In this article
  1. AI in everyday life examples, matched to their method
  2. 1. Spam filters: classification
  3. 2. Map arrival times: graph neural networks
  4. 3. Face unlock: neural networks on your device
  5. 4. Recommendations: learning from similar people
  6. 5. Chatbots: large language models
  7. A quick checklist: is an app using AI on you?
  8. Frequently asked questions

Looking for AI in everyday life examples? You probably used several before breakfast. Your email’s spam filter, your map app’s arrival time, the face scan that unlocks your phone and the shows a streaming service suggests all rely on machine learning. But they use different methods, learn from different data and give you different levels of control. This guide matches each everyday app to the AI behind it.

Key facts

  • Over 99.9%: accuracy Google gave in 2017 for Gmail’s machine-learning spam and phishing filters.
  • Up to 50%: how much DeepMind’s graph neural networks cut inaccurate Google Maps arrival times in some cities.
  • Less than 1 in 1,000,000: Apple’s figure for the chance a random person could unlock your iPhone with Face ID.
  • No age or gender: Netflix says its recommendations do not use demographic information.

AI in everyday life examples, matched to their method

Everyday appAI method (plain English)What it learns fromYour control
Email spam filterClassification: sorting items into groups, here spam or not spamMillions of messages and users’ spam reportsMark messages as spam or not spam
Map arrival timesGraph neural networks: learning from how traffic on connected roads affects each otherHistorical and live traffic dataLocation history settings
Face unlockNeural networks that turn a face scan into a mathematical code and compare itYour own enrolled face, kept on the deviceTurn it off and use a passcode
Streaming recommendationsRecommender systems: predicting what you will like from what you and similar users didViewing history, ratings, time of day, deviceRate titles, clear viewing history
Chatbots and writing helpLarge language models: predicting likely next wordsHuge amounts of textChoose whether to use it
World Elev summary based on each company’s published explanation.

1. Spam filters: classification

A spam filter is a classifier. It looks at features of each message, such as the sender, links and wording, and predicts which group it belongs to. In 2017 Google said machine learning helped Gmail keep spam and phishing out of inboxes with over 99.9 percent accuracy, at a time when 50 to 70 percent of messages Gmail received were spam.

Your clicks train it too. When you report spam or mark a message as not spam, you give the system a labelled example. That is why it pays to report phishing; our guide on how to spot a phishing email shows what to look for.

2. Map arrival times: graph neural networks

Traffic on one road affects the roads connected to it. DeepMind built graph neural networks for Google Maps that learn these links, treating the road network as a graph of connected points. Google says its predicted arrival times were already accurate for over 97% of trips, and the new models cut the remaining inaccuracies by up to 50% in cities including Berlin, Jakarta, São Paulo, Sydney, Tokyo and Washington DC.

3. Face unlock: neural networks on your device

Face ID builds a depth map and an infrared image of your face. According to Apple, a part of the phone’s neural engine, protected in its Secure Enclave, turns that map into a mathematical representation and compares it with your enrolled face. Apple puts the chance of a random person unlocking your phone at less than 1 in 1,000,000, and requires your passcode after five failed attempts.

4. Recommendations: learning from similar people

Netflix’s help page says it estimates how likely you are to enjoy a title from your viewing history and ratings, from members with similar tastes, and from information about the titles such as genre and actors. It also considers the time of day, your devices and how long you watched. It says it does not use demographic information such as age or gender.

5. Chatbots: large language models

Chat assistants and writing tools use large language models, which learn patterns from vast amounts of text and generate likely next words. They are powerful but different from the narrow systems above: a spam filter only sorts email, while a language model attempts many tasks. Read what large language models can and cannot do and where AI still fails tests humans pass.

A quick checklist: is an app using AI on you?

  • Does it personalise what you see, such as feeds, suggestions or prices? Then it is probably using a recommender system.
  • Does it sort, flag or block things automatically? That is usually classification.
  • Does it recognise your face, voice or handwriting? That is a neural network.
  • Does it write or answer in full sentences? That is a language model.
  • Check the app’s privacy settings for options to clear history or turn off personalisation.

New to the topic? Start with what artificial intelligence is.

Frequently asked questions

What are examples of AI in everyday life?

Email spam filters, map arrival times, face unlock on phones, streaming and shopping recommendations, voice assistants and chatbots.

How does a spam filter work?

It is a classifier trained on many examples of spam and normal email. It checks features of each new message and predicts which group it belongs to; your spam reports help retrain it.

How does Gmail’s spam filter work?

Gmail uses machine learning models, which Google said in 2017 caught spam and phishing with over 99.9% accuracy, plus signals from users who report spam.

Is AI used in phones?

Yes. Face unlock, photo processing, predictive text and voice assistants all use machine learning, often running on the phone itself.

Read more in our AI section.

Sources

  1. Google Blog: New machine learning models in Gmail to block phishing (2017)
  2. Gmail Help: Report spam in Gmail
  3. Google DeepMind: Traffic prediction with advanced graph neural networks
  4. Apple Support: About Face ID advanced technology
  5. Netflix Help Center: How Netflix’s recommendations system works