Artificial Intelligence, explained calmly

Part 1 of 13

The world changed fast, and nobody quite explained it.

Today, with an ordinary computer, someone can create a video of a person saying things that person never said. The right face, the right voice, the right mannerisms.

That is the work of something people call Artificial Intelligence. It shows up in the news, in conversations with your kids, on your own phone. And it is going to change a lot: some of it for the better, some not.

The good news is that understanding the essentials does not take any computer knowledge. It just takes a little patience. That is what this guide offers.

And if you feel like you have fallen behind: you have not. Almost everyone is learning this at the same time you are.

Part 2 of 13

Artificial Intelligence is a very well-trained word guesser.

You know how, when you are texting, your phone suggests the next word? You type "See you at" and it offers "noon".

Artificial Intelligence, which we will call AI, was trained by reading an enormous amount of text: books, newspapers, websites, conversations. Imagine someone who read a whole encyclopedia, then every library in the world. That is more than one person could read in thousands of lifetimes.

From reading so much, it got very good at continuing sentences. So good that it seems to talk, explain, and even give advice. But underneath, it is always doing one calculation: "what word makes the most sense right now?" Then the next one. And the next.

It is the same with photos and voices. It has seen millions of faces and heard millions of voices, and learned to imitate them. That is why it can make a fake video look so much like the real thing.

The names you will hear on TV, like ChatGPT, Gemini, Meta AI and Claude, are programs of this same kind, made by different companies.

Part 3 of 13

It is not magic. It has no soul. It has no will of its own.

Many people sense that this is something otherworldly. That is understandable: a machine that answers like a person is unsettling.

But there is no spirit or intention behind it. There is calculation, the same kind a calculator does, just billions of times a second. AI does not know who you are, does not feel anything, and does not want anything. It imitates the way we speak.

That is also why it makes mistakes, and makes them with confidence. When it does not know something, it sometimes invents a good-sounding, wrong answer, like someone who never says "I don't know". For anything about health, money or official documents, always check with someone you trust.

A tool is neither good nor bad. The same knife that cuts bread can also hurt someone. What decides is the hand, and the heart, behind it.

This is not only a question for engineers.

Voices from many different faiths and philosophical traditions are asking the same question. As machines take on more of our work, how do we keep people, not profit, at the centre?

One example: in 2026, Pope Leo XIV devoted his first major encyclical to exactly this. It echoes a 19th-century predecessor, Leo XIII, who wrote about protecting factory workers during the first wave of industrial machines.

Whatever your own beliefs, the question is the same one this guide is trying to answer. Keep people in charge, and treat this new tool the way we have learned to treat any powerful one.

Part 4 of 13

When someone says "the robots" work for them, here is what that means.

So far we have talked about AI that talks: you ask, it answers. The next step is AI that acts. Experts call this an "agent". We will use the plain term: a helper robot.

It is not a metal robot walking around the house. It is a program that receives a task and does it on its own, inside the computer. It looks up information, writes a text, fills in a spreadsheet, checks another robot's work.

Think of a construction site. The engineer designs the project and makes the decisions. The foreman organizes the crew. The workers do the building. Today, one person can have several robots doing the workers' job, and sometimes even the foreman's.

But the site still has an owner. It is the person who decides what to do, checks whether it turned out well, and answers for the result. If the wall comes out crooked, it is not the brick's fault.

Part 5 of 13

You have probably already used AI. You just did not know its name.

• On WhatsApp, text messages, or Facebook. That colorful blue-and-purple circle in WhatsApp is Meta AI. You can talk to it like a person.

• On the phone. When you call somewhere and a recorded voice understands what you say back, it is often AI.

• At the bank. The alert about "unusual activity on your card" comes from an AI watching out for you.

• On your map app. The route with less traffic is calculated with AI.

• In your photos. When your phone groups "all the family photos" together, it recognizes the faces with AI.

• On TV and social media. Many suspiciously perfect images and videos are already made with AI.

Part 6 of 13

Some jobs have already changed. Others are changing now. Others will take longer.

You have already seen this happen. The telephone operator who connected calls, the typist, the bus fare collector, the person who developed your photographs. The machines arrived, some jobs disappeared, others were born.

It is similar with AI, only much faster. And now it is reaching office jobs, not just factory ones.

• Telemarketing and customer service

  Already changed a lot

  Many calls and support conversations are already handled by robots. People are left with the hard cases.

• Admin and office work

  Changing now

  Scheduling meetings, answering emails, organizing paperwork: AI already does a good part of this. What is left for people is what needs trust and people skills.

• Software development and project coordination

  Changing now, and fast

  Here the change is moving faster than almost anywhere else. Robots already write a large share of computer programs. Whoever coordinates projects now coordinates people and robots at the same time, and needs to know exactly what to ask for and how to check it.

• Civil engineering

  Will take longer

  AI helps calculate, draft, budget and review designs. But a building site is made of concrete, weather, terrain and people. And nobody lets a robot sign off on responsibility for a building.

• Healthcare

  Everything is ready, but it has not happened yet

  AI can already read imaging scans, help with triage, and summarize patient records, in some cases with excellent accuracy. But in healthcare, change moves slowly, and for good reason: there are laws, liability, trust and lives at stake.

  It is the perfect example of something the technology already allows, while society is still deciding how to use it. And it is exactly the kind of conversation healthcare workers themselves will need to have.

• Retail sales

  Will take longer

  Online stores use AI to suggest clothes and write ads. But anyone who sells in person knows something AI does not: the eye for what suits someone, honest conversation, and years of earned trust.

• Electrical engineering

  +Here AI is creating work

  AI needs so much electricity that the world is building new power plants, transmission lines and substations. You will see why in the next part.

Nobody knows the exact pace. But one rule of thumb helps: repetitive tasks done on a computer change first. Work that needs hands, presence, responsibility and care for people changes last.

Part 7 of 13

Behind AI is a huge warehouse, full of computers, that uses energy and water.

When you talk to an AI, the answer is not produced inside your phone. It is calculated far away, in buildings the size of several football fields, packed with computers. These buildings are called data centres.

Picture an iron left on all day, multiplied by millions. The International Energy Agency estimates that, by 2030, these data centres will use as much electricity as the whole of Japan uses today.

This matters because energy and water are not infinite, they cost money, and they are scarce in many places. Someone always pays that bill.

And why so much money?

Big technology companies are spending hundreds of billions of dollars a year on these warehouses. It is a race: whoever has the best AI will sell it to the whole world.

It is similar to the oil race, or the electricity race, a hundred years ago. Whoever wins this race will have a lot of money and a lot of influence. That is why so many people, on every side, are watching this closely.

Part 8 of 13

The same tool can help, and it can deceive.

Part 9 of 13

These days, seeing and hearing is no longer enough to believe.

A few seconds of someone's voice, taken from a video or a voice message on WhatsApp, text, or social media, is all it takes. With that, a scammer can make that voice say anything. With a few photos, they can make a video.

The most common scams look like this:

• A child's or relative's voice asking for money urgently: "I changed my number", "I've been in an accident", "send money right now".

• A famous doctor recommending a miracle cure. The faces of well-known TV doctors have been used in countless fake ads for cures that do not exist.

• A TV host or celebrity offering an investment, a prize, or easy money.

• A call "from the bank" asking for your password or a code that arrived by text.

These scams mostly target people over 60. It is not a lack of intelligence: the scam bets on kindness and trust.

In the United States alone, the FBI's Internet Crime Complaint Center recorded over 200,000 victims aged 60 and older in 2025, with losses topping $7.7 billion. More than 3,100 of those complaints specifically mentioned AI, with losses over $352 million.

Part 10 of 13

Before you forward something, ask five questions.

Not all fake content wants your money. Some of it wants your anger or your fear, because what moves us is what spreads the most. And with AI, thousands of pieces can be made every day, about any topic: health, faith, politics, sports, celebrities.

In 2023, a photo of Pope Francis in a stylish white winter coat went around the world. Millions of people believed it. It was fake, made by AI.

This happens even at the highest levels. In the United States, in 2024, a political consultant faked President Biden's voice in phone calls, to confuse voters from his own party.

In 2025, President Trump posted an image of himself dressed as the Pope, made by AI.

When important people use these tools, they start to feel normal. And it gets harder and harder to know what is real.

So the rule applies to everyone, on every side: nobody is proof. Not a president, not a celebrity, not the family group chat.

And if you have already forwarded something you later found out was fake: everyone has done that, including very well-educated people. It is not the fault of whoever shared it. It just got too easy to fake.

When in doubt, do not forward it. Nobody wants to accidentally spread a lie about someone, or pull a friend into a scam.

Part 11 of 13

AI knows a great deal. But it does not know what matters to you.

AI can write a lovely birthday message. But it is you who knows your friend is going through a hard time and really needs, instead, a visit.

People still decide what is right, care for someone, and take responsibility. They notice what was not said, stay present, hold on to faith, and love.

Anyone who works with AI every day learns this fast. The robots do a lot, but someone still needs to know what to ask for, check what comes back, and say "no, not like that". Human judgment became more valuable, not less.

Part 12 of 13

If you want to try it, start here. Calmly, and with no risk.

Meta AI already comes built into WhatsApp. Look for the colorful blue-and-purple circle and tap it. Then just write to it like you would to a person. If you cannot find it, ask someone in the family to show you: it takes a minute.

• Ask for an explanation

  "What does the word inflation mean? Explain it simply."

• Ask for help writing something

  "Help me write a warm birthday message for a friend I have known for years."

• Ask for a recipe

  "I have chicken, rice, and green beans. What can I make for dinner?"

Never send an AI photos of documents, card numbers, passwords, or medical results with your name on them. And remember it can be wrong. For anything about your health, always ask your own doctor.