An introduction to machine learning

If you had told technology experts 20 years ago what we could do with machine learning today, they would probably be skeptical, to say the least. The term was first coined by Arthur Samuel, an American pioneer of artificial intelligence research in the 1950’s. So why, after all these decades, has it gained fresh momentum only in the last few years? The easy answer is that both the data storage and the data processing capacities have grown tremendously, to the point where it is now profitable for businesses to use machine learning. A smartphone in your pocket now has more storage and compute power than a mainframe in the 80's.

The ability to automatically apply complex mathematical calculations to big data – over and over, faster and faster is only just a recent development. The technocrats of Silicon Valley can thank the cloud for the leaps and bounds machine learning have made in recent times, cloud hosting platforms, for instance, Microsoft Azure have been the catalyst for the next wave of tech start-ups like Graphcore becoming a member of the holy grail of all clubs. The unicorn club that is.

An IBM mainframe in the 1980's.

Machine learning is unquestionably the latest buzzword in the tech landscape as it is one of the most interesting and promising subfields of computer science. ML is a new programming paradigm, a new way of communicating your wishes to a computer. To put it in perspective, let’s say that you want a machine learning program to identify a comic book without your intervention. In a traditional algorithm-based approach, you would code this program, or you may perhaps create an algorithm using data about the structure and style of a typical comic book. However, in a machine learning-based approach, your algorithm will learn to identify comic books all by itself by learning the style and structure of a typical comic book; simply by using data of a million comic books, you would feed to the algorithm.

While traditional algorithms are created and perform to produce the desired result, machine learning algorithms are created with the intent of having the program “learn” how to produce the desired result with minimal human intervention. Machine learning is not magic. It is simply part of the ever-evolving analytical landscape that is laying the foundation for a technocracy society where technological solutionism becomes the answer to almost every problem.

A baby learns to crawl, walk and then run. We are in the crawling stages when it comes to applying machine learning. Its true potential is still unknown since we’ve yet to cross the threshold. The complexity and size of our data sets are growing faster than our computing resources and therefore place’s a considerable strain on our computing fabric. Although the field of quantum computing is in its infancy stages the monumental impact it could have on the development of machine learning and other subsets of artificial intelligence is unprecedented.

Google with its infinite resources have developed a quantum computer they claim is 100 million times faster than any of today’s systems. The application of such tools will drastically improve our machine learning and artificial intelligence capabilities by uncovering patterns or anomalies extremely quickly. Quantum computing is a promising emerging technology and its full potential will only be realised when it starts to become commercialised. I highly recommend watching the age of AI, the 8-part YouTube documentary series hosted by Robert Downey Jr, Will give you tremendous insights in ways artificial intelligence and its subsets are changing the word today. 

Sundar Pichai, pictured with the Sycamore Quantum processor