How machine learning will change society

The potential for machine learning is huge, and no one knows that more than Danny Lange, VP of Machine Learning at the US-based games engine Unity.

Having previously worked for Uber and Amazon, Lang believes that future uses of machine learning will stretch across all sectors, disrupting most of what we know.

"If we take a step back then the real disruption in the machine learning technology sector, is the disruption itself," explains Lange in an interview with Techworld at Web Summit. "This is because we are going from programming computers to training computers."

Read on for five ways machine learning will impact society...



We already know that through algorithms, small and hard to spot abnormalities can be discovered in all kinds of things, from legal documents to financial papers. So it's fitting that the same can be said for healthcare.

Spotting irregularities in test results is just the start.

"Anything that has to do with specialists, such as doctors that need to learn a lot of things, computers will be able to learn that as well," Lange says.

Adding machine learning into historic sectors will have to keep humans at the centre, Lange says. It won't take over, but will co-exist.

"In the grand scheme of things, our focus as humans will be much more on the human aspect. For example, if you're a patient, you're not going to interact with a machine."



The transport industry as a whole is slowly adapting to new emerging technology. We're starting to see more sensors on roads and rail to monitor degradation and smarter tech being used more than ever before.

With machine learning, you're able to expose vehicles to millions of potential scenarios and make sure the computer in the car, bus or truck acts in a certain way.

"If we look at self-driving cars, if the computer within the car can learn as we know it can to drive that car around that's a massive step forward," explains Lange.

"The reason we've not been able to do this until now is because we've had lots of software engineers, but we need to programme it to handle all kinds of situations and that's where machine learning has come into it."

"I think transportation is a big one, because transportation will be made much more efficient, not just by the creation of self-driving cars however.

"We’re already feeling it by better routing and better sharing of resources."



The use of VR and AR in education has already been proven, with student surgeons able to practice life-like medical procedures and children able to explore countries without leaving the classroom.

"Education is going to be big, says Lange. "We've not got a system that can learn how to teach you on a very individual basis, rather than one size fits all."

While the classroom teacher will still exist, the ability to quickly curate lessons for struggling individuals is possible with machine learning, he explains.



Retail will benefit hugely from data and the meaningful insights and actions machine learning can throw up.

It's a sector crying out to be disrupted and machine learning could do just that.

"I want to emphasis retail, an area which people think is going to die. But I don't think that. I think that retail is going to be able to use these technologies such as facial recognition and provide a personalised experience in store.

"If they knew what you looked like, from last time you visited, they can start showing you things that they know that you're interested in."



Within the entertainment industry new technology can emerge quite quickly, from new lighting and sound techniques to all out CGI advancements.

If machine learning is introduced the experience as a consumer will totally change.

"There's not going to be a movie with a fixed sequence of interactions. It's going to be a very different experience.

"You're going to be able to impact what you watch. You'll be able to be a part of it."

Copyright © 2017 IDG Communications, Inc.