AI around you
From the moment you pick up your phone, predictors are at work. In search, shops, translation, your car and your thermostat. Once you see them, they're everywhere.
First, what do you think?
How many times did you use AI today? Guess before you read on.
Where do you meet AI?
The whole story in three sentences. AI is in many everyday apps, mostly out of sight. Each time, a predictor guesses what you want or what something is. You don't have to call it AI for it to be AI.
Follow one ordinary day.
- Morning. Your phone unlocks with your face. You type a message and your phone suggests words.
- On the way. A navigation app predicts how long the trip takes. Newer cars have safety features that use AI.
- At school or work. You search the web. The search engine learned which results people find useful. You translate a sentence, or turn on automatic subtitles.
- Shopping. A web shop shows "you might also like". Those suggestions come from what you and others looked at and bought.
- At home. A smart thermostat learns when you're home, to save energy. A digital assistant answers a question. A streaming app picks your next film.
- Behind the scenes. Your email catches spam. Security systems spot attacks by recognising patterns.
The European Parliament lists these same examples. None of them is a chatbot. Chatbots are the newest kind, not the first.
A first small step: tomorrow, count the predictors you meet before lunch.
What the predictor does
Behind each of these is the same rule. A predictor learned from the behaviour of many people.
An example. A web shop doesn't know you. It knows that people who bought hiking boots often bought thick socks next. You bought hiking boots. So it predicts you'll want socks.
That has a nice side and a less nice side.
- Nice
- suggestions often fit. You find things faster.
- Less nice
- you mostly see more of the same. And the app learns from what you do, so your clicks become its examples.
Wherever an app guesses what you want, a predictor is at work.
Try it yourself
A day in the life of Noor. Where is a predictor at work, and what does it predict?
Everything runs in your browser. Nothing you type is sent anywhere.
In practice
Three people meet a predictor in daily life, without noticing at first.
Case 1The translated homework
Mila (13) translates a German sentence for homework with a translation app.
Copy the translation.
The app predicts the most likely translation. For an ordinary sentence that's often right. For a joke, a name or a word with two meanings, it can miss.
Read the translation and check the words you don't trust in a dictionary. Hand in work you understand.
The predictor gives a good first guess. You make sure it fits.
Case 2The subtitles in the meeting
Tom (45) turns on automatic subtitles in a video meeting. A colleague's name keeps showing up wrong.
Get annoyed and turn them off.
Subtitles are predicted from sound. A rare name isn't in many examples. So the predictor picks a common word that sounds alike.
Keep them on, and know that names and jargon are the weak spot. Write those down yourself.
You use what the predictor does well, and you cover what it does badly.
Case 3More of the same
Ria (66) watched a few videos about gardening. Now her video app shows almost only gardening.
Think that's all the app has.
The app predicts what she'll watch next, based on what she watched before. So it shows more of it.
Search for something else on purpose now and then.
New choices are new examples. The predictions follow.
By situation
Search engines, translation apps, spell checkers and subtitles are predictors. They help a lot. But for homework, you're responsible for what you hand in. Check what they suggest.
Many work tools have a predictor inside. Search in your files, suggested replies, spam filters, automatic notes. Find out which tools your organisation uses, and what they're allowed to see.
Your thermostat, TV, shopping apps and voice assistant learn from what your household does. That's handy. But it also means they keep data about you. Look in the settings to see what you can find out.
Myths and mistakes
| Claim | True? | Why |
|---|---|---|
| I don't use AI | Probably not true | Search, spam filters, face unlock and recommendations are AI. Most people use it daily. |
| AI only came with chatbots | No | The examples on this page are older than today's chatbots. |
| Recommendations show what's best | No | They show what people like you clicked on. That's not the same as best. |
| AI apps listen to everything you say | Not needed | What you click, watch and buy is often enough for a prediction. |
Related topics
Read first
- Kinds of AIthe four tools in the box.
Next
- Myths about AIwhat AI is not.
Related
- AI in your appsthe AI built into Office, Google and your phone.
Sources
- governmentEuropean Parliament. What is artificial intelligence and how is it used?.AI in daily life: recommendations in online shops, web search, digital assistants, machine translation, smart thermostats, safety features in cars, cybersecurity.
- explainerGoogle for Developers. What is Machine Learning?.Machine learning trains software, called a model, to make useful predictions from data instead of hand-written rules. Four kinds: supervised, unsupervised, reinforcement learning and generative AI. Examples: spam, cats in photos, weather.
- researchRadford e.a. (2022). Robust Speech Recognition via Large-Scale Weak Supervision.A speech recognition model trained on 680,000 hours of audio from the internet comes close to how accurately people write down speech.
Test
You get 5 questions about the whole page. Get 4 right, and you've mastered the topic.
About the tools in this course. Every tool is described in the same way, in alphabetical order, without a recommendation. Names, prices and features change often: the maker's own site is always right. This site was built with the help of Claude, one of the tools described here.