Machine Learning Trends In 2016 And Beyond: How They'll Affect Your Life
It's no secret that machine learning, sometimes incorrectly called artificial intelligence, has had a major impact on the way consumers interact with the world. Banks and credit card companies employ "smart" algorithms to detect fraud, Amazon uses them to recommend products you "might also like," traffic control systems use them to speed the flow of cars during rush hour - and Siri depends on machine learning to be able to understand your requests and respond to them properly.
The use of these machine learning algorithms is expanding rapidly throughout every type of business, and their power to boil down data to predict consumer needs and behaviors is growing by the year. Here are a few trends to expect.
Expect Better E-Commerce Search Results Than Ever
Current algorithms being used on e-commerce sites depend primarily on the simple query you type into the search box, like "blue widgets" or "hotels in Lexington." But retailers are beginning to factor other information into the results they return, thanks to advanced machine learning.
The website Data-Informed.com describes one example. Chain retailer The North Face has combined its own system with IBM's natural-language learning system, Watson, to create an amazing personal shopping experience. The example the company uses to demonstrate the power of this approach is a shopper who types in a phrase like "I'm staying in a cabin in Iceland this December." The ML system combines data from its catalog of products, a list of items needed for a cabin stay, and information on Icelandic weather in December to produce a short list of suggested purchases more personalized than any competitor can offer.
Expect to see this approach become common across the web in the next few years.
Would You Mind Answering This Short Survey?
Most of the online interactions driven by machine learning are a linear process. The user makes a request, the machine learns a little about the user and returns results it believes will be a good match. Another request, another slightly "smarter" response, and so on.
Expect online retailers to begin implementing a short quiz, administered when you visit their site but before you start shopping. The answers to a few questions will give the machine much more data to go on, so it can provide results or show products tailed explicitly to your specific needs and preferences.
Siri Will Get Even Smarter
The extent to which machine learning is used by Apple's devices wasn't fully understood until this month, when BACKCHANNEL revealed how dependent the iPhone has become on ML algorithms. You may not understand how your phone was able to identify a caller for you when they're not in your contact list, or how a Maps location miraculously appeared for your next appointment without your asking for it, but it was because of machine learning.
Siri has gotten much better at interpreting your requests in the last two years; you don't have to use certain phrases or speak slowly anymore, and that's largely due to machine learning advances. So what's next? An even smarter personal assistant - with a more human voice, again made possible by artificial intelligence. Apple says tests have shown that a more personable voice makes Siri more trustworthy and improves user willingness to use their iphone even more.
Machine learning and artificial intelligence may sound a bit scary, considering the way it's often portrayed in sci-fi horror flicks. But the next time you have a satisfying experience online or with your smartphone, remember that it's largely possible thanks to machine learning.