Quick introduction to Artificial IntelligenceJuly 27, 2018
For a thorough introduction to AI, it is very important to know the difference between human intelligence, and computer programs. Humans can take in data, process it, and come to conclusions on their own, even if the computing speed and the memory are low. In case of computer programs, although there is much higher computing speed and memory involved, there is no way to come to a conclusion other than the pre-existing set of programs in the system. In layman’s terms, computer programs cannot learn on their own, and this is where artificial intelligence comes in. It is primarily a way for the computers to take in data, learn from them and apply what has been learned in real life situations at a much higher speed.
The concept of AI was conceived as early as 1940, and since then low-level AI has been used to test various theories, and to get a better understanding of the human brain. In recent years, Artificial Intelligence has been gaining quite a lot of momentum because of the availability of cheaper super-fast computers, faster bandwidth connections, and considerable investments in cloud computing. These advancements are propelling artificial intelligence to newer heights, and many industries are actually changing their policies to adapt to an AI boom. Even so, the current state of AI is considered weak at best when compared to its fullest potential.
A concept of singularity is often associated with AI, which says that AI would replace all humans after it has been perfected, and we would live in a post-human era. In the Singularity Summit of 2012, it was said by experts that this might come as soon as 2040. However, there are many who believe that it would be a long time before we get replaced by machines, simply because the perfect AI would take a lot of time to be ‘perfected’.
Teaching machines how to think is no easy feat, but we are slowly getting there. The U.S. Army came out with an unmanned rover which has shown instances of using what it learned to manoeuvre around obstacles. Although being in its rudimentary stages, AI is being applied to a lot of things around us which we do not notice. Google uses AI to suggest search results, aeroplane companies use AI for auto-pilot, spam filters in emails and the list go on.
To perfect AI, it is of paramount importance that the workings of the brain is understood, and can be recreated. If this is not achieved, then perfecting AI would take a lot longer. However, that has not stopped people from making programs which allows something called deep learning. It is considered one of the most advanced aspects of AI, where artificial neural networks are created to mimic the working of the human brain. AI can also be applied to natural language processing, where the communication with computers can be done by natural speech instead of computer language.
Machine learning is another type of AI which uses algorithms to learn from a set of given data. This does not use a specific set of rules and codes for generating intelligence. Generally, programmers compile a code and provide an input, and the computer gives the output. But in the case of machine learning, the inputs and outputs are provided, and the computer makes up the code to match the set of inputs and outputs. Once this is achieved, it is then used to derive the output for a different set of inputs, thus enabling the computer to use what it has learned at a rudimentary level.