The Latchkey Club Daily Draft — 2026-06-30
Teleprompter / Blog Script
Welcome back to the channel, guys.
Today I wanted to talk about AI agents, but not in the normal hype way.
Because apparently now everything is an agent.
Your email has an agent. Your browser has an agent. Your calendar is going to have an agent. Somebody is probably working on an AI agent for your refrigerator so it can judge your cheese choices with confidence.
And the word that keeps showing up is coworker.
AI coworker.
Digital employee.
Virtual teammate.
And I get why companies use that language. It sounds friendly. It makes the tool feel less like software and more like help. And honestly, some of these tools are helpful. I use them every day. I build with them. I have had AI do work for me that, a few years ago, would have taken hours or would have required a person with a completely different skill set.
So I am not anti-AI.
But I do think we need to be careful about giving the robot a job title.
Because the minute I start calling it my coworker, something subtle changes. I may start treating the answer like it came from a person who understands consequences. I may start assuming it knows what matters. I may start relaxing my own responsibility because the “assistant” sounded very confident and used bullet points.
And if you have ever been in a workplace long enough, you know confidence and competence are not always the same person.
Sometimes they are not even in the same building.
So, let's get into it.
I saw a headline recently that basically said AI agents are not your coworkers. That line stuck with me, because I think that is exactly the distinction we need, especially for people our age.
AI can help.
AI can draft, summarize, compare, organize, write code, find patterns, create checklists, remind you of things, pull information together, and reduce a lot of the friction that wears you down during a normal week.
But it is not a coworker.
A coworker has a life. A coworker has judgment shaped by consequences. A coworker knows what it feels like when a mistake lands on a customer, a family, a patient, a team, or your own reputation. A coworker can be wrong too, obviously. Let’s not get sentimental. We have all sat in meetings where the spreadsheet was technically present but wisdom was not.
But a human being carries accountability in a way software does not.
AI does not wake up at 2 in the morning thinking, I hope that advice was right.
It does not have to call the customer back.
It does not have to look your spouse in the eye and explain why the calendar got messed up.
It does not have to retire on the plan it recommended.
You do.
That is the part I think we cannot hand away.
And maybe that is where age helps. Not because being older makes us immune to bad decisions. I have ordered things online with a level of confidence that did not survive the shipping box.
But by 55 or 60, you have lived with enough consequences to know that the last step matters.
The follow-through matters.
Who owns the decision matters.
Who checks the work matters.
And that is the difference between using AI like a tool and treating AI like an authority.
A tool needs a person holding it.
An authority gets obeyed.
Those are not the same.
I think Gen X has an interesting advantage here because we came up in a world where tools were mostly dumb. And I do not mean that as an insult. A hammer did not pretend to be your life coach. A calculator did not tell you it had reflected deeply on your tax situation.
The tool did the thing it could do, and you were still responsible for whether you used it well.
Now the tools are more powerful and more persuasive. The interface got smoother. The voice got nicer. The answer got faster.
But the basic question has not changed.
What is this tool for?
Where is it likely to fail?
How do I check it?
What decision still belongs to me?
Those are grown-up questions. And I mean that in the best way.
If you are using AI to help with retirement planning, let it organize the questions. Let it explain vocabulary. Let it compare scenarios. Let it make a list of things to ask a real planner.
But do not let it become the planner.
If you are using AI for health, let it help you understand a term, summarize notes, prepare better questions for a doctor.
But do not let it become the doctor.
If you are using AI at work, let it draft, research, code, analyze, and clean up the mess of information we all swim in now.
But do not let it become the person who owns your judgment.
And if you are using AI for family logistics, reminders, emails, schedules, college paperwork, caregiving stuff, whatever it is, let it carry some of the mental clutter.
But do not let it convince you that automation is the same thing as attention.
That distinction matters.
Because one of the temptations with powerful tools is to step back too far.
At first, you are using the tool.
Then the tool starts making things easier.
Then you start trusting it because it has been right a few times.
Then one day you realize you are not supervising anymore. You are just approving whatever it hands you because you are tired and it sounds official.
That is the dangerous spot.
And honestly, that danger is not new. We have done versions of this forever.
We trusted GPS until it sent us into a neighborhood we did not recognize and then had the nerve to say “make a U-turn” like this was all our idea.
We trusted forms, systems, vendors, dashboards, and policies that looked official but still needed a person with common sense to ask, wait, is this actually right?
AI is just a more convincing version of that problem.
So maybe the practical move is this: do not give AI a job title. Give it a job description.
Not, “Be my financial advisor.”
Try, “Summarize this retirement article and list five questions I should ask a professional.”
Not, “Manage my health.”
Try, “Turn these appointment notes into a short checklist of follow-ups I can verify.”
Not, “Run my life.”
Please do not ask software to run your life. Most of us are barely qualified and we live here.
Try, “Look at this messy week and help me identify the three things I cannot afford to drop.”
That is different.
Specific task. Limited scope. Human review.
That is how I am trying to think about it.
AI is useful when I point it at a defined problem. It gets dangerous when I let it define the problem for me.
And the older I get, the more I think our job is not to be impressed by the tool. Our job is to govern the tool.
That sounds dramatic, but I mean it in a very normal Tuesday-afternoon way.
Govern the tool that drafts the email.
Govern the tool that summarizes the document.
Govern the tool that suggests the plan.
Govern the tool that says it found the answer.
Because your experience is still needed. Your judgment is still needed. Your values are still needed. Your relationships are still needed.
The machine can move fast, but it does not know what you owe people.
It does not know which promise matters most.
It does not know whether a technically correct answer is a wise answer.
That is where we come in.
And I actually find that encouraging.
Because a lot of the AI conversation makes older people feel like they are behind. Like the world changed again, and now we have to sprint to catch up with a bunch of twenty-five-year-olds who were born understanding settings menus.
But maybe the real skill is not sprinting.
Maybe the real skill is supervision.
Maybe it is knowing how to assign a task, inspect the result, ask the follow-up question, catch the weird assumption, and decide whether the output belongs in real life.
That is not flashy.
But it is valuable.
And if you have managed people, raised kids, fixed problems at work, helped aging parents, dealt with contractors, balanced family responsibilities, or just lived long enough to know that clean plans can get messy fast, you already have some of that muscle.
You know that delegation does not erase responsibility.
So maybe that is the posture for this AI moment.
Do not fear it.
Do not worship it.
Do not hand it the keys and go take a nap, tempting as that sounds on some days.
Use it.
Train it for narrow tasks.
Check it.
Keep the responsibility where it belongs.
Anyway, that is what I have been thinking about.
Maybe AI can do more than I expected.
But that does not mean it gets to become more than it is.
It is a tool. A powerful one. Sometimes a weird one. Sometimes a useful one.
But the judgment still has to come from somewhere else.
And for those of us in the second half of life, that may be the opportunity.
Not to pretend we are young.
Not to chase every feature.
But to take the judgment we have earned and put it to work with better tools than we ever had before.
Video Prompt Script — Questions to Answer Without Reading
Use these as prompts. Don't read them on camera; answer them naturally.
Opening: What do you think when you hear companies call AI agents “coworkers” or “digital employees”?
- Follow-up: Why does that language bother you, even though you use AI every day?
Concrete example: Where has AI actually helped you recently — coding, summarizing, reminders, family logistics, work documents, or planning?
- Follow-up: Where do you still need to verify it before trusting the output?
Main distinction: What is the difference between giving AI a job title and giving it a job description?
- Follow-up: Can you give examples like “summarize this article” versus “be my planner”?
Gen X angle: How did growing up with “dumb tools” shape the way Gen X understands responsibility?
- Follow-up: What changed when tools started sounding confident and human?
Aging/work/retirement angle: Why is this especially important for people in their fifties and sixties dealing with work changes, retirement planning, health, scams, and family responsibilities?
- Follow-up: What decisions should never be fully outsourced to AI?
Practical takeaway: What is one safe way viewers can use AI this week without handing it too much authority?
- Follow-up: What should they check before acting on the answer?
Closing: What responsibility still belongs to us, no matter how powerful the tools get?
- Follow-up: Ask viewers where they draw the line with AI.
Title Options
- Don’t Give AI a Job Title
- AI Is Not Your Coworker — And That’s Good News
- The Gen X Advantage With AI: Supervision
Thumbnail / Onscreen Text Options
- AI IS NOT YOUR COWORKER
- GIVE IT A TASK, NOT THE KEYS
- THE ROBOT DOESN’T OWN THE DECISION
Shorts / Reels Cutdowns
- “Job title vs. job description” — Cut from “do not give AI a job title” through the retirement/health examples.
- “Confidence and competence” — Use the line about confidence and competence not always being the same person, then land on verification.
- “The Gen X AI advantage” — Clip the section about growing up with dumb tools and learning that the person using the tool still owns the result.
Viewer Question
Where do you draw the line with AI — what tasks are you comfortable handing it, and what decisions still have to stay human?