AI you already use

You practiced noticing behavior, not badges. Maps suggesting the quieter route, Photos grouping faces, or Netflix lining up another episode you might like — those systems guess from patterns in data. A toaster that browns bread on a timer, a calculator that multiplies, or a light switch that flips on — those follow fixed rules someone wrote. Same “smart” feeling in the store aisle does not mean same technology under the hood.

Think of three everyday tells:

1. It personalizes from your past. Playlists, For You feeds, “people also bought,” spam filters learning your junk folder.
2. It interprets messy input. Voice assistants turning speech into text, camera apps finding pets in photos, maps estimating traffic from crowd movement.
3. It can be wrong in weird ways. Wrong face tag, absurd route detour, or a recommendation for something you already hate — classic AI miss, not a broken spring.

Contrast: a microwave counting down minutes is precise and dull. It never “learns” that you burn popcorn every Saturday.

• Choosing a new app: ask “does this guess, personalize, or recognize?” vs “does it only follow buttons and timers?”
• Explaining AI to a friend without sci-fi — point at Maps live traffic or Photos search for “beach.”
• Deciding when to trust a suggestion: if it learned from your habits, check whether those habits still match what you need today.

“AI-powered” on a product page is ad copy. A smart bulb with a schedule is often just remote control + rules. Conversely, some quiet features (autocomplete, face unlock, translation) are AI even when nobody brags. Don’t upgrade trust just because a logo glows purple.

On your phone, open one maps app, one photo app, and one music or shopping app. For each, name the guess it makes (traffic, faces, taste). Then name one home tool that never guesses (kettle, lamp, basic calculator).