Of all the technological advances mankind has achieved over the millennia that have changed the way we live and work, Machine Learning (ML) may well prove to be the one with the most powerful and disruptive effects. And while one application – autonomous vehicles – is still in its infancy, ML applications in many other business applications are already here and more are on the way.

ML grew out of the computer science of artificial intelligence (AI). Broadly speaking, AI refers to computers programmed to perform tasks that normally require human intelligence, like recognizing images and speech and making decisions. As remarkable as that is, ML goes one step further: it creates computers that can learn from experience and actually improve their performance.

Another way of putting this has to do with the way ML tames Big Data. We’re all familiar with the Big Data challenge: the Internet and a proliferating number of sensors in everyday items are generating massive amounts of data that exceed our ability to extract actionable information. ML programs enable computers to sort through immense data streams at lightning speed and find meaningful patterns that they weren’t pre-programmed to recognize. In turn, ML programs can use those patterns to make predictions of future events.

Here’s how Josefin Rosén, an ML and AI expert, describes it:

Machine learning enables computers to find hidden insights without being explicitly programmed where to look. They can change their behavior and improve their algorithms by themselves, and every time an error is made and recognized, the algorithm corrects itself and begins another iteration of learning.

The first ML application was created in 1952 by a pioneer in AI and computer gaming at IBM named Arthur Samuel. But as an independent discipline, machine learning didn’t begin to flourish until the 1990s, when the field changed its goal from achieving artificial intelligence to creating computers than can independently solve problems of a practical nature.

Now, businesses are rushing into the field with billions of dollars being invested. A recent survey of 500 chief information officers across 11 countries found that 89 percent said their organizations are using, piloting or developing strategies to machine learning. And it’s no wonder: just as AI has been used to automate thousands of manual tasks, ML has the potential to automate far more sophisticated functions that rely on data analysis used by business, government, medicine and more. The benefit is expected to be counted in trillions of dollars of increased productivity, improved customer service, creation of new products and marketing strategies, and millions of lives saved and made more enjoyable.

Machine learning applications have already started to seep into our daily lives. We have smart appliances and smart homes, voice-interactive digital assistants that can operate other devices and look up information for us, like restaurants with food we like or where our kinds of movies are playing. More and more service companies are using “chatbots” to interact with customers online, ML-powered digital assistants that respond to keyboard messages.

Behind the scenes, Uber and Lyft rely on ML applications to match drivers and riders and support dynamic pricing, rates that change based on supply and demand for rides by location and time of day. Google and other search engines use ML to improve each user’s search results employing back-end algorithms that watch how far down in the results users click on search items.

ML is also making headway in two industries with which I’m familiar: insurance and fleet.

  • User-based insurance. Drivers shopping for car insurance with Progressive can opt to install a company-provided telematics device that measures how much risk for an accident they pose, based on how often they speed and engage in harsh braking (a sign of distraction or tailgating). Safe drivers are offered premium discounts of up to $150.
  • Real-time repair estimates. Many insurance carriers are experimenting with ML-based technologies to help drivers who’ve had an accident receive real-time repair estimates by taking photos of their vehicle damage with a smartphone camera. The app is being built based on thousands of images of damaged vehicles and their repair costs.
  • Accident reduction. Fleets are beginning to turn an ML-assisted way of identifying high-risk drivers called predictive or prescriptive analytics. These algorithms use records of driver behavior and various demographic and industry data to predict the probability that they’ll be involved in a collision over the following 12 months. Once those that pose the greatest risk are identified, fleets can remediate their poor driving behavior with additional training and coaching.
  • Predictive maintenance. Based on historical data, Programs have been developed that can predict when auto parts or systems are likely to fail, and recommend replacements before those failures occur.

There’s more coming, in every industry. It’s part of what historians are calling the Fourth Industrial Revolution, in which disruptive technologies and trends such as the Internet of Things, robotics, virtual reality and artificial intelligence are changing the way we live and work. As with the previous three revolutions, many jobs will become obsolete and disappear, but – and I’m confident of this – many more will be created, and mankind will be much better off.

Competition in the business world will depend more and more on the extent to which companies harness machine learning and the related digital tools of the Fourth Industrial Revolution. The survivors will be those who master them strategically and master them first.

Recent Articles

  Staying safe on the roads during an emergency evacuation is critical to your safety,..

Read More

When you drive, there are many things that can distract you, but the road you..

Read More

Connected fleet safety isn’t just a fancy term; it’s a big change in how we think about keeping vehicles safe. It started with simple tracking systems but has…

Read More