What is AI? Defining a concept
Valérie Bécaert Valérie Bécaert
November 12 4 min

What is AI? Defining a concept

Artificial intelligence is everywhere: in our collective imagination and increasingly in the world around us. There are many vague ideas floating around about what AI means — “smart computers”, “artificial life” — but for a meaningful understanding of all the things AI can do for your business, it’s worth stopping to fully consider - what is AI?

The ability to reason

Originally defined by pioneering computer scientist John McCarthy, the term “artificial intelligence” refers to "the science and engineering of making intelligent machines." But what do we mean by “intelligent?”

Intelligence is usually defined and measured by the capacity to acquire and process knowledge, using it to “understand” concepts and make correct “judgments”. These abilities to learn and make decisions are two of the key attributes that make AI useful to us.

Learning machines

An AI learns by being provided with data, which it processes and uses as the basis for its decisions. That data becomes part of a store of learnings that it draws on in the future.

When a machine learns, it can make decisions. It can analyze data, categorize it, and make recommendations based on what it already knows. And the more an AI system knows, the better decisions the system can make.

When a machine is able to make decisions autonomously with a good enough degree of ever-improving judgment, it can perform tasks for you; quickly, precisely, and tirelessly. That means AI systems can take some of the strain off human workers when it comes to the repetitive tasks they’d sometimes sooner avoid.

A helpful worker

Popular culture sometimes paints an AI system as a person or a creature. It’s more of a tool, or perhaps an obedient robot. But it’s an extremely sophisticated tool that can help with many of our manual and cognitive tasks.

Give an AI system information and it can sort it, analyze it, and act on it – for instance, judging insurance claims in bulk and determining the appropriate payouts, or detecting possible instances of fraud. It can process these claims quickly and accurately, taking a large workload of painstaking, time-consuming tasks off the insurance underwriter’s desk.

AI systems can also monitor industrial machinery, analyze its performance, and detect a glitch or an impending breakdown — or even predict when a failure will occur based on historic equipment lifecycle data. That means maintenance personnel are burdened with fewer routine checks and can dedicate more time to higher-level work such as finding new ways to boost production output throughout the facility.

Perhaps these examples sound futuristic, but they’re only two of many readily-implemented examples of how AI systems are capable of assisting in a task, increasing efficiency and optimizing organizations’ processes.

Digital interactions with the physical world

As AI systems progress, they’re able to interact with our world in new, more useful ways than ever before. Machines now have the processing power, code complexity, and sensory sophistication to examine this video, audio and other sensor data in real time, quickly and deeply.

Just as an airplane’s autopilot makes the pilot’s life easier, AI systems offer the chance to help drivers on their journeys. Autonomous vehicles can use visual data to analyze their surroundings and navigate, using GPS and proximity sensor data to do so with precision. And an AI system can perpetually learn from its experiences on the road and improve, making each journey with greater efficiency and safety.

More data, more possibilities

Thanks to the Internet of Things (IoT), an ever-expanding network of smart devices throughout the world, today’s AI systems can potentially access a larger amount of data than ever before — and be even more useful to mankind.

This gives us the opportunity to analyze, understand and change our world in ways never possible before. Climate sensors thousands of miles apart send back data to be compared. Carbon emissions from Beijing to Montreal to Addis Ababa are recorded and analyzed simultaneously.

Now, valuable data is created every single time a transaction is made, or when goods are transported, or a product is manufactured. With the IoT there to collect it, it’s instantly added to the huge mass of information that AI systems can learn from.

Imagine the potential insights we can gain from ever-smarter AI systems with a wider, deeper view of our world. Our digital helpers’ usefulness will only grow and grow.