Welcome, let's start slowly
Hello, and welcome. I'm glad you're here.
Maybe you've heard "AI" on the news every day lately. Maybe it sounds exciting, or a bit worrying, or like hype. All of those reactions make sense. We'll go one small step at a time, and you don't need any technical background.
By the end of this lesson you'll be able to say what AI is in your own words. You'll also know a few big moments in its history and be able to tell three kinds of computer program apart.
So, what is AI, really?
Here's the plainest definition I know. Artificial intelligence (AI) is computer software that does tasks we'd normally say need human thinking, like recognising a face, understanding a spoken question or writing a sentence.
That's all it is. It isn't a robot, and it isn't a brain in a box. It's software that handles jobs that used to need a person.
Why do people find it confusing, then? A lot of it comes from films. So let's clear up two common myths.
Myth one: AI is a thinking, feeling mind. It isn't. Today's AI has no feelings, wishes or awareness. It can sound friendly, but that's a skill with words. Nobody is "in there".
Myth two: AI is always right because it's a computer. It isn't, and this one really matters. AI can make mistakes, and sometimes it sounds completely confident while it does. We'll look at why in the next level. For now, just remember that "sounds sure" doesn't mean "is correct".
A short trip through history
You might think AI appeared out of nowhere a few years ago. In fact the idea is about 75 years old.
In 1950 the British mathematician Alan Turing published a paper called "Computing Machinery and Intelligence". He asked whether machines could think. He also suggested a test, the "imitation game", now often called the Turing test. The question it asks is whether a person, chatting by text, could tell a machine from a human.
Next, in the summer of 1956, a group of researchers met at Dartmouth College in the United States. Their proposal, led by John McCarthy, is credited with bringing in the term "artificial intelligence". Many people see that workshop as the moment AI became a proper field of study.
Let's jump to 1997. IBM's computer Deep Blue beat the world chess champion Garry Kasparov in a six-game match. It was the first time a computer had beaten a reigning world champion under tournament conditions. People around the world noticed.
Then in November 2022 a company called OpenAI released ChatGPT, a chatbot you could simply type to. Within days more than a million people had tried it. That's when AI moved from the science pages into everyday conversation.
So in about seventy years, AI went from an idea on paper to something you can chat with on your phone.
Three kinds of computer program
This part clears up a lot of the jargon.
Ordinary software follows fixed rules. A person writes exact instructions, and the computer carries them out the same way every time. Think of a calculator, or a recipe followed to the letter. Two plus two always gives four. The program never learns anything new on its own.
Machine learning learns patterns from examples. Nobody writes every rule. Instead, the program is shown lots of examples and finds the patterns itself. Picture teaching a child what a dog is. You don't hand them a rulebook. You point at lots of dogs, and they work it out. A spam filter in your email works a bit like this. It has seen huge numbers of junk messages and learned what junk tends to look like.
Generative AI creates new things. It's a kind of machine learning that doesn't just sort or spot things. It produces new text, pictures or sound. When you ask a chatbot to write a birthday poem, or ask an image tool to draw "a cat in a raincoat", that's generative AI. The word "generative" just means "it makes stuff".
Here's a quick way to remember them. Ordinary software follows, machine learning learns, and generative AI creates.
Let me check with you. Suppose an app suggests the next word as you type a text message. Which kind is that? It learned from huge amounts of writing which words usually come next, so it's machine learning, not fixed rules.
Your small, safe thing to try
You don't need to sign up for anything today.
Open the messages app on your phone and start typing a sentence, like "See you". Look at the little word suggestions above the keyboard. Tap one, then another, and watch the sentence build.
That's machine learning working quietly in your hand. It isn't thinking. It's guessing the likely next word from patterns. Notice how sometimes the guesses are spot on and sometimes they're a bit silly. Keep that in mind, because it's a lovely preview of the next level.
Key takeaways
- AI is software that does tasks we'd usually say need human thinking. It isn't a mind with feelings.
- AI can be wrong, even when it sounds confident.
- Key moments: Turing's paper (1950), the Dartmouth workshop (1956), Deep Blue beating Kasparov (1997) and ChatGPT's release (November 2022).
- Ordinary software follows fixed rules, machine learning learns patterns from examples, and generative AI creates new text or images.
- You've probably been using AI for years already, for example in your phone's word suggestions.