What is AI, really?

The short answer

AI (artificial intelligence) is computer software that learns from examples, so it can do jobs that used to need a person’s judgment: recognizing a voice, spotting junk email, predicting traffic, or writing a reply.

You don’t need to understand the math to use it, just like you don’t need to understand engines to drive a car.

How is AI different from ordinary software?

Think of two cooks.

• The first cook follows a recipe exactly. Step 1, step 2, step 3. If something isn’t in the recipe, they’re stuck. That’s ordinary software: a calculator, for example, follows exact rules that people wrote, so 2 + 2 is always 4.
• The second cook has tasted thousands of dishes. Give them ingredients they’ve never seen together, and they can make a good guess at what will work. That’s AI: it learns patterns from lots of examples, then uses them on new things.

1People gather lots of examples

2AI learns the patterns in them

3It meets something new

4It makes its best guess

International experts put it this way: the OECD (an international organization of governments) defines an AI system as one that, from the information it receives, works out how to produce predictions, content, recommendations or decisions. ^[1][2] In plain English: it takes something in, and works out a useful guess, suggestion or answer.

You already use AI every day

AI isn’t only chatbots. Chances are you’ve used it today without noticing:

• Your email’s spam folder. Google says its AI-powered tools stop more than 99.9% of spam, phishing and malware from reaching Gmail inboxes, blocking nearly 15 billion unwanted emails every day. ^[4]
• Arrival times in your map app. Google Maps combines past traffic patterns with live conditions and uses machine learning to predict how long your trip will take. ^[3]
• “Hey Siri” on an iPhone. The phone uses a type of AI called a neural network to recognize those two words when you say them. ^[5]
• Chatbots like ChatGPT, Claude, Gemini and Copilot. These are the newest kind of AI, and the next lesson explains them.

AI isn’t new

The name “artificial intelligence” goes back to a summer research project at Dartmouth College in 1956, where a small group of scientists met to explore whether machines could be made to learn. ^[6] What’s new is how powerful, and how easy to use, AI has become in the last few years.