What is AI (and What It Isn't)
Jump to section
AI is already everywhere around you
Before you opened this course today, you probably already used AI multiple times. Did your phone suggest a word while texting? That's AI. Did Netflix recommend a show? Also AI. Did Google Maps find the fastest route? AI again. Did your email spam filter catch a phishing attempt? AI. Did Spotify build your Discover Weekly playlist? Of course, AI. Artificial intelligence isn't something from the future — it's technology you already use every day without even thinking about it.
By 2025, ChatGPT surpassed 300 million weekly active users — and that doesn't count the billions who use AI indirectly through Google, Apple, Microsoft, and other products. AI is no longer a novelty for tech enthusiasts — it's the infrastructure of the modern digital world, as taken for granted as electricity or the internet.
What AI actually is
Artificial intelligence is software that learned to recognize patterns in massive amounts of data. Think of it as a student who read billions of pages of text and learned to predict what word is most likely to come next. It doesn't understand the world the way humans do — but it's incredibly good at finding patterns and generating responses based on them.
When you ask AI 'What is the capital of France?', it doesn't flip through an encyclopedia. Instead, based on the billions of texts it has seen, it knows that 'Paris' most often follows 'the capital of France is'. It's more sophisticated than just guessing the next word, but the principle is the same — pattern recognition.
Types of AI you encounter
There are different types of AI. Narrow AI is specialized for one task — face recognition, translation, movie recommendations. This has been working for years. Generative AI is what everyone's been talking about since 2022 — models like ChatGPT, Claude, and Gemini that can generate text, images, code, and audio. Then there's the concept of Artificial General Intelligence (AGI), which would have human-level capabilities — that doesn't exist yet, and experts disagree on when (or if) it will arrive.
When people say 'AI', they usually mean generative AI — tools like ChatGPT or Claude. That's what we'll be working with in this course. You don't need to understand the differences between AI types in depth — just know that the chatbots you'll be using are just the tip of the iceberg.
What AI is not
Let's bust some myths. AI is not a conscious being. It doesn't have feelings, opinions, or free will. When ChatGPT says 'I'm happy to help!', it doesn't mean it — it's just generating text that makes sense in context. It doesn't have a 'bad day' or a 'good mood'. It's software.
AI is not infallible. It can sound confident even when it's wrong. Always verify important information from an independent source. This problem is called 'hallucination' and we'll cover it in the last lesson.
AI is also not creative in the human sense. It can generate poems, stories, and images — but it does so by combining patterns from its training data. It doesn't invent anything truly new, even if the result looks original. When AI paints a picture 'in the style of Monet', it doesn't mean it understands impressionism — it recognized patterns in millions of images and generates something that statistically matches them.
Quick way to spot AI text: if a response starts with phrases like 'Great question!' or 'Of course!', that's a typical AI pattern. A human expert would just answer. Focus on the substance of the response, not its polite wrapping.
A brief history: how we got here
People have been talking about artificial intelligence since the 1950s, but for most of us, it became real in late 2022 when ChatGPT by OpenAI reached 100 million users in just two months. Since then, the field has evolved enormously. Today's AI tools are multimodal: they understand text, images, voice, and code. They have web access, can analyze documents, generate images, and even control computers. What ChatGPT could do in 2022 is a fraction of what AI can do today.
Key milestones: 2017 — Google publishes 'Attention Is All You Need', the foundation of the Transformer architecture. 2020 — OpenAI releases GPT-3. 2022 — ChatGPT launches the era of accessible generative AI. 2023 — Google Gemini, Anthropic Claude, and Meta Llama enter the scene. 2024 — AI becomes multimodal with tool use and web browsing. 2025 — AI agents begin autonomously solving complex tasks, models gain context windows exceeding one million tokens.
You don't need to understand how AI works under the hood to use it effectively. Just like you don't need to understand a combustion engine to drive a car. This course focuses on practical use, not theory.
Think about your yesterday and find at least 7 situations where you (maybe unknowingly) used AI. Write them down and for each, try to guess what type of AI powers it (narrow AI / generative AI). Starting examples: search engines, navigation, recommendations, autocorrect, email spam filters, unlocking your phone with your face, voice assistants, auto-generated video captions, translation tools, personalized ads, bank fraud detection, battery optimization.
Hint
Try going through the apps you use most frequently — most of them use AI today. Don't forget less obvious examples: a smart alarm that adjusts to your sleep patterns, or automatic photo sorting by faces.
Open any AI chatbot (ChatGPT at chat.openai.com, Claude at claude.ai, or Gemini at gemini.google.com) and try asking these questions: 1) A factual question you know the answer to ('What's the population of London?'). 2) An opinion question ('Which programming language is the best?'). 3) A creative task ('Write me a haiku about Mondays.'). 4) An intentionally nonsensical question ('When did Napoleon visit the Moon?'). Notice how AI responds to each type — where it's confident, where it hedges, where it surprises you, and how it handles the nonsensical question.
Hint
For the Napoleon question, watch whether AI rejects the nonsense or tries to fabricate an answer. For the opinion question, notice how AI balances different perspectives rather than giving one specific answer. This isn't a flaw — it's intentional behavior.
Pick one real task from your day — something you actually need to do. For example: drafting a reply to an email, brainstorming a gift for a colleague, planning dinner from ingredients you have at home, or explaining a complex concept simply. Give it to an AI chatbot and see how quickly you get a usable result. Note: How long would this have taken without AI? Is the result usable as-is, or does it need editing?
Hint
Start with something simple. If the result isn't great, try adding more context — who, what, why, for whom. You'll learn about context and specificity in lesson 4 on prompting.
What is AI — a visual explanation
- AI is software that recognizes patterns in data — it's not a conscious being with feelings or opinions
- You already use AI dozens of times daily, mostly without realizing it (recommendations, navigation, spam filters, autocomplete)
- Generative AI (ChatGPT, Claude, Gemini) can generate text, images, code, and audio — with new capabilities added every month
- AI can sound confident even when wrong — always verify important information from an independent source
- You don't need to understand the technical details to use AI effectively — this course is about practice, not theory