Artificial intelligence, a term coined by John McCarthy in 1955, is transforming us into an automated and machine-driven society. In everyday life, we use a lot of artificial intelligence. Companies are pumping in huge investments for research and implementation of artificial-intelligence-based products, which they believe will transform their businesses.
In 1950, Alan Turing came up with the notion of machine intelligence. The Turing Test gave an idea how we can decide whether a machine is intelligent or not? He wrote in his paper Computing Machinery and Intelligence that when a machine’s activities are indistinguishable from a human, it is evidence of a machine exhibiting intelligent behaviour. The paper is regarded as the foundation of artificial intelligence.
The common tasks that we do everyday, such as talking, thinking, reasoning, playing games, recognising people and perception, were difficult for computers to perform, initially. Computers are good at mundane tasks, like performing arithmetic calculations. They perform such tasks for which they are programmed. They cannot reason on their own—a capability only possessed by intelligent beings. But now it has been shown that machines can be simulated to exhibit intelligence by mimicking the way the human intelligence works.
For creating and simulating intelligence, we need knowledge-gathering and representation, learning, planning, perception, and problem-solving. Of these, knowledge-representation and learning are the two most heavily investigated areas in the last decade. It involves learning from huge quantities of data, storing the knowledge in representative form, and using it to create an understanding of the real world. This sub-area of artificial intelligence is called machine learning. There are a lot of potential applications where it can be utilised. Imagine feeding into a computer enormous quantities of stock market data. The computer can learn stock market trends and predict market values of shares in advance. This area is called computational finance—a major utility of investment firms.
Language translation is another application. Almost all language translators in the market use this technology. Speech recognition engines identify words spoken by users, which are utilised in voice-based internet searches and commands. This has been used in voice-based artificial intelligence systems like Microsoft Cortana, Google Now and Apple Siri. These are intelligent personal assistant systems.
IBM’s Watson is a question-answering system, which was developed for competing in the quiz show Jeopardy. This system consistently outperformed human users in the quiz, indicating the power of artificial intelligence systems. In 1997, IBM’s Deep Blue defeated world chess champion Garry Kasparov. From that point, artificial intelligence researchers started to believe that it can not only mimic human intelligence, but also surpass it.
In 2014, Facebook’s DeepFace surpassed human capabilities in facial recognition, using a convolutional neural network algorithm, which mimics the working of human visual cortex—the part of the brain which humans use for vision and perception.
Silicon Valley giants Google, Facebook, Apple, Microsoft, IBM are all investing hugely into artificial intelligence. This has led to a demand for artificial intelligence skills in the job market. Working and succeeding in this field requires basic knowledge of computer science as well as specialised skills in artificial intelligence, logic, probability and statistics. In fact, universities are introducing courses on artificial intelligence not only in their advanced degree programmes, but also in undergraduate studies. We are also seeing a trend where organisations are hiring students who are good in artificial intelligence and related skills.
From targeted advertisements to computer gaming, from finding shortest routes for cabs in crowded cities to predicting stock market behaviour, artificial intelligence is everywhere. It is not too ambitious to predict that this technology will take on the most challenging problems we will face in the future and help us create a better future for ourselves.
The author is a PhD researcher, Centre for Artificial Intelligence, IIIT-Delhi. Views are personal