Say the words ‘Artificial Intelligence’ and it all seems very X Files or Men in Black – certainly not something that is every day at least. But what if you were wrong? After all, Facebook’s founder, Mark Zuckerberg, announced that this year he wants re-create his very own Jarvis from Iron Man – he’s actually going to build his own robot assistant!
However, when it comes to Artificial Intelligence (AI) in IT, it is really very much a ‘thing’ and it can certainly help to benefit your business too, beyond being a virtual butler and making you a cup of tea…
There’s something called ‘Machine Learning’ that is part of the AI family. Essentially it means computers have the ability to learn without being explicitly programmed – no hours of coding required. Rather, machine learning focuses on the development of computer programs that can teach themselves to grow and change when exposed to new data.
Unlike data mining, which searches through data and looking for patterns to present in a meaningful way to be understood by humans, machine learning uses data to improve the program’s own understanding and adjusting the program’s actions accordingly – it’s about patterns, probability and predictions based on what the program has learned through the inputting of historic data.
It means that through looking and learning from volumes of historic data, machine learning produces actionable results. It really is a very clever piece of kit for any business or organisation that needs to consider external factors affecting business decisions – anything from market challenges to customer behaviour.
For example, imagine you’re a business and you sell umbrellas. Through machine learning, by inputting the data from historical weather patterns for the past year, the results may show you that in 2015, December was the wettest month of the year and so may predict that the same will happen this year. However, by ‘feeding’ the programme more historic data going back many years, the results may show you that actually, December isn’t always the rainiest month of the year and that it tends to be more around April/May time instead. In essence, the more info you can give the programme, the more it learns and better predicts the most probable outcomes.
Thinking about the umbrella business, armed with this information, you can use the results from the machine learning exercise to inform your purchasing decisions. The system could, for example, learn what the typical linkage is between the actual weather, the forecast weather and how many people buy your umbrellas. If you know it’s less likely to rain in August, you don’t need to order in as many umbrellas. If you know it’s likely to be a very wet September, you can make sure you have enough stock to meet demand and set your price point accordingly. While some of the results (like this one!) may be intuitive and obvious, this isn’t always the case.
It’s all really impressive stuff and it’s a product offering Amazon Web Services is taking to market. And what’s great is it’s not really expensive either (certainly not as costly as Mr Zuckerberg’s bill for building his own Jarvis!). With Amazon Machine Learning there’s no upfront hardware or software investment, and you pay as you go, so you can start small and scale up if you need to. The predictions are presented in real-time too, so it helps when you have to make a quick business decision with little time to consider the options.
AI and machine learning may sound like the stuff of MI5 and the Secret Service, but the truth is most businesses, regardless of shape, size or sector could benefit from the insight it can offer. It’s an affordable and effective way to stay ahead of the competition and something you might like to consider as part of your IT and big data strategy sooner rather than later.