What is Machine Learning?
And does it affect your work as a pharma marketer? Can you separate the hype from the hope? In this article, we give you an overview of the basics of machine learning, and explain how and why it should matter to you.

Put simply: Machine learning is teaching a machine to learn. Okay, we know you need a little more than that. But that’s the truth.

“Machine learning is the science of getting computers to act without being explicitly programmed,” according to Andrew Ng, the former chief scientist of Baidu, China’s answer to Google.

It’s All About Math
The heart of machine learning is math: statistics, probability, algorithms – and do you remember functions in algebra class? Turns out there really was a practical application for them after all. Machine learning teaches a computer to find functions – equations that work not only for the examples that it has, but for unknown ones in the future. Machine learning teaches a computer how to predict.

Ng explains further: “In the past decade, machine learning has given us self-driving cars, practical speech recognition, effective web search, and a vastly improved understanding of the human genome. Machine learning is so pervasive today that you probably use it dozens of times a day without knowing it. Many researchers also think it is the best way to make progress towards human-level AI.”

And so, if the heart of machine learning is math, data is the air it needs to breath. Or, as the Royal Society puts it, “Data is the fuel for machine learning.” Just as human learning relies on examples, machine learning needs the same thing – in much, much larger quantities. Humans can take a few examples and extrapolate from that. Machine learning needs large amounts of data to recognize patterns, test them, and use them to make predictions.

Can It Live Up to the Hype?
Machine learning, consulting group Gartner says, is at the “Peak of Inflated Expectations” in their “Hype Cycle,” a yearly graph that tracks emerging technologies. However, it also says, “Smart machine technologies will be the most disruptive class of technologies over the next 10 years due to radical computational power, near-endless amounts of data, and unprecedented advances in deep neural networks that will allow organizations with smart machine technologies to harness data in order to adapt to new situations and solve problems that no one has encountered previously.”

So what’s the truth? How much does machine learning matter? What will it really change? What do you need to know about it? And how will it affect your specific role?

Current and Potential Use
Machine learning is so common that you probably don’t notice how many times a day it touches your life. You might think of it only in cutting-edge applications like self-driving cars; but from your spam filter to the spam itself; machine learning is everywhere.

One of the places where machine learning in pharma has the most potential is in drug discovery.

Applications for healthcare are enormously varied, though. Machine learning could increase the chance that a physician’s prescription will be the likeliest treatment to help a patient. It can make it possible to prevent disease outbreaks before they happen. Everything from decoding physicians’ handwritten prescriptions to drawing conclusions from enormous data sets is made more possible with the computing power that backs machine learning.

At the Digital Health Coalition’s Spring Summit this May, panelist Havilah Clarke opined that the biggest current opportunity for AI in pharma is in marketing. She related her experience using natural language programming, a machine-learning-assisted process, to analyze news coverage of a campaign far more rapidly and accurately than her competition.

Machine learning can help answer one of the greatest questions on marketers’ minds: what’s the ROI? It’s often been impossible to parse out the specific impact that a campaign has had. But if it were possible to plot with a high degree of confidence the results that would’ve occurred with or without marketing intervention, solid answers could become more common.

As marketers increasingly personalize messaging, machine learning can also be used to better predict customer behavior and advise as to what messaging will be best received by a given customer in a given moment.

Where Do We Go From Here?
Some are concerned by the potential power of machine learning. One of the most widely cited cases of this is an open letter written in 2015 to urge against the development of “offensive autonomous weapons,” which was signed by thousands of artificial intelligence (AI) researchers and tech royalty, including Stephen Hawking, Steve Wozniak, Noam Chomsky and Elon Musk.

At the same time, others see the potential that machine learning has in improving human learning. One of the humans who played – and lost — against the Google AI in the game of Go last year pointed out that those games had taught him to think better and play better.

With the help of machine learning to calculate, analyze and predict, who can say how our own potential can increase? We may be able to automate manual processes and free ourselves up for creativity and inspiration in unimaginable new ways.

But in the near term, machine learning is making small but deliberate progress in our lives and in our work. We are now accustomed to using voice dictation or facial recognition, getting accurate predictions for the weather and what we’re likely to buy next. Machine learning is working its way into our profession as well, and if we remain alert to its potential, we can use it to improve how rapidly and accurately we’re able to convey information to patients, physicians, caregivers, payers and other stakeholders.