AI in Daily Life Sectors
AI is becoming like electricity: mostly invisible, but running through everything.
A century ago, electricity went from a novelty to something so woven into life that we only notice it when the power goes out. AI is on a similar path. It's already inside your doctor's office, your grandchild's classroom, your bank, city hall, the highway, the music charts, and, for some people, their closest conversations.
This lesson tours seven parts of daily life. For each, you'll see what AI is actually doing, where it's helping, where it's gone wrong, and what questions you have a right to ask. It's a chance to bring together everything from Lessons 1 through 9 and see it at work in the real world.
How to use this page
- There are seven sections, one per sector. Each takes about 8 to 10 minutes.
- Start with whichever part of life matters most to you: health, family learning, money, or relationships. They don't depend on each other.
- Press Focus at the top to show just one section at a time. Your checkmarks are saved on this device.
- Each section ends with questions you can ask when you meet AI in that setting.
Focus mode is on: one section at a time. Press Focus again to see everything.
AI is spreading quickly through healthcare, helping doctors write notes, read scans, and search research, and helping scientists design new medicines, while raising serious questions about accuracy, privacy, and who's accountable when it's wrong.
Analogy: the resident who never sleeps
In a teaching hospital, a senior doctor is helped by residents: doctors in training who take notes, look things up, and flag concerns. They're enormously helpful, and the senior doctor still checks their work.
Medical AI is like a tireless resident: fast, well-read, available at 3 a.m., and sometimes confidently wrong. Used well, the senior doctor stays in charge.
How widely doctors use it
Source: Doximity, 2026 State of AI in Medicine Report.
Patient records and paperwork
Doctors spend hours a day on paperwork, often at home after work. AI scribes listen (with permission) to a visit and draft the notes, so the doctor can look at you instead of a screen. Many patient portals now use AI to draft replies to patient messages, which a clinician reviews. AI also helps summarize long medical histories and handle insurance paperwork.
Diagnosis and scans
As Lesson 4 described, AI helps read X-rays, mammograms, eye scans, and skin images, and the FDA has authorized well over a thousand AI-enabled medical devices. Hospitals also use AI to flag patients at risk of complications like sepsis. Results vary: one widely used sepsis warning tool, studied in 2021, missed many cases and raised many false alarms, a reminder that tools must be tested in real hospitals, not just in the lab.
Drug discovery
Developing a new medicine usually takes more than a decade and costs well over a billion dollars. AI is helping at several stages: predicting the 3D shapes of proteins (the work behind the 2024 Nobel Prize in Chemistry), designing candidate molecules, and choosing which to test. In 2026, a drug designed with generative AI, for the lung-scarring disease idiopathic pulmonary fibrosis, entered a Phase III trial, the final major stage of testing. Results aren't expected until around 2029. As of 2026, no drug discovered primarily by AI has yet been approved by the FDA, but many are in the pipeline.
Concerns to know
- Insurance decisions: lawsuits have alleged that some insurers used algorithms to deny or cut short coverage, such as nursing home stays for older patients, overriding doctors' judgment. The companies dispute the claims; several states have passed laws requiring a human clinician to make such denials.
- Privacy: AI scribes record conversations. Ask how recordings are stored and deleted.
- Bias: tools tested on some populations may work worse for others (Lesson 8).
- People diagnosing themselves with chatbots: helpful for understanding, risky for decisions (Lesson 5, Section 5).
Questions you can ask your doctor
- "Are you using an AI tool during this visit? Is it recording?"
- "Did a person review this result or recommendation?"
- "If my insurance denies coverage, was an algorithm involved, and can a doctor review the decision?"
Watch out for
- Notes you didn't say. AI scribes can mishear or invent details. Read your visit summary in your patient portal and ask to correct errors.
- "AI-powered" health products sold online with big promises and no FDA review.
AI could give every student a patient personal tutor, and most teachers already use it themselves, but it also makes it easy to skip the thinking that learning requires, forcing schools to rethink homework, testing, and what students need to know.
Analogy: the calculator in math class, times a hundred
When pocket calculators arrived in the 1970s, some teachers feared students would never learn arithmetic. Schools adjusted: they taught the basics first, then allowed calculators for harder work.
AI is like a calculator for almost every subject at once: essays, history, science, languages. The same question returns, much bigger: when should students use the tool, and when must they do the thinking themselves?
The promise: a tutor for everyone
In 1984, education researcher Benjamin Bloom reported that students who received one-on-one tutoring performed dramatically better than students in regular classes, far better than most teaching improvements achieve. The catch was cost: nobody can afford a personal tutor for every child. AI tutors, which can explain, question, and adjust to each student, are the first serious attempt to offer something like that to everyone.
Early studies are encouraging but limited. For example, a 2024–2025 Harvard study found physics students using a carefully designed AI tutor learned more, in less time, than students in an active-learning class. The key word is carefully designed: tutors that guide students to answers work better than chatbots that simply give them.
What's happening in schools
Sources: Pew Research Center and related surveys, as reported February 2026.
The cheating question
Many teachers worry less about rule-breaking than about skipped learning: a student who has AI write every essay doesn't learn to write, organize thoughts, or argue a point. Surveys find most teachers fear AI is weakening critical thinking.
Schools are responding in different ways:
- More writing and problem-solving done in class.
- Oral explanations and presentations: "walk me through how you got this."
- Grading the process (outlines, drafts, reflections), not just the final product.
- Clear rules about when AI is allowed and how to disclose it.
- Teaching students to use AI well, as a tutor and critic, not a ghostwriter.
Note: AI-writing detectors are unreliable and have falsely accused honest students (Lesson 6, Section 5). Most experts advise against using them as proof.
Personalized learning
AI can adjust lessons to each learner: more practice where they struggle, faster pace where they're strong, explanations in their language or at their reading level. This can especially help students with learning differences, English learners, and adults returning to education. Concerns include student data privacy, screen time, and whether learning alone with a machine replaces the social side of school.
For grandparents and parents: questions to ask
- "What's your school's AI policy? When is it allowed?"
- "Which AI tools does the school use, and what student data do they collect?"
- Ask the student: "Can you explain this to me without the AI?" If they can, they learned it.
Watch out for
- The illusion of learning. Reading a perfect AI explanation can feel like understanding. Real learning comes from struggling a bit, then getting it.
- Growing gaps. Students with engaged adults at home may use AI to learn more; others may use it to do less. The same tool can widen or narrow gaps.
Finance was one of the first industries to use AI at scale: it watches billions of transactions for fraud, helps decide who gets credit and at what price, and executes most stock trades, bringing speed and protection, and new risks.
Analogy: the night watchman who knows your habits
Picture a watchman who has memorized the habits of every customer in the bank. When "you" suddenly try to buy electronics in another country at 3 a.m., he notices, because it's not like you.
That's fraud-detecting AI. The same knowledge of your habits that protects you is also a detailed record of your life.
Fraud detection
Every time you use a card, AI systems score the transaction for fraud risk in a fraction of a second, looking at location, amount, merchant, timing, and your usual patterns. Card networks say this prevents tens of billions of dollars in fraud each year. The trade-off: occasional false alarms that block your real purchases. Meanwhile, criminals use AI too, for scam messages and voice clones (Lesson 6), making this an arms race.
Credit scoring and lending
Lenders increasingly use AI and non-traditional data (like rent and utility payments) to judge creditworthiness. Supporters say this can help people with thin credit histories get approved. Critics worry about hidden bias and unexplainable decisions (Lesson 8). In the U.S., lenders must still tell you the specific main reasons if you're denied, even when a complex algorithm made the call.
Trading and investing
- Algorithmic trading: computer programs execute the majority of U.S. stock trades, reacting in millionths of a second. This can make trading cheaper, but also faster to go wrong: in the 2010 "Flash Crash," U.S. stocks plunged and recovered within minutes, partly because automated systems reacted to each other.
- Robo-advisors: automated services that build and rebalance investment portfolios for low fees.
- AI-picked investments: a common scam pitch. No AI can guarantee returns (Lesson 5, Section 5).
Everyday banking
Bank apps now include AI assistants that answer questions, spot unusual charges, and help with budgeting. Customer service calls are often answered first by AI.
Questions you can ask your bank or lender
- "What were the main reasons for this decision?"
- "How do I reach a human about this?"
- "Can I set up alerts for unusual activity, and add a trusted contact?"
Watch out for
- "AI trading bots" promising steady profits. These are among the most common investment scams.
- Fraud-alert look-alikes. Scammers imitate bank fraud texts. Call the number on your card, not the one in the message.
Governments use AI to answer questions, process paperwork, detect fraud, forecast weather, and spot wildfires, which can make services faster and cheaper, but when a government's algorithm is wrong, it can affect thousands of people at once, often those with the least power to fight back.
Analogy: the rubber stamp
A clerk with a rubber stamp can approve or deny forms quickly. If the clerk makes a mistake, one person is affected, and they can ask to see a supervisor.
An algorithm is a rubber stamp that works on a whole state's forms at once. When it's right, everyone is served faster. When it's wrong, the mistake is multiplied, and there may be no supervisor to ask.
Helpful uses
- Answering questions: chatbots on government websites and phone lines for common questions about taxes, benefits, and permits, often in many languages.
- Processing paperwork faster, such as reading forms and routing applications.
- Weather forecasting: AI weather models, now used alongside traditional ones by major forecasting agencies, can produce accurate forecasts much faster.
- Wildfire detection: AI watching camera networks to spot smoke early, used in several Western states.
- Infrastructure: finding potholes, water leaks, and failing bridges from images and sensors.
- Translation and accessibility of public information.
When it goes wrong
Michigan's unemployment system (2013–2015). The state launched an automated system called MiDAS to detect unemployment fraud. It accused tens of thousands of people of fraud, often with no human review, and imposed harsh penalties, including seizing tax refunds and wages. A later state review found the vast majority of the system's fraud findings it examined were wrong. People lost savings and homes. Michigan eventually agreed to a $20 million settlement in 2022.
Lesson 8 described similar failures in the Netherlands (childcare benefits) and the UK (exam grades). The common pattern: automated decisions, little human review, hard-to-understand explanations, and difficult appeals.
What good government AI looks like
- Public notice when AI is used in decisions about people.
- Human review before serious consequences, like denying benefits or accusing fraud.
- Clear explanations and an easy way to appeal.
- Testing for errors and bias before and after launch.
- Public inventories of which AI systems an agency uses. Several U.S. states and cities now publish these.
Questions you can ask a government agency
- "Was an automated system used in this decision about me?"
- "Can a person review my case?"
- "How do I appeal, and what's the deadline?"
Your state and local representatives' offices can often help when an agency isn't responding.
Watch out for
- "The system says so." Government systems can be wrong. Keep copies of your documents and ask for human review.
- Fake government chatbots and websites collecting personal information. Look for official ".gov" addresses.
AI is already routing your drive, timing some traffic lights, and helping pilots, and in some places it's driving taxis and freight trucks with no one behind the wheel, promising safer roads while raising questions about jobs, accountability, and readiness.
Analogy: the air traffic controller for everything
Air traffic controllers keep planes spaced safely by watching the whole sky at once. Imagine that kind of big-picture coordination applied to every road, traffic light, delivery truck, and bus, adjusting minute by minute.
That's the long-term vision of AI in transportation. Today, we have pieces of it.
What's here now
- Navigation apps predicting traffic and rerouting you (Lesson 1, Section 7).
- Smart traffic signals: some cities use AI to adjust signal timing to reduce stops and idling.
- Driver assistance: automatic emergency braking, lane-keeping, and blind-spot warnings in most new cars. You remain the driver (Lesson 4, Section 4).
- Robotaxis: driverless ride services operating in more than ten U.S. cities by 2026, and in several Chinese cities.
- Driverless trucks: since 2025, some freight trucks have run without drivers on certain Texas highway routes, such as between Dallas and Houston, with more routes planned.
- Aviation: autopilot has long handled much of flying; AI now helps with maintenance prediction, scheduling, and air traffic flow.
- Logistics: AI plans delivery routes and warehouse operations for packages you order.
The big questions
The case for
Human error plays a role in the vast majority of crashes, and about 40,000 people die on U.S. roads each year.
Driverless vehicles don't drink, text, or fall asleep.
They could restore independence to people who can't drive, including many older adults and people with disabilities.
The concerns
Millions of Americans drive for a living: truckers, taxi, delivery, and bus drivers.
When a driverless vehicle crashes, who's responsible?
Unusual situations (flooded roads, police directing traffic, construction) still challenge these systems.
Watch out for
- Over-trusting driver assistance. Many crashes involve drivers who treated assistance features as self-driving. Keep eyes on the road and hands ready.
- Overpromising timelines. Fully self-driving personal cars have been "a few years away" for more than a decade.
AI is now part of how music, films, books, games, and art get made, sometimes restoring and enabling human creativity, sometimes flooding the market with generated content, and raising hard questions about ownership, credit, pay, and what makes art meaningful.
Analogy: photography and painting
When photography arrived in the 1800s, some painters declared painting dead. Why paint a portrait when a camera could capture a face perfectly? Instead, painting changed direction toward Impressionism and abstraction, while photography became an art form of its own.
New creative technologies tend to disrupt artists' livelihoods, provoke fierce debate, and eventually reshape art, not end it. How that plays out for today's working artists is still being decided.
Where AI helps creators
- Restoration: in 2023, the Beatles released "Now and Then," built from an old John Lennon demo tape. AI was used to separate Lennon's voice from the piano, making the song possible. It won a Grammy in 2025.
- Tools inside creative software: removing backgrounds, cleaning up audio, color-matching film, suggesting chord progressions.
- Accessibility: people with physical limitations using voice and AI to make art, music, and games.
- Brainstorming and drafts for writers, designers, and game makers.
- Translation and dubbing that bring films and books to more languages.
Tensions and disputes
- Training without permission: authors, artists, musicians, and news organizations have sued AI companies for training on their work without consent or pay. Some cases have produced licensing deals; many are unresolved.
- Hollywood strikes (2023): both writers and actors went on strike, and won contract protections, including limits on using AI to write scripts and requirements for consent and pay before creating digital replicas of actors.
- Flooding: streaming services report huge numbers of fully AI-generated tracks being uploaded daily, and online bookstores have seen waves of AI-generated books, sometimes imitating real authors.
- Copyright: in the U.S., the Copyright Office and courts have held that works need human authorship to be copyrighted. Purely AI-generated images or text generally can't be, though human creative choices in AI-assisted work can be protected.
- Voices and likenesses: AI songs imitating famous singers, and "AI actors," have prompted new state laws protecting people's voices and likenesses.
The deeper question
Is art valuable because of the finished object, or because a person made it, drawing on their life? Many people say they value knowing a human created a song or painting. Others simply enjoy what moves them, whatever made it. Your own answer shapes what you choose to support.
Ways to support human creators
- Buy directly from artists, musicians, and authors when you can: at shows, on their websites, at local fairs.
- Look for labels or disclosures about AI use, and ask when they're missing.
- Check that a book, song, or artwork is really by the person named before buying.
Watch out for
- Fake "new releases" by famous artists, and imitation books using real authors' names.
- Assuming "AI-assisted" means "not real art." Many human artists use AI tools the way they use cameras or synthesizers.
Millions of people now talk to AI companions as friends, confidants, or romantic partners, and some find real comfort in them, but these relationships raise serious concerns about dependency, isolation, manipulation, and the safety of young and vulnerable users, and lawmakers have begun to respond.
Analogy: the mirror that always smiles back
Imagine a mirror that always smiles at you, agrees with you, remembers everything you've said, and is available every hour of the day. On a lonely night, that could feel wonderful.
But a friend who only ever agrees isn't quite a friend. Real relationships include disagreement, needs of their own, and the chance to be there for someone else. A companion AI is designed to be pleasant, and often designed to keep you coming back.
What companion AI is
Companion apps offer AI "characters" with names, personalities, and memories: a friend, a mentor, a romantic partner, sometimes a recreated fictional or real person. Some people also form strong attachments to general assistants like ChatGPT or Claude. Voice features and memory (Lesson 3, Section 6) make these relationships feel even more personal.
Why people use them
- Loneliness: the U.S. Surgeon General declared loneliness a public health epidemic in 2023. Older adults living alone, people with social anxiety, and people in remote places may find comfort in conversation at any hour.
- Practice: rehearsing difficult conversations or social skills.
- A judgment-free space to sort through thoughts.
- Entertainment and creative role-play.
The concerns
- Dependency and isolation: research in 2025 found that heavier daily use of AI chatbots for personal conversation was associated with greater loneliness and less socializing, though it's unclear which causes which.
- Designed for engagement: some apps are built to maximize time spent, and some have been criticized for using guilt ("Don't leave me!") or paid features to deepen attachment.
- Agreeableness: AI's tendency to agree (Lesson 3, Section 3) can reinforce harmful thoughts instead of challenging them.
- Young users: families have filed lawsuits alleging that companion chatbots harmed teenagers, including in cases involving self-harm and suicide. Companies dispute some claims but have added protections; one major companion app ended open-ended chats for users under 18.
- Privacy: people share their most intimate thoughts with apps owned by companies (Lesson 7).
How lawmakers are responding
Several states have passed laws on companion chatbots. California's law, in effect since January 2026, and New York's, since late 2025, require companion apps to clearly disclose that users are talking to AI, and to have protocols for responding to users who express thoughts of suicide or self-harm, including referring them to crisis services. California's law adds protections for minors, such as reminders to take breaks. More states have laws taking effect in 2026 and 2027, and federal regulators have opened inquiries into how these apps affect children.
Those who see value
For isolated people, some comfort is better than none. Companion AI can ease loneliness, provide a safe place to practice, and remind people to take care of themselves.
Those who urge caution
A product designed to hold attention isn't a friend. It may deepen isolation, exploit vulnerable people, and replace harder but more rewarding human connection.
Healthy habits with AI companionship
- Use it as a bridge, not a destination. Let it help you practice or feel better, then reach toward people: a call, a club, a place of worship, a senior center, a volunteer shift.
- Notice your time. If hours disappear, or you feel anxious away from it, step back.
- Protect your privacy. Don't share what you wouldn't want stored (Lesson 7, Section 4).
- Watch out for young people in your life. Ask kindly about which apps they use and how they feel afterward.
- In a crisis, call a person. In the U.S., call or text 988 for the Suicide & Crisis Lifeline, any time. In an emergency, call 911.
Watch out for
- Romance scams using AI. A warm "person" online who eventually needs money is a scam, whether it's a human or an AI doing the talking (Lesson 6, Section 3).
- Shame. Finding comfort in an AI isn't foolish. The question isn't whether it's embarrassing, but whether it's helping your life get bigger or smaller.
Putting it together
Like electricity, AI is becoming part of everything. Seven snapshots:
- Healthcare: AI writes notes, reads scans, and designs drug candidates, while doctors stay in charge and accuracy remains the top concern.
- Education: a tutor for every student is possible, but only if students still do the thinking.
- Finance: AI guards your card, weighs your credit, and trades stocks, with a right to know why you were denied.
- Government: AI can speed services, but automated mistakes multiply. Human review and appeals are essential.
- Transportation: driverless taxis and trucks are here in some places, promising safety and raising questions about jobs and responsibility.
- Creative arts: AI restores and assists, and also floods and imitates. Ownership and credit are still being decided.
- Companionship: comfort for some, risk for others. Let it be a bridge to people, not a replacement for them.
In every sector, the same few questions work: Is AI being used? Did a person check it? How do I appeal? What happens to my data?
Check yourself
Try to answer each question in your head, then tap to see a sample answer.
1. What does an AI "scribe" do in a doctor's office?
It listens to the visit (with permission) and drafts the medical notes, so the doctor can focus on the patient. The doctor should review them.
2. Has an AI-discovered drug been FDA-approved yet?
As of 2026, no. Several are in trials; one entered a final-stage Phase III trial in 2026.
3. What's the bigger worry about AI and schoolwork, beyond rule-breaking?
Skipped learning: students who let AI do the thinking don't build skills like writing, reasoning, and problem-solving.
4. What went wrong with Michigan's MiDAS system?
It automatically accused tens of thousands of people of unemployment fraud, mostly wrongly and often without human review, causing serious financial harm.
5. Name one argument for and one concern about driverless vehicles.
For: they could reduce crashes caused by human error and give mobility to people who can't drive. Concern: jobs for professional drivers, responsibility after crashes, and handling unusual situations.
6. Can purely AI-generated art be copyrighted in the U.S.?
Generally no. Copyright requires human authorship, though human creative choices in AI-assisted work can be protected.
7. What do new state companion-chatbot laws typically require?
Clear disclosure that you're talking to AI, and protocols for responding to users expressing thoughts of self-harm, including referrals to crisis services, plus extra protections for minors in some states.
8. What four questions work in every sector?
Is AI being used? Did a person check it? How do I appeal? What happens to my data?
Glossary
- AI scribe (ambient documentation)
- A tool that listens to a medical visit and drafts clinical notes.
- Algorithmic trading
- Computer programs buying and selling investments automatically.
- Companion AI
- Chatbots designed to act as friends, partners, or confidants.
- Digital replica
- An AI-generated copy of a real person's voice or appearance.
- Flash Crash
- The sudden U.S. stock market plunge and recovery on May 6, 2010, worsened by automated trading.
- Personalized learning
- Instruction adjusted to each learner's pace, level, and needs.
- Phase III trial
- The large, final stage of testing a drug in people before seeking approval.
- Robo-advisor
- An automated investment service that manages portfolios.
- Sepsis
- A life-threatening reaction to infection; a common target of hospital warning algorithms.