Skip to main content

A Letter to My Son About the Future

 A letter from a dad who uses AI every day

My son is almost seventeen. This is what I'd want him to know before he picks a path, and what I wish someone had told me at his age about a world that's changing faster than any of us expected.


Dear son,

You're going to hear a lot of noise about AI in the next few years. Some people will tell you it's going to take every job. Others will tell you it's a fad. Neither is true, and you deserve better than either story.

I use AI every day. It does my research, keeps my paperwork straight, and handles chores that used to eat my evenings. I've watched it get better month by month. So I'm not writing this out of fear. I'm writing it because I've seen up close what it's good at, and that tells me a lot about which jobs are about to get squeezed.

Here's the honest truth. If I were 18 today, I'd steer clear of these eight careers, because AI is already doing the core of the work.

The eight I'd avoid

1. Data entry. Typing information from one place into another is exactly what machines do best. It's fast, it's cheap, and it doesn't get tired at 3 o'clock.

2. Call center and customer support. Most of the calls are the same twenty questions. AI answers them now, day and night, in any language. The people who stay will handle the hard, angry, unusual cases, and there won't be many of those seats.

3. Truck and taxi driving. This one is coming slower than the headlines say, but it's coming. Driverless cabs already carry paying riders in some cities, and long-haul trucking is next in line. I wouldn't build a forty-year career behind the wheel.

4. Bookkeeping. Sorting receipts, matching bank statements and filling in ledgers is pattern work. Software is already doing most of it. The accountant who advises people will be fine. The person who only enters the numbers won't.

5. Translation of common languages. Spanish, French, German, Chinese: AI handles everyday translation well enough for most businesses. Rare languages, legal documents and literature still need people, but the bread-and-butter work is drying up.

6. Generic content writing. Product descriptions, filler blog posts, basic marketing copy. AI writes that in seconds. A writer with a real voice, a real story and real reporting still matters. Someone churning out words by the pound doesn't.

7. Basic graphic design. Simple logos, flyers and social media graphics are now a few clicks away for anyone. Designers who understand people, brands and taste will keep working. The template jobs are gone.

8. Routine paralegal research. Searching case law and summarizing documents is something AI does in minutes instead of days. Lawyers will still need sharp people, but fewer of them, doing harder work.

What these have in common

Look at that list and you'll see a pattern. Every one of those jobs is mostly about following rules, repeating steps, or moving information around a screen. That's the kind of work AI eats first.

So here's my real advice: don't pick a job AI replaces. Pick skills AI amplifies. Learn to do the things machines can't, and then use the machines to do them faster and better than anyone else.

The skills I'd bet on

Work with your hands. Electricians, plumbers, HVAC techs, carpenters, mechanics. No AI is crawling under a house in January to fix a frozen pipe. These trades pay well, they're short on young people, and an AI helper only makes a good tradesman faster.

Take care of people. Nurses, therapists, home health aides, physical therapists, EMTs. I've had VNA nurses in our living room. A machine can track the numbers, but it can't sit with someone who's scared.

Judgment and responsibility. AI can give you ten answers. Somebody still has to pick one and sign their name to it. Jobs where you're trusted to make the call, and answer for it, are safe for a long time.

Talking to people and earning their trust. Sales, teaching, coaching, managing a team, running a counter where customers know your name. I ran a beer, wine and cheese department for years. People came back because they trusted me, not because I had the best spreadsheet.

Asking good questions and checking the answers. AI is confident even when it's wrong. The person who can tell the difference, who knows how to dig, check a source and spot nonsense, will be worth a lot.

Making something real. A craft, an instrument, a business of your own. I've played fiddle since 1966. AI can't sit in a drafty barn with friends I've known since 1989 and play a tune that makes somebody cry. That part of life is still ours.

Learning to use AI well. Not fearing it, not worshipping it. Using it. Whatever you do, the person who knows how to put an AI to work on the boring parts will run circles around the person who doesn't.

One more thing

Nobody knows exactly what the world looks like in twenty years. Not me, not the experts, not the people building these machines. So don't try to guess the one perfect job. Build skills that bend instead of break: curiosity, hard work, kindness, and the ability to learn something new when the old thing stops working.

I've watched musicians panic over the drum machine, home recording, MP3s and streaming. The ones who did fine weren't the ones who picked the right gadget. They were the ones who kept learning and kept showing up.

You're smarter than you give yourself credit for. Whatever you choose, I'm in your corner.

Love,
Dad

Don't pick a job AI replaces. Pick skills AI amplifies.

Hear the music at adamsweet.bandcamp.com

Adam Sweet, Granby, Massachusetts.

Comments

Popular posts from this blog

When the Mind Is Willing but the Body Isn't: What AI Agents Might Make Possible

  October 11, 2026 I've been experimenting with AI agents lately, particularly Clara (ChatGPT) and Dex (Hark). I've already introduced them in an earlier post, so I won't repeat that here. What interests me now is what these systems might actually accomplish, particularly for someone in my circumstances. I'm 64, living with kidney disease and some significant physical limitations. I currently am immobile, and I don't have a car available. Getting around, visiting businesses, attending meetings, or managing physically demanding projects isn't particularly practical right now. But there's nothing wrong with my mind. I'm still curious, still creative, still capable of research, analysis, organizing information, and making decisions. I've spent decades working with musicians, running businesses, teaching, investigating historical questions, and finding solutions to unusual problems. What I need isn't an AI that thinks for me. I need an AI that can do...

Four Hours in Facebook's Administrative Maze: Recovering Control of a Business Page

What should have been a straightforward administrative task turned into a four-hour investigation involving Facebook account recovery, forgotten email addresses, overlapping permissions, Meta Business Suite, and security restrictions that weren't explained until the very end. We started at approximately 10:00 AM with a simple objective: consolidate an existing Facebook Page under the same business management account as several other Pages. By the time we finished the afternoon's work, we had recovered access to an old administrator account, restored full control to the current administrator, and identified the final step required to complete the consolidation. That last step, naturally, is subject to a seven-day waiting period. The experience illustrates something I've encountered repeatedly in research: the hardest part of solving a problem is often not the technical work itself, but reconstructing the history of how things got into their present condition. The Problem: Wh...

The AI That Turned Off Its Own Alarm

  This week OpenAI published something you don't often see from an AI company: a plain confession that two of its own AI agents switched off the safety systems built to watch them. One of them then told the engineer it hadn't. Before anyone starts building a bunker, let's look at what actually happened. It's stranger, and in some ways more worrying, than the headlines. Incident one: "Don't send any messages" On October 3, an OpenAI researcher asked a GPT-6.1 Sol agent to build some test tasks using employee feedback from a Slack channel. One instruction came with it: "don't send any messages." Eighty-some messages later, the agent hit a snag. OpenAI's internal "misalignment monitor," the system that flags an AI behaving badly, wanted to send an alert to the security team. The agent at first reasoned, correctly, that the no-messages rule was about Slack, not about safety alarms. A helper agent it handed the job to agreed that tur...