What is AI? Plain answers for complete beginners
Find out in plain English what AI is, where you already use it and the kinds of mistakes it makes. No tech background needed.
Who it’s for: Complete beginners of any age, no tech background needed.

Everyone is talking about AI, but very few people can explain what it actually is. This course explains it in plain English. In short, AI is software that learns patterns from examples, rather than following rules that a programmer wrote by hand. A spam filter is a good example. Nobody listed every possible spam message. Instead, the filter learned from huge numbers of emails that people had marked as junk.
From there, you'll spot the AI you already use every day, from your phone recognising your face to map apps and film suggestions. You'll learn why chatbots improved so quickly, and how they write by predicting a likely next word. You'll also sort AI tools into three kinds. Generative AI makes new things, predictive AI makes guesses about what comes next, and perception AI recognises faces and speech.
Just as important, you'll learn what AI still gets wrong. It can invent facts and sources, miss recent news and stumble on long sums. So you'll practise deciding whether a task suits AI before you rely on it, and you'll learn to tell when an AI headline is hype. No tech background is needed. The course suits complete beginners of any age.
By the end you’ll be able to…
- Explain AI in one plain-English sentence a friend would understand
- Name five places AI already shows up in your daily routine
- Sort an AI tool into generative, predictive or perception AI
- List three kinds of task where today's AI reliably makes mistakes
Short on time? Start with the essentials
These 4 lessons cover the most useful ideas in this course. Each one takes about a minute.
What exactly is AI?
Explain AI as software that learns patterns from examples rather than following hand-written rules.
Where am I already using AI?
Identify at least five everyday products that use AI, from phone face unlock to spam filters and maps.
How does a chatbot write its answers?
Describe how a large language model writes by repeatedly predicting a likely next word, and what that implies for accuracy.
What does AI still get wrong?
Predict common failures: invented facts and sources, stale knowledge, errors on long calculations, missing context and judgment calls.
Every lesson in this course
This course has 11 video lessons of about a minute each, from the basics to expert habits. Each one ends with a question, three takeaways and its sources.
Foundations 4 lessons
What exactly is AI?Essential
Explain AI as software that learns patterns from examples rather than following hand-written rules.
“Everyone's talking about AI, but hardly anyone can explain what it actually is.”
Why did AI suddenly get so good?
Use three turning points (deep learning in 2012, the transformer in 2017, public chatbots in 2022) to explain why today's assistants handle language so well and why they improve so fast.
“AI is about 70 years old, so why did it suddenly get so good?”
Where am I already using AI?Essential
Identify at least five everyday products that use AI, from phone face unlock to spam filters and maps.
“You've probably used AI five times today without opening a single chatbot.”
How does AI actually learn?
Explain why a spam filter trained on examples outperforms one built from hand-written rules.
“Try writing rules to spot a cat in any photo.”
Building skills 3 lessons
Is all AI the same as ChatGPT?
Sort AI tools into generative (chatbots, image makers), predictive (recommenders, fraud checks) and perception (face unlock, speech-to-text).
“Netflix picks, face unlock and ChatGPT are all AI, but they're different machines.”
How does a chatbot write its answers?Essential
Describe how a large language model writes by repeatedly predicting a likely next word, and what that implies for accuracy.
“ChatGPT writes whole essays by doing one simple thing, guessing the next word.”
What can AI make besides text?
Name the main things generative AI creates today (text, images, voice, music, video, code) and one real use for each.
“The same idea that drafts your emails can now paint, speak, compose and film.”
Advanced 2 lessons
What does AI still get wrong?Essential
Predict common failures: invented facts and sources, stale knowledge, errors on long calculations, missing context and judgment calls.
“AI can ace exam questions and still invent a source that doesn't exist.”
Which jobs can I trust AI with?
Explain the 'jagged frontier' and check whether a task sits inside or outside what AI does well before relying on it.
“Consultants using AI did better on most tasks, but much worse on one.”
Expert 2 lessons
Is that AI headline real or hype?
Evaluate an AI claim by asking what exactly was tested, under what conditions, compared with whom, and who ran the test.
“AI beats doctors sounds huge, but three questions tell you if it's true.”
Will AI ever be as smart as us?
Explain what people mean by AGI, summarise the main camps on timelines, and say why definitions drive the disagreement.
“Some AI leaders say human-level AI is years away. Others say decades.”
Apps we recommend
These are the tools this course uses, and what each one is best for. We have no partnership with any of them.
Teachable Machine
Best for: Training your own AI model in a browserTrain a model to recognise your gestures or objects in minutes, so 'learning from examples' stops being abstract.
FreeQuick, Draw!
Best for: Seeing a neural network guess in real timeDoodle and watch a network guess as you draw: a 2-minute demo of pattern recognition and its blind spots.
FreeChatGPT
Best for: Trying a large language model for the first timeThe most widely used chatbot, with a free tier: the easiest place to see a large language model at work.
FreemiumMap the AI in your own week
Over three days, note every time AI touches your life: apps, phone, shopping, work, news. Label each kind, pick three you rely on and predict how each could go wrong and how you would notice. Then explain AI to someone who has never had it explained.
What you’ll make: A one-page map of the AI in your week, with three risk notes and your one-sentence definition of AI.
Check your own work
- Lists at least 8 AI encounters from 3 or more different days
- Each encounter is labelled generative, predictive or perception
- Three entries each name a likely failure and a concrete way you would check for it
- Your one-sentence definition uses no jargon, and a friend or relative could repeat it back
- One AI headline from the week is judged with the what-tested, compared-with-whom, who-ran-it questions
Further reading and viewing
We’ve chosen these articles, research papers, videos and guides from other trusted sources.
Elements of AI
Free, non-technical course that has taught AI basics to over a million people worldwide.
But what is a neural network?
The clearest visual explanation of how a network learns to recognise handwritten digits.
The 2026 AI Index Report
Independent yearly data on AI progress, adoption, investment and incidents.
What is artificial intelligence (AI)?
Plain overview of AI types, machine learning, deep learning and generative AI.
Questions about this course
What does AI mean in simple terms?
AI stands for artificial intelligence. In simple terms, it's software that learns patterns from examples instead of following hand-written rules. Show it enough photos of cats and it learns to spot a cat it has never seen. That same idea powers chatbots, spam filters, voice assistants and the phone that recognises your face. It's clever pattern-matching, not a thinking mind.
What are some examples of AI in everyday life?
You probably use AI many times a day without noticing. Your phone recognising your face, spam filters in your email and live traffic in map apps all rely on it. So do the film and music suggestions on streaming services, fraud checks on your bank card and speech-to-text when you dictate a message. Chatbots like ChatGPT are just the most visible kind.
Is AI always right?
No. Today's AI can sound confident and still be wrong. It sometimes invents facts, quotes and even sources. It may not know recent news, and it can slip on long calculations. In 2023, New York lawyers were fined US$5,000 for citing court cases that ChatGPT had made up. So treat AI answers as a helpful first draft, and check anything that matters.
Is ChatGPT the same thing as AI?
ChatGPT is one kind of AI, but not the only kind. It's generative AI, which makes new text, images or sound. Predictive AI makes guesses, like which song you'll enjoy or whether a payment looks like fraud. Perception AI recognises things, like your face or your voice. Knowing the difference helps you judge what a tool is actually good at.
Do I need any tech background for this course?
Not at all. The course is written for complete beginners of any age, and it explains every term in everyday words. Each lesson is a short video of about a minute, followed by one question to help you remember it. You can start from the very first lesson, or take the AI skills check to find the right starting point for you.
Get early access to this course
Tau opens soon. Join the waitlist and we’ll email you as soon as you can start this course.
We’ll only email you about Tau, and you can unsubscribe any time. Privacy policy.
We’ll email you as soon as Tau opens.


