Thanks to the likes of Google, Amazon, and Facebook, the terms artificial intelligence (AI) and machine learning have become much more widespread than ever before. They are often used interchangeably and promise all sorts from smarter home appliances to robots taking our jobs.
Read more: The UK has a new AI centre – so when robots kill, we know who to blame
But while AI and machine learning are very much related, they are not quite the same thing.
AI is a branch of computer science attempting to build machines capable of intelligent behaviour, while Stanford University defines machine learning as “the science of getting computers to act without being explicitly programmed”. You need AI researchers to build the smart machines, but you need machine learning experts to make them truly intelligent.
Big technology players such as Google and Nvidia are currently working on developing this machine learning; desperately pushing computers to learn the way a human would in order to progress what many are calling the next revolution in technology – machines that 'think' like humans.
Over 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. But how does it work?
Let’s take a very simplified example. When you make a typo, for instance, while searching in Google, it gives you the message: "Did you mean..."? This is the result of one of Google's machine learning algorithms; a system that detects what searches you make a couple seconds after making a certain search.
For example, suppose you were searching for 'WIRED' on Google but accidentally typed 'Wored'. After the search, you'd probably realise you typed it wrong and you'd go back and search for 'WIRED' a couple of seconds later. Google’s algorithm recognises that you searched for something a couple of seconds after searching something else, and it keeps this in mind for future users who make a similar typing mistake. As a result, Google 'learns' to correct it for you.
While this is a very basic example, data scientists, developers, and researchers are using much more complex methods of machine learning to gain insights previously out of reach. Programs that learn from experience are helping them discover how the human genome works, understand consumer behaviour to a degree never before possible and build systems for purchase recommendations, image recognition, and fraud prevention, among other uses.
So now you have a basic idea of what machine learning is, how is it different to that of AI? We spoke to Intel’s Nidhi Chappell, head of machine learning to clear this up.
“AI is basically the intelligence – how we make machines intelligent, while machine learning is the implementation of the compute methods that support it. The way I think of it is: AI is the science and machine learning is the algorithms that make the machines smarter.
“So the enabler for AI is machine learning,” she added.
Chappell went on to explain that machine learning is the fastest growing part of AI, so that’s why we are seeing a lot of conversations around this lately. Even though it’s a small percentage of the workloads in computing today, it’s the fastest growing area, so that’s why everyone is honing in on that.
"Simple examples are when you go to a new place and search online for ‘top things to do’, the order you see them in is defined by machine learning, and how they are ranked and rated, this is all machine learning,” Chappell said, adding that it’s the same story for when news is trending.
Read more: DeepMind's AI learned to ride the London Underground using human-like reason and memory
“AI has become so pervasive in our lives we don’t come to recognise that it’s powering a lot of things,” she added. “You probably use it dozens of times a day without knowing it.”
Elsewhere, Facebook is attempting to demystify the concepts in a series of videos and blog posts.
"Your smartphone, house, bank, and car already use AI on a daily basis," explained Facebook engineering leads Yann LeCun and Joaquin Quiñonero Candela. "Sometimes it’s obvious, like when you ask Siri to get you directions to the nearest gas station, or Facebook suggests a friend for you to tag in an image you posted online. Sometimes less so, like when you use your Amazon Echo to make an unusual purchase on your credit card and don’t get a fraud alert from your bank.
"AI is going to bring major shifts in society through developments in self-driving cars, medical image analysis, better medical diagnosis, and personalised medicine. And it will also be the backbone of many of the most innovative apps and services of tomorrow."
The pair continued that AI isn't magic, it's just maths - albeit really hard maths.
Reinforcement learning: This type of learning concentrates on how an AI 'agent' should behave in order to get the most out of its work. The machine picks an action or a sequence of actions, and gets a reward. This is used when teaching machines to play and win games but needs a large number of trials to learn even simple tasks.
Supervised learning: This is when researchers tell the machine what the correct answer is for a particular input. For example, they show it an image of a car and tell it the correct answer is “car.” It is the most common technique for training neural networks and other machine learning architectures.
Unsupervised learning/predictive learning: Humans and animals learn, typically, in an unsupervised manner by watching how the world works and by observing our parents. However, no-one is there to tell us the name and function of every object we perceive so we have to teach ourselves basic concepts such as: the world is three-dimensional, objects don't disappear spontaneously and objects that are not supported fall. Researchers don't know how to do this with machines at the moment, at least not at the level that humans and animals can.
But in order for AI to progress, machine learning must make big jumps in terms of performance, and this is rarely possible in the traditional high-performance computing world, where problems are well-defined and optimisation work has already been happening for many years.
Machine learning algorithms still have room for improvement, and that’s why a lot of the large technology companies are making it a central focus to their strategy, and working tirelessly to make it more intelligent, in order to push forward and create the next innovation, such as completely autonomous and 100 per cent safe self-driving cars.
This article was originally published by WIRED UK
