This article was first published on Medium. You can take a look at it here
Hollywood has created an image that equates Artificial Intelligence (AI) with iRobot, The Terminator and Ex Machina. While these are valid examples of AI, it is not all encompassing. AI is not a new technology, it has been around since the 1950's when John McCarthy coined the term Artificial Intelligence.
So what exactly is AI? Artificial Intelligence is defined as the science of making computers execute actions that require intelligence. In the context of AI, intelligence is the ability to adapt behavior to fit new circumstances.
The goal of AI is to create a computer that can think like a human. But why are we trying to make computers more like us? Isn't a computer more complicated than my brain? False, the most complex network/system in the world is the human brain. By teaching a computer how to think like a human, we can solve hard problems and improve on existing services such as speech recognition!
You're saying a computer can learn?
Yes, a computer can learn, but it is a somewhat slow process and the computer needs a lot of training. There are multiple forms of learning, the simplest is trial-and-error. Trial-and-error is where the program will try out actions at random until it finds success. Rote learning is when the program remembers the successful action and is able to produce that action the next time it is given the same problem. Trial-and-error and rote learning are relatively easy to implement, generalization learning is a little more challenging. Generalization learning allows the program to perform better in scenarios the program has not seen before.
Cool, the computer "learned" things, now what?
As I mentioned before, the implementation of AI can help solve complex problems in a variety of fields. Three popular fields are gaming, vision systems and speech recognition. AI has been implemented in gaming where it plays a crucial role in strategic games such as chess and tic-tac-toe. Vision systems incorporate AI to help systems understand, interpret and comprehend visual input on the computer. An example of a vision system is facial recognition on a camera. Speech recognition devices, such as Amazon's Alexa, use AI to listen and comprehend human speech.
While the goal of generalized learning might be years away, the impact of AI can already be felt in the short term through Weak AI applications. According to AI philosophy, there are two major types of AI: Weak AI and Strong AI. Weak AI is focused on developing technology that can act like a human. Weak AI applications make humans feel that the machine is acting intelligently, but in reality they are not. An example of a Weak AI application is a computer playing chess. The chess application isn't actually thinking or planning, it's making moves based on what it "learned" from human input.
Strong AI on the other hand, is focused on developing technology that can think and function similar to humans, not just mimic human behavior. We are still in the process of creating Strong AI applications, but one day, these applications will actually act and think just as we do. Strong AI applications are what we see portrayed in movies such as iRobot.
AI is a very broad field and has many subdomains such as machine learning and deep learning. As computers become more powerful, the AI revolution will continue and I'm sure we will see exciting products and services emerge from the market.
This is the fourth post in my "What is" tech blog series. I'll be writing more every week here and on my blog!