A technology has invaded your life without you knowing
I’ll go as far to say that it would be difficult, and noticeable, to go a day without machine learning.
It’s hard given our daily routines, to recognize how different our lives are today than ten years ago. I remember being told in math class how the calculator I had was more powerful than what astronauts took to the moon. Now the smart phone I’m using to type this up, is miles more powerful than that, and able to keep up with a super computer from the not too distant past, Deep Blue. (The computer that beat chess grandmaster Garry Kasparov)
The way we connect with one another, the way we plan our lives, or even days, is so different than any other time in history. All of this change has come thanks to the rapid improvement of technology, that has entered all aspects of our life.
Technology catalyzes change and it comes in waves. Putting it simply; you have your science fiction authors that will dream something up, your researchers that test and retest to create something novel, your early adopters who take the first plunge, then years / decades later you have the mainstream benefit.
The printing press took generations before it increased literacy, but once it did it catalyzed the Renaissance, which lead to the Enlightenment, which lead to the birth of our modern government. My iPhone however, took less than ten years to go from early adopters to an application revolution. That mobility ended the cycle of capabilities from a desktop, to a backpack, to now, where I can do most of my tasks from something that slides into my pocket.
This same cycle is playing out right in front of us with machine learning- and I’ll go as far to say that it would be difficult, and noticeable, to go a day without machine learning. Yet, I doubt you even give it much thought.
Let’s say you’re new to Amazon. Just got your prime membership and you’re quickly hooked. You do your normal shopping at Target, but instead of buying anything there on impulse, you start to look up the price on Amazon. Like me, you start to learn the patience, mixed with a Christmas Day feeling, when your ordinary goodies arrive two days later.
Quickly and freakishly, you start to notice recommendations that not only catch your eye, but start to seemingly read your mind. “I just ran out of _insert item here_, how did it know?” Or even the classic, “I was just talking with _x_ about _y_, is my phone listening to my conversations?!”
No, it’s *probably* not. But what it is doing, is leveraging machine learning, against on a pretty good picture of who you are. What kind of habits you have, how you like to spend your time, your guilty pleasures, hell it probably even knows your diet. (I’m also particularly fond of when a machine learning algorithm for Target that knew someone was pregnant before she did)
We see machine learning all around us, but like a fish swimming along, sometimes we forget that we’re in water. The first industry to really adopt and exploit machine learning is the ad industry. There’s a lot of press right now about what platforms do with your data and who they sell that data to for advertisements (rightfully so). What’s often missed though is the technology behind it.
We see machine learning all around us, but like being a fish swimming along, sometimes we forget that we’re in water.
If a company can build a robust profile from your buying and browsing habits, so robust that it can guess that your pregnant before you even know, what could the benefits be for other use cases? How might we be better able to use medical data to find illnesses? Or spot trends that lead to quicker and more accurate diagnoses?
Taking the Amazon example I gave earlier and apply it instead to a medical use case, I’ll dive a bit deeper. You just downloaded a new application that your doctor suggested you use to help monitor your health.
You accept the terms and conditions of the application which securely tracks lots of data on you; phone usage, app usage, location tracking, etc. Moving past the obvious and concerning big brother problems with such an application, let’s try and dream up what the benefit could be.
The application runs daily and from time to time asks you questions about how you’re feeling. Mood, pain levels, outlook and more. This feedback is a necessary piece that the program uses to build a better picture of you. From advertising and shopping, that feedback comes from your purchases and taking actions in the app, such as liking a post.
So you’ve had the app for a few weeks and you go back to your doctor. The two of you sit down, talk about how you’re feeling, then review what insights the app has generated. It notices that when you spend more time on social media your mood decreases. When you sleep less your back pain increases. Easy stuff first.
Then it starts diving a bit deeper. It starts presenting literature to both you and your doctor that relates to a stomach issue that you’ve been struggling with. It shows how people with your lifestyle, age and diet better manage it. Making it easier for the three of you (patient, doctor and application) to create a detailed health plan, where your doctor can now check in on your progress in a more informed manner.
Imagine the possibilities for mobile healthcare. You’d have the potential to converse with your doctor about your health, from anywhere, with more information about your well being than he would have otherwise.
This doesn’t just end here. What if your insulin pump connected to this same application? Or if you used your camera as a way of monitoring your pulse and heart rate? Enter in your temperature, connect to a blood pressure monitor, what else could this application find?
None of this is science fiction, all of this is possible today, right now. Devices like a blood pressure monitors or the like aren’t just becoming cheaper, they’re becoming smarter. Why not take all of that data and help create an application that makes the patient and doctor smarter?
An app like this can give us an opportunity to learn the best way possible, through context. I’m sure we all struggle fully being aware of what habits we should change or how to create new ones. Having a little health assistant along the way can help nudge us towards healthier behavior.
Keeping us honest with our habits, encouraging us to exercise or eat right (this could even provide us with recipes based on the diet my nutritionist and I agreed on). Layering on top of that with articles better informing us about our health and habits. If done right this could be a breakthrough in healthcare.
The biggest barriers to this becoming a reality, other than investment into it, are security and thoughtful design. If the user doesn’t trust that every detail of their life being catalogued somewhere won’t get stolen or misused for someone else’s profit, then it’s benefit will be quickly go up in smoke.
Same with design. Navigating through it, the tone of voice when giving you insights or asking for feedback- when and how frequently it does this too. Are there any moments of delight? Not just a slick interface with some nice animations (which is important), but something that makes me feel delighted with my health. Gamification is of course one way of going about this, but is there something more personal that we can design for?
This all paints just one picture, other applications of this can include diagnostic devices running analysis on their patient data de-identified to the individual. An application like that could spot patterns in diseases, health issues or provide more insight into disease clusters.
And to think, we’re only getting started.
If you liked this post please leave me a comment or a few claps! If you have any questions, or are looking to build your own machine learning app, you can find me over at LinkedIn, or at our website 3ct.co.