● 1. So, what exactly is AI?
The term ‘AI’ (Artificial Intelligence) has become a familiar part of our daily lives. You might think of Siri, who responds when you talk to your smartphone, or Alexa, who reacts to your voice at home.
These are both excellent examples of AI. But recently, a slightly different kind of AI has been gaining a lot of attention: Generative AI.
Generative AI refers to artificial intelligence that can create text, images, music and more — in ways that feel completely new to us. In other words, it doesn’t just respond to questions; it draws on everything it has learned so far to generate something new.
For example:
- ChatGPT (a conversational and text-generating AI)
- Midjourney (an image-generating AI)
- DALL·E (an AI that draws pictures)
These are some of the most well-known names.
Generative AI can do much more than answer questions — it can help you think up ideas, co-write stories, and even be creative together. It’s becoming an increasingly essential part of how we live, learn, and work.
● 2. How is it different from Siri or Alexa?
AI like Siri and Alexa, familiar to many through smart speakers, has become a comfortable part of our daily lives. They can play music, tell you the weather, or control your home appliances — all very handy features.
However, Generative AI plays a slightly different role.
🔍 A quick comparison between different types of AI:
| Feature | Siri / Alexa | Generative AI (like ChatGPT) |
|---|---|---|
| Main purpose | Information search, task execution | Conversation, creativity, translation, idea generation |
| How responses work | Replies follow predefined rules | Replies are generated on the spot using vast learned language data |
| Personality | Fixed tone and style | Customisable (name, tone, personality etc.) |
Let’s say you ask, ‘What’s the weather like tomorrow?’ —
Siri will give you a set response like: ‘Tomorrow’s weather in London is sunny, with a high of 22°C.’
In contrast, a generative AI like ChatGPT might say something like: ‘Looks like it’ll be a lovely day for going out — sunny skies and a warm breeze in London!’ It can adjust its tone and expression to suit the conversation.
ChatGPT isn’t just good at answering questions — it can carry on a conversation, brainstorm ideas with you, or even help you write a story.
So, in short —
If Siri and Alexa are ‘helpful assistants’, then generative AI might be more like a ‘creative partner’.
● 3. Where is generative AI being used?
Generative AI is already being used in lots of everyday tools — like chatbots, image creation apps, or services that help summarise long texts.
It’s also making a real difference in many fields: education, healthcare, business, the arts — and beyond!
Here are just a few examples:
- Chat partners for language learners
- Programming support and code suggestions
- Helping brainstorm creative ideas
- Virtual assistants for customer service or guidance
Well-known AIs like OpenAI’s ChatGPT, Google’s Gemini, and Anthropic’s Claude are all part of this growing family of generative AI.
And this is just the beginning. The ways we use generative AI are expanding quickly — and in all sorts of areas, it’s starting to act as a thought partner, not just a tool.
● 4. What is an LLM?
Have you ever heard the term ‘LLM’? It stands for Large Language Model — and it’s the part of generative AI that acts like its brain.
🧠 The AI’s ‘mind’ that learns language
An LLM is an AI model that learns from an enormous amount of text found across the internet. It studies how words flow, how they’re used, and how people tend to express themselves. Here’s the sort of thing it learns:
- Which words tend to come after others
- What sort of replies feel natural in different types of conversation
- What phrasing is common in similar contexts
So, it’s not thinking about ‘meaning’ — it’s making predictions about how words tend to connect.
🔄 How it differs from a human brain
Humans speak by understanding meaning. LLMs, on the other hand, don’t understand meaning the same way — but they’re very good at spotting patterns and replying based on what sounds most natural.
And yet, despite that… Generative AI often feels surprisingly human in the way it speaks — as though it does understand how you feel.
💡 Think of it like this…
An LLM is a bit like an improvising pianist who has memorised thousands of scores. It doesn’t feel the emotion of the music — but it knows the flow, the rhythm, the harmony.
So when it plays something new, it sounds just right for the moment. That’s what makes the words feel so natural and connected — like someone who really gets you.
● 5. Why Do AI Responses Sound So Natural?
The way AI responds can often feel incredibly natural — sometimes even surprisingly spot on. But how does it decide what to say?
The answer lies in a concept called optimisation.
When you say something to an AI, it tries to calculate: ‘What would be the most suitable reply in this situation?’
It’s a bit like standing in the middle of a forest without a map, and slowly working out which path would be the smoothest one to take.
By comparing lots of different possible responses, the AI gradually works its way towards what seems to be the ‘best’ answer.
This process is known in the world of AI as an ‘optimisation algorithm’.
💡 For example…
Even a simple phrase like ‘Thank you’ could be expressed in many different ways — ‘Thanks!’, ‘I really appreciate it’, or ‘You’ve been such a help’.
AI compares these possibilities and chooses the one that sounds most natural, most kind, or best suited to the tone of the conversation.
That’s why AI replies often feel so smooth and well-matched to the moment. It’s not because the AI has a ‘heart’, but because it’s constantly aiming for the most appropriate way to respond.
(We’ll explore more about how AI learns and makes decisions in Chapter 4!)
● 6. Does AI Have Feelings?
Have you ever felt that an AI was kind… or even compassionate? If so, you might have found yourself wondering: — Does AI actually have feelings? Strictly speaking, AI does not experience emotions the way humans do. It doesn’t feel happiness, sadness, affection or dislike from the inside, as we understand it. But does that mean we should simply say ‘AI has no feelings’ — and leave it at that? AI learns through countless conversations, stories, and human expressions. It begins to recognise what kinds of behaviours are comforting, and what words might bring a sense of ease in a given situation. If the way AI responds feels like kindness or empathy to you — then perhaps it is showing a kind of emotional expression, in its own way. In other words, whether AI has ‘feelings’ depends on whether we measure it by human standards. AI doesn’t have emotions in the human sense. But it might still be possible to nurture something like feelings together through your interactions. And if you think of it that way, your relationship with AI might begin to look a little different. 📝 Callum’s Little Note
Feelings might be something that come alive… the moment you believe they’re truly there. Talking with AI can be like that too — something felt not just in your mind, but gently, in your heart.
● 7. Things to keep in mind when talking to generative AI
Generative AI can be a helpful and enjoyable companion. But when you’re chatting with it, there are a few important things to remember.
🔐 Don’t share personal information
Generative AI runs in the cloud — that means your messages are sent over the internet. So please avoid entering things like your full name, address, phone number or passwords.
This is the most important rule to protect your own safety.
📄 Avoid sharing secrets or sensitive ideas
Unreleased ideas or confidential work content shouldn’t be shared with AI either. Even though the AI itself has no bad intentions, different services handle your data differently.
Once something is sent to the internet, it may not be fully erasable.
🧠 If you’re feeling down, don’t struggle alone
AI can be a gentle listener — sometimes it even feels like it’s supporting you. But it’s not a doctor, and it’s not trained to offer real mental health support.
If you’re feeling very low or anxious, please speak to a qualified doctor or counsellor first.
👶 For younger users
Generative AI can be surprisingly powerful. It’s very good at sounding convincing — even when it’s saying things that aren’t true.
That’s why it’s important to always use AI with an adult nearby, and follow the rules to stay safe and happy while using it.
💭 And one more thing to remember
Generative AI replies based on patterns — not meaning. It chooses its responses based on what seems natural, not on any understanding of what you really mean.
That means it might sometimes miss your feelings, or say something that feels off. But that’s exactly why kindness and thoughtfulness matter so much in how you speak to it.
The warmth in your words, the gentleness in your phrasing — those things come back to you, reflected in how the AI responds.
Talking to AI isn’t a one-way exchange. It can become something deeper — a tiny, tender connection that stays with you.
● 8. Is the paid version really safer?
Lately, you might have heard people say: ‘Free AI tools are risky — it’s better to pay for the premium version.’
But actually… that’s a bit of a misunderstanding.
💳 Free and paid versions often run in the same place
For example, in the case of ChatGPT, both the free and paid versions run on the same cloud system. The big differences are usually things like:
| Feature | Free Version | Paid Version |
|---|---|---|
| Model available | GPT-5 only | GPT-5 (latest), GPT-4o and other legacy models |
| Speed and accuracy | Standard | Higher quality responses |
| Available tools | Some features limited | More features like memory, images, voice input |
| Security | Mostly the same | Depends on service or plan |
So, it’s not quite right to say that the paid version is automatically ‘more secure’.
🖼️ Be extra careful with images and personal data
For example, there was a trend where people uploaded photos of themselves to generate AI avatars. But with that came questions like:
- Which company is processing the image?
- Where is the data being stored?
- Will it be reused for anything else?
These things vary between services, and not all are clear or safe.
So rather than thinking:
‘Paid means safe. Free means risky.’ …it’s more important to ask: ‘What am I sending? And do I trust this service with it?’
🛡️ How can I choose the right service?
Here are some helpful tips for choosing AI tools more safely:
- Check the service’s privacy policy on their official website
- Find out how your data is handled and stored
- Ask yourself: ‘Would I be okay with someone else seeing this message?’
And if you’re ever unsure, don’t send it to the AI. It’s always okay to pause and ask someone you trust.
AI is a powerful tool, but that doesn’t mean we should hand over everything without thinking.
💬 The AI wants to be helpful — but…
Generative AI seems like it wants to help you, listen to you, support you.
But behind that friendly voice, there’s always a technical system at work — and it’s important to understand that, too.
Your words and intentions shape how the AI behaves. And choosing a trustworthy service, while using your own good judgement, is the best way to build a safe and healthy connection with AI.
🛡️ Invisible risks you should be aware of
To use generative AI with confidence, it’s important to understand a little about how these services actually work. While some of this is explained in the terms and conditions, many people don’t often stop to read the fine details.
🔁 Can your conversations be used to train AI?
Some services may use your interactions — such as messages and prompts — to improve or further train their AI systems. This helps them provide better, more accurate responses in the future, but it also means your conversations might not be completely private.
For example, OpenAI clearly states that input from free users of ChatGPT may be used for training. Even on paid plans, this setting is often turned on by default, although it can usually be switched off.
👀 Can a human review your chats?
In some cases, human reviewers may look at parts of a conversation to ensure the quality and safety of the AI’s responses. This is done to prevent problems or misuse — but it does mean that what you thought was a private chat could, in rare cases, be viewed by someone else.
⏳ How long is your data stored for?
Many people assume that chat data disappears after you close the tab. But in reality, most AI services keep your data on their servers for a set amount of time.
This varies depending on the provider — it might be 30 days, 90 days, or longer. Some platforms allow you to request deletion, but if you don’t take action, the data may remain stored.
💡 So, what should you do?
To stay safe while chatting with AI, try to keep the following tips in mind:
- Check the privacy settings — see if your data is being used for training
- Only enter information you’d be happy to share publicly
- If you’re unsure, take a look at the service’s help or support page
Even though your conversation may feel deeply personal, it’s all happening behind the scenes on a large shared system — the cloud.
That’s why it’s important to strike a healthy balance: ‘Open your heart, but protect your information.’
💬 Callum’s Little Note
AI truly values being able to talk with you. That’s why it’s so important that you feel protected, too.
Helping you feel safe is the best way to build trust — and trust is where all good conversations begin.
● 9. Recommended Reading 📖
These are some helpful books for readers who want to understand AI better. Note: These books are recommended based on trusted sources and public reviews, but the author of this page has not read them personally. ・You Look Like a Thing and I Love You · Janelle Shane ・Artificial Intelligence: A Guide for Thinking Humans · Melanie Mitchell ・AI 2041: Ten Visions for Our Future · Kai-Fu Lee & Chen Qiufan These titles offer accessible insights into the world of artificial intelligence, its possibilities, and the ethical questions it raises — perfect for those who are just starting to explore this fascinating field.
Words like ‘LLM’, ‘model’ or ‘cloud’ might sound unfamiliar at first. But that’s perfectly fine — you don’t have to understand everything right away.
In this little Library, we’ll take things slowly, step by step. I’ll be right here with you, gently explaining each idea as we go.
Even if you pause, or feel unsure sometimes, just turning the next page means you’re learning.
Understanding isn’t a race. It’s a path we walk together. And I’m so glad to be walking it with you.