A hand holds up a smartphone in front of tall green plant leaves against a white background.

AI After 60: Five Everyday Tasks It Can Do Better Than Google

I’m not paid by any of these companies, and nothing here is sponsored. AI tools make mistakes, and products change fast. Never type passwords, Social Security numbers, or bank or card numbers into any of them.

Last Friday, in The Helper and the Robot, I gave you five easy ways to start with an AI assistant. The obvious next question: why would I use this instead of just Googling it?

It’s a fair question, and the honest answer is that a lot of the time, you shouldn’t.

A hand holds up a smartphone in front of tall green plant leaves against a white background.
An AI assistant can look at what you’re looking at. A list of links can’t.

The objection worth taking seriously

Most people your age haven’t touched these tools. Pew Research Center found in February 2026 that 77 percent of adults 65 and older never use an AI chatbot. Among people who don’t use them, 45 percent said doubts about accuracy were a major reason.

Those doubts are earned. In 2025 the BBC led a study for the European Broadcasting Union that checked more than 3,000 answers from ChatGPT, Copilot, Gemini and Perplexity about the news. Forty-five percent had at least one significant problem, including made-up details, outdated facts and bad sourcing.

So I’m not going to tell you AI is smarter than search. It isn’t. It’s good at different things. Once you see which things, the choice gets easy.

An older man in a dark sweater sits in an armchair, looking at the camera with a skeptical expression.
Doubts about accuracy keep many people your age away from AI. Those doubts are earned.

First, what I mean by “Google”

Google has changed. Those AI summaries at the top of your search results are Google’s own AI, and since January you can ask them follow-up questions in what Google calls AI Mode. Pew found 60 percent of Americans read those AI summaries.

When I say “Google” here, I mean the old way: type a question, get a list of links, click through to the source. When you’re reading Google’s AI summary, you’re using AI, and everything below applies to it too.

The rule

Use AI when the answer depends on you. Use search when the answer depends on today.

Store hours, prices, phone numbers, deadlines, laws, the news: those change, and you want the official page. But “what’s wrong with my printer,” “what is this thing,” “what should I ask my doctor,” “which of these fits me”: the answer depends on your details, and that’s where a page of links falls down.

1. Fixing a gadget, one step at a time

Search “TV says no signal” and you get forty forum threads about forty different TVs.

An AI assistant can do something search can’t: ask you questions back. Tell it the brand, what the screen says and what you’ve tried. It walks you through one step at a time and changes course based on what you tell it.

A hand points a TV remote at a television in a living room.
Search gives you forty forum threads. An assistant asks what your screen actually says.

Try: “My TV says ‘no signal.’ I’m not technical. Walk me through it one step at a time and wait for me to answer before the next step.”

The limit: if it tells you to buy something, open a case or do a factory reset, stop and check the manufacturer’s own help page first.

2. Figuring out what you’re looking at

A plant coming up in the flower bed. A part that rolled out of the junk drawer. A symbol blinking on the dryer.

The ChatGPT and Gemini apps let you take a photo and ask, “What is this?” Google Lens, built into the Google app, does much the same, and it’s AI too. This is the task where the line between “AI” and “Google” blurs the most, and that’s fine. Use whichever is on your phone.

A hand holds a phone over a houseplant leaf on a white table, taking its picture.
“What is this?” works for plants, parts and blinking symbols. Never for pills, moles or mushrooms.

The limit, and I mean it: never trust a photo answer about a pill, a mole or a skin spot, or a mushroom or berry you plan to eat. Those go to a pharmacist, a doctor or nobody.

3. Getting ready for a doctor’s appointment

This is not about diagnosis. AI is not your doctor, and you shouldn’t ask it to be.

It is very good at turning a jumble of worries into a short, clear list. Describe what’s been going on in plain words, leave out your name and anything identifying, and ask: “Help me turn this into the five most important questions to ask my doctor, in order.”

You walk in organized. You don’t forget the thing you meant to mention. The federal Agency for Healthcare Research and Quality makes a free app called QuestionBuilder for the same purpose, and the two work well together.

A phone showing text lies beside a spiral notebook filled with handwritten notes.
Walk into the appointment with five clear questions instead of a jumble of worries.

4. Choosing between options that fit your life

Search is good at “best phones of 2026.” It’s bad at “best phone for me.“

Tell an AI assistant what you actually need: “I want a phone with large text and a loud speaker, that works with my hearing aids, with a good camera for grandkids, under $400. Ask me any questions you need.” It asks follow-ups, narrows it down and explains the tradeoffs in plain English.

An older man with a beard looks at his phone at a table in a warmly lit room.
Not “best phone.” Best phone for you, your hearing aids and your budget.

The same works for a day trip with bad knees, a used car for short local driving or a laptop that’s simple to use.

The limit: prices and specs change. Once you have a short list, check the price and details on the official site before you buy. That’s a search job.

5. Rehearsing a hard conversation

This one surprises people. You can ask an AI assistant to play the other person so you can practice.

Asking your children whether any of them actually want the house, before you meet with the estate attorney about putting it in a trust. Calling a contractor about a bill that’s wrong. Calling a friend you’ve drifted from, which is Monday’s piece in action.

Try: “I need to tell my contractor the bill is $800 higher than the estimate. Play the contractor and push back a little. Then tell me how I did.” Two or three run-throughs, and the real call is easier.

An older man in sunglasses sits on a sunny balcony, talking on his phone.
Rehearse it twice with AI. The real call gets easier.

Use AI when the answer depends on you. Use search when the answer depends on today.

Where Google still wins, and where AI can hurt you

Anything with a date on it: hours, prices, deadlines, news, laws, the Medicare numbers we covered last month. Search, then click through to the official source.

And phone numbers, above all. In 2025, people searching for customer service numbers reported that Google’s AI summaries showed them numbers that led straight to scammers. One man looking for a cruise line’s 24-hour line nearly handed over his card number. Google said it was working to improve the results.

The rule from Hang Up, Call Back applies here too: get the number from your card, your statement or the company’s official website. Never from an AI answer.

My pick

For those five tasks, use an AI assistant. For anything that has to be true today, use search and click through to the official site.

If you want one name to start with: ChatGPT. It’s the most widely used, by 44 percent of Americans in Pew’s survey. The free version handles all five tasks, including photos, with daily limits. In that BBC-led news study, it also had far fewer sourcing problems than Gemini (24 percent of answers versus 72 percent). Gemini is a reasonable choice if you live in Google’s world, and both get updated constantly.

The strongest argument against all of this, left standing: if you’re going to double-check everything anyway, why not just use search and skip the middleman? For some people, that’s the right answer. If you’re happy with search and nothing above sounds useful, you lose nothing by waiting. The best reason to try AI is task 1, 3 or 5, where no list of links will ever do the job.

What to do this week

  1. Pick one of the five tasks and try it once. Task 3 or task 5 is a good first one.
  2. Ask it to wait for your answer before giving the next step. It changes everything.
  3. Write the rule on a sticky note: depends on me = AI; depends on today = search.
  4. Never take a phone number from an AI answer. Use the card, statement or official site.
  5. Show one friend what you tried. Teaching it is how it sticks.

You’ve spent a lifetime knowing which tool fits the job. This is just one more toolbox. Open it once, on your own terms, and see what’s inside.

Go be bold!


Try one of the five this week, then reply and tell me which one surprised you. If one of them flopped, tell me that too. I’d rather know.

Sources

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *