May 17, 2017
Any article or blog post with the words “Machine Learning” has freaked me out for years because all I can envision is something like this showing up at my front door:
While that day is probably still a few decades away, I have become more and more obsessed with how digital interfaces are getting smarter by learning and serving engaging experiences that are completely suited to a user’s preferences. The days of just publishing content that a copywriter thinks might be relevant are quickly coming to an end.
The above three are a few of the best examples of interfaces learning and accommodating to human preferences. The more you use their services by navigating categories, searching and watching/purchasing products, the more the interface is learning about your personal tastes.
By “learning,” the machine can now slowly increase the likelihood that it’s going to provide content that you’d want to engage with next. It’s a massive win-win for both the user and for the service by reducing the time it takes to reach both parties’ intended goals.
A great example of how machine learning is solving human preferences would be the way my two-year-old son navigates Netflix to find one of maybe four shows he loves to watch. He can navigate this without any help from me or my wife.
He started playing with the interface when he was a little older than one and it didn’t take him long to understand how to get to his specific user account, slightly scrolling up/down or left/right for his favorite or “most recently watched” shows, and then understand that pressing on the album cover will start the show. Netflix has learned that he consistently plays his go-to shows and the service always makes those readily available to him.
If you were to look at my user account, I have quite a bit more variety in my browsing recommendations. I’m not always shown everything that I’ve watched recently at the top of the screen. Along with beautifully designed style patterns within the interface, Netflix understands the preference and viewing differences between humans and accommodates us accordingly.
Programmers and experience architects are creating extremely complex algorithms in the back-end of these services, and the way they write, which accommodates learning and evolving, is so fascinating. In my opinion, this will continue to be the direction in which UX design moves, and it’ll only get more complex with advancements in augmented- and virtual-reality environments.
Users demand relevant content when browsing or engaging with a product, and if your interface isn’t providing recommendations that fit their desired preferences, your product isn’t going to survive long. Easier said than done, I know, but you have to push to satisfy on the highest level, and machine learning is at the forefront of this movement.
I’d love to hear about anything you’re doing within the machine-learning space, so feel free to reach out to me at email@example.com. Cheers.