns in it. It tries to draw relationships among the data. The computer tries to learn from the data feed into it.
Next, we use the rest 20% of the data to test its accuracy. We feed input data into the computer and wait for its output. We cross verify this output with the actual values and check its accuracy. If we are not satisfied with the accuracy, we can tweak the algorithm to give more accurate results. We repeat this process until we are satisfied with the output. This way the computer learns to make accurate weather predictions for the year 2020.
Comparison between Artificial Intelligence and Human Intelligence
Humans naturally have the traditional five senses: sight, taste, touch, hearing and smell. Humans are equipped with these senses to navigate complex environments in order to survive.
AI tries to mimic human behaviours in a machine. For certain tasks, an AI might be equipped to detect light using sensors or to respond to sound such as an Amazon Alexa device. However, so far machines still struggle to understand sensory inputs it was not priorly trained to recognize, just try speaking to voice recognition software in an uncommon accent!
The human brain is much more powerful than the most advanced supercomputer ever built. There are so many processes taking place at the same time inside your brain, although difficult to estimate, researchers believe that the human brain can process information at 1 exaFLOP, which is equivalent to a billion billion calculations per second.
On the other hand, AI still has not reached this speed or complexity. Currently, the fastest supercomputer in the world is the Tianhe-2 which has a maximum processing speed of 54.902 petaFLOPS. A petaFLOP is a quadrillion calculations per second, a far cry from the speed of the human brain.
However, with advances in quantum computing, machines could one day outperform the human brain, as ultimately the human brain is still limited by its organic and material nature.
Human intelligence is based on various stimuli received by the brain throughout our life. Many complex processes and stimulants are flowing through your body that helps you in decision making. The human brain is capable of making decisions, even in situations that we have never encountered in our lives.
However, AI learns through examples and training. It cannot respond to an unknown situation. It builds its knowledge based on past data and values. Upon the development of a machine capable of adapting to entirely novel situations, it would no longer be an AI, instead earning the moniker Artificial General Intelligence (AGI). So far humanity is a long way off creating a true AGI.
Human intelligence is prone to bias. Our decision-making process is influenced by biases. For example, if a judge has to pass a sentence between his daughter and an unknown person, his judgment will generally be biased in the case of his daughter.
On the other hand, AI is not biased. No matter what the circumstance, it will always give an honest result. The bias for an AI instead comes from the data it is trained upon, a limited data set will increase the bias present in the outcomes. For instance, training an AI to recognise horses in photos by inputting that the subject has four legs would lead the AI to potentially determine all four-legged animals as horses.
Human intelligence is full of feelings. We are filled with feelings like love, hatred, guilt, happiness, sadness, and whatnot. For many people emotions are a core part of the human experience.
AI is yet to have feelings. We have not been successful in instilling feelings into machines. The esoteric nature of emotion for humans is something that researchers do not fully understand, while there are chemical processes behind the emotions that humans feel, that could potentially be replicated in a machine, some may argue that they are still not true feelings.
When it comes to mathematical calculations and predictions, human intelligence is a bit slow. Our intelligence is not trained to work with huge volumes of data. Try to calculate in 897 x 78657x 4645373 in your mind, you will certainly struggle!
Artificial intelligence was specifically created to help humans with these calculations. We have developed huge supercomputers that can do millions and billions of calculations within seconds. The human mind struggles with mathematics beyond a certain amount of figures, the complexity of our brains struggle to be singularly focused on one vast task. So while a human brain might be more powerful than a supercomputer, its complex juggling of tasks is its ultimate downside.
In Part 2 we will further examine the differences between humans and AI