Technologies such as Artificial Intelligence and Machine Learning are hard at work blurring the lines that divide the fictional world from the real one. The technology that powers fictional superhero Iron Man’s, aka Tony Stark’s, magic suit to protect the world is also powering the real-life autonomous cars and the eponymous Siri. The idea of everyday objects being interconnected to each other to become smarter, more intelligent and responsive to improve our life quality is a reality that we are moving closer to every day. Our search engines are becoming more intelligent owing to these technologies as they work towards achieving an ‘understanding’ of context. It is because of the tremendous value that they can deliver, the adoption of technologies such as AI and Machine Learning is becoming more mainstream in commercial and industrial applications in areas such as healthcare, finance, aviation, defense etc. Today, over 50% of enterprise IT organizations are experimenting with AI and its concepts to use the huge volumes of data at their disposal for actionable insights. As Big Data and IoT adoption grow, we can expect that by 2022 the AI and Machine Learning market will cross USD $40 billion.
Having said this, most of us are liberal enough and end up using the terms Machine Learning and Artificial Intelligence interchangeably. After all, in our defense, we say that both these technologies are about manipulating the huge volumes of data and use this data to draw intelligent conclusions. It is true that both these concepts are used in context with business intelligence, big data, and analytics. However, it is important to note that Artificial Intelligence is a much larger entity and puts the outcomes from Machine Intelligence to use. Machine Learning, on the other hand, utilizes big data and uses it to create new models whose outputs can be leveraged and consumed by Artificial Intelligence initiatives. To put is quite simply, Artificial Intelligence is the umbrella that encompasses Machine Learning.
Lynne Parker, director of the division of Information and Intelligent Systems for the National Science Foundation, calls Artificial Intelligence “a broad set of methods, algorithms, and technologies that make software ‘smart’ in a way that may seem human-like to an outside observer.” Concepts of Machine Learning, Neural Networks, natural language processing, robotics etc. thus become a part of the Artificial Intelligence universe. Machine Learning, refers to a vast variety of algorithms and methodologies that allow software to improve performance over time as it obtains more data by recognizing trends from the data or by identifying the categories that the data fits in to make accurate predictions. Instead of hand-coding software routines with specific instructions, in Machine Learning, the machine is trained to use these vast data sets using algorithmic approaches.
The focus of Artificial Intelligence is to empower and enable machines with so much intelligent data that they become capable of coming up with solutions to problems on their own. To put it quite simply, Artificial Intelligence aims to ‘humanize’ machines by making them so smart that they would, by conversation, convince a human that they are not robots. While Artificial Intelligence is yet to reach the capabilities as depicted in movies like Her or Terminator, these systems today are working towards passing the Turing Test to achieve human-level performance in all cognitive tasks. In order to pass this test, the system has to possess natural language processing capabilities to communicate effectively in a known language, knowledge representation capabilities to store the information, and automated reasoning capabilities in order to use the stored information – the aim is to be able to answer the key questions and arrive at successful and new conclusions. Machine Learning, on the other hand, offers the data necessary for a machine to learn and adapt when exposed to new data. It generalizes information from large data sets. It detects and infers patterns which can be applied to come up with new solutions and actions.
Machine Learning demands the use of inferential statistics to develop self-learning algorithms while Artificial Intelligence aims to create a ‘system’ that mimics human responses to situations and circumstances. Needless to say, the goal of Artificial Intelligence is to use the data and its associated calculations to come up with solutions of its own.
Despite the science fiction feel that both these technologies have, it is a reality that these are evolving from the sci-fi era to on-the-ground reality. Top enterprises like IBM, Google, Facebook etc. are investing heavily in Artificial Intelligence and Machine Learning projects and are investigating how Artificial Intelligence can be applied to business solutions. The success of Artificial Intelligence and Machine Learning projects like IBM supercomputer Watson predicting patient outcomes more accurately than physicians themselves and the successful application of Machine Learning in real-world applications as demonstrated by Amazon, Netflix etc. are cementing the fact that these technologies, do indeed, have a bright future.