What Is AI?

Michael Notter, Christian Luebbe & Catherine Dietrich @ EPFL Extension School · 12 minutes

What Is Artificial Intelligence & Why Is Everybody Talking About It?

You live in the age of intelligent machines!

Artificial intelligence (AI) is already here, and it’s here to stay. Every day, new breakthroughs are made in fields like bioscience, medicine, and robotics that are attributable to AI. And these breakthroughs are occurring with increasing regularity as AI gets stronger and stronger.

In some areas, AI has already left us behind. For example, there are no board games left where we can outperform AI. Only a decade ago, people doubted whether artificial intelligence would ever be able to defeat a professional player at Go, a game infinitely more complicated than chess. And then, in 2016, the world’s number one Go player, Lee Sedol, lost to a computer program.1 That’s just one example of how AI continues to surpass our expectations and impress us with new achievements.

Our culture is also fascinated by AI’s possibilities, and movies and TV shows often explore what it might mean for the future of the human species. But what was once science fiction is very quickly becoming science fact. AI is giving birth to completely new industries, creating new jobs and attracting huge investments. It also promises to transform how we work and how we live our lives. So, why don’t we understand it better?

What Is AI?

AI describes the broad discipline of developing machines that can emulate human intelligence by perceiving, learning, reasoning, planning, and solving problems.

More exactly, AI refers to the technologies and algorithms that are used to train systems to perform these intellectual tasks by themselves.

This quickly raises a lot of questions. What makes AI “intelligent”? How “smart” is it? How does it “learn” things? How competent and trustworthy is it? How powerful will it be in the future? Most importantly, what will it mean for humanity as a whole? What are the philosophical, ethical, and social implications of AI? What happens when machines can think like us?

Let’s start with a few basic questions:

  1. What role does AI already play in our lives?
  2. Why is there a widespread mistrust of AI? Is it justified? And how can we combat this?
  3. Why should we inform ourselves about AI and its transformational effects on our lives?

It is just a regular AI kind of day.

You might not realize it, but AI is already a big part of your everyday life. AI unlocks your smartphone when it recognizes your face or fingerprint in the morning. You use another AI as you open your weather app to see if it’s going to rain or not. As you make for the car with your umbrella, you use another AI to plan your quickest route through the traffic. By this point in the day, you have already used three AIs and you’ve only just left the house.

Detecting your face, guiding you through the city or predicting the weather. AI is already all around us.

And it doesn’t stop there. At work, AI has screened your email for spam and organized your inbox into folders. AI suggests and spell checks your replies. You pay for lunch with a banking app that uses AI to protect you from fraud. Back in the office, AI continues to support you through the working day by streamlining and optimizing manual tasks.

You arrive home after a busy day and check your social media accounts. They’ve already been pre-filtered with AI to suit your tastes and behaviors. So have your suggested films and TV shows when you’re deciding what sort of entertainment you’d like to stream before bed. Finally, AI switches off the bedroom light with a simple voice command.

Why do so many of us distrust AI?

Even though most of us interact with AI multiple times every day, there is still widespread unease and distrust of the technology. AI has the potential to change our lives in profound and dramatic ways. That’s unsettling, especially for those of us who do not really understand it. There’s also the question of whether AI will be used in negative or harmful ways. Reasons people give for being fearful of AI include:

We have concern for the future of jobs.

AI systems are much better than humans at laborious and repetitive tasks, and the automation of more and more manual work will have consequences for workplaces and employment opportunities. Which jobs are safe from automation? What are we going to do when the machines are doing everything for us?

We fear of losing control.

The assumption that AI generated outcomes will always be positive should make all of us uncomfortable. We should ask ourselves what happens if an AI makes a “bad” decision and we can’t explain it? What happens when human intelligence is only second best? How do we make decisions and where do we place our trust?

What if AI is super addictive?

We can already see addictive behaviors in the way humans use technology, with gaming and social media being two obvious examples. What happens to us if AI introduces a level of sophistication to the digital experience that makes it even harder to switch off our devices? What does it mean for our children?

Monopolies in the Industry

The development of important AI products requires a lot of funding, big data, advanced infrastructure, and expertise. That means only a handful of big companies have the necessary resources. Can we trust a relatively small group of people to develop technology that will transform the world? Where is the regulation to protect the public interest?

What makes us human anyway?

As AI continues to develop and outperform humans, we will be forced to confront deep philosophical questions. What makes us who we are? What is our role in a world where we are always second best? How will we distinguish ourselves from the machines?

These are valid concerns that need to be addressed - which is the goal of That's AI. By unpacking them we can start to ask ourselves why there is so much mistrust of AI. Is it because we are afraid of the technology itself? Or are we more worried about the potential for human beings to misuse it? It’s frequently said that AI is only as good as the humans behind it. As AI learns from us, it takes on our biases and prejudices, as well as our abilities.

AI can be liberating when it’s used with good intentions, but it can also exploit us if it’s misused. For this reason, it’s essential that we demystify AI and the uncertainties around it. With greater understanding comes greater trust, and from there we can begin to explore the enormous potential for opportunity and innovation that AI presents to us.

Why is it important to understand AI?

It’s hard to overstate just how all-pervasive and transformational AI will be in the coming years. But while AI promises to deliver impressive achievements and discoveries, history shows us that new technologies also introduce unintended consequences and fresh challenges. Educating ourselves about AI gives us the opportunity to pre-empt some of the problems we will need to solve and prepare for this uncertain future.

By automating specific tasks, AI gives us more time to concentrate on challenges that require human ingenuity. This means that new jobs will be created, some will be redefined, and some will disappear altogether. So while AI will solve a lot of problems in our societies, it also threatens to disrupt our economic and social structures. It may also exaggerate existing inequalities, negative human behaviors, and alarming ecological trends. We need to ensure that AI empowers all of us and doesn't end up concentrating power in the hands of only a few individuals.

That starts with educating ourselves about AI because this technology will soon affect every aspect of our daily lives. Everyone should have some basic understanding of how AI works. We don’t all need to become AI engineers, but we do need to know enough to meet the challenges that AI will bring, as well as seize its opportunities. A common understanding will also help us to share the benefits of AI and avoid negative outcomes.

It is a world of infinite possibilities.

AI will introduce all sorts of dilemmas and moral questions into the public conversation. It will challenge how we see the world – and each other – in ways that are impossible to foresee. But we already know enough to come up with a few difficult questions. Try the following thought experiments and get a sense of the world we’re about to enter into.

  • There is a car accident in which two self-driving cars are involved. Who is responsible? How do we decide?
  • You’re due to undergo heart surgery. The hospital gives you a choice: would you like the human surgeon with 30 years of experience, or the highly skilled machine trained on 1,000 years’ worth of simulations covering every conceivable scenario?
  • If / when machines are able to simulate human intelligence, will there still be a distinction between humans and machines? Will robots be entitled to the same legal protections? What status will they have alongside us?
"Your AI-car clearly didn't look ahead!" - "Well, your AI-car shouldn't have been driving like that."

As you can see, it’s not difficult to pose some very challenging questions about how AI will fundamentally change our world – and we’ve only just scratched the surface. There will be many important conversations to be had about AI and its implications for society. It’s vital that we all gain a basic understanding so that we can be meaningful participants in those discussions.

AI means different things to different people.

We also need to keep in mind that artificial intelligence has no universally agreed definition. Scientists, media outlets, corporations, and the general public each have their own ideas and assumptions when it comes to AI. No two groups see it in quite the same way.

For example, some people use phrases like “the AI” or “an AI” when what they’re actually referring to is a system, machine, or service that’s built using AI technology. Other people describe AI in human terms by attaching character traits to it, like “opinionated” or “stubborn”. And for some people, AI is used simply as a catch-all term for any complex system they can’t comprehend. For all of these reasons and more, our general definition of AI is very, very broad.

The problem with this is that our inability to talk about AI in a shared language means we frequently misuse the term. News articles, marketing campaigns, TV and films add to the confusion by using AI as an easy headline or plot point in a story. Without a common understanding of what AI is and does, it’s almost impossible to have meaningful conversations about it. Building that understanding is a really important task.

This is why we have built the “That’s AI” platform.

The aim of That's AI is to demystify artificial intelligence by addressing these questions in ways everyone can access and understand. On the platform, you will find content that is designed to educate and inform you on the subject of AI – and you don’t need to be an IT expert to get started.

This technology is too important to be left to a handful of computer scientists and engineers. It will have far-reaching consequences for all of us, and so we need to develop a shared understanding to work through them. We’ve built That's AI so you can join us on the AI journey!

  1. For more on the story of how AlphaGo won against Lee Sedol, check out the corresponding Wikipedia article or the AI developers’ homepage


AI Is All Around Us