Calvin's Updates

The Latchkey Club Daily Draft — August 27, 2026

Teleprompter / Blog Script

A few months ago, I started thinking of age 57 as the beginning of a five-year countdown.

Five more years to get healthier. Five more years to get serious about retirement. Five more years to pass along as much as I can at work before I leave.

That sounded responsible when I first said it.

But I noticed something hiding inside the phrase “five more years.” It can sound like a plan, or it can sound like I’m already standing near the exit with my hand on the door.

Welcome back to the channel, guys. Today I wanted to talk about the strange stretch of time when retirement is close enough to see, but still far enough away that you can’t really treat work like it’s over.

Five years is not forever. But five years is also not a long weekend.

Think about how much work changed between 2021 and 2026. Remote work changed. Meetings changed. AI went from something most people heard about in the news to something that can write, research, organize information, build software, and show up inside tools we already use.

I don’t know exactly what the next five years will bring. I’m fairly sure it won’t be nothing.

That creates a real question for people our age. How much should we keep learning when we may be close to leaving?

I understand the temptation to say, “I’m not starting over now. I’ll just keep doing what I know until retirement.” There’s a practical side to that. Time and energy are limited. I don’t need to learn every new app, follow every model release, or spend my evenings trying to impress somebody with how current I am.

At this age, I’m trying to protect my health and my attention too. Staying relevant at work doesn’t help much if I arrive at retirement exhausted and unable to enjoy it.

But there’s another risk in deciding too early that learning is finished.

I read an AARP article recently about research into AI and older workers. It said workers 55 and older in jobs that are more exposed to AI have been somewhat more likely to leave the workforce since ChatGPT arrived. The researchers were careful about the limits of the data, so I don’t think we can say AI caused every one of those decisions.

Still, the explanation made sense. If you believe you only have a short time left, the payoff from learning a difficult new tool may not feel worth it. A younger worker might think they have twenty years to benefit from the effort. An older worker may look at the same training and think, “By the time I figure this out, I’ll be gone.”

I’ve had some version of that thought.

The problem is that retirement dates don’t always arrive on our schedule. A company changes direction. A department reorganizes. Health changes. Family needs change. The work itself may change enough that the job you planned to hold for five more years isn’t really the same job anymore.

If I mentally retire before I financially retire, I may give up some of my ability to choose how the last part goes.

I don’t mean that as a speech about constantly reinventing yourself or racing younger people. They have strengths I don’t have, and I have things that took a long time to learn.

It’s about not confusing selectivity with disengagement.

There are too many tools now, and they change too quickly to master all of them.

What works better for me is starting with one real problem.

There is a repeated task that wastes time. There is information scattered across systems. There is a process that makes sense in my head but is hard to explain to somebody else. There is a document that keeps getting rebuilt from scratch.

Then I ask whether a current tool can help with that specific thing.

That is how I learned most of what I now do with AI. I didn’t sit down and decide to become an AI expert. I had a Python script that wasn’t working, asked for help, and slowly realized the tool could do more than answer one coding question. Later, I saw that an agent could help build a larger system. Each step came from a real problem sitting in front of me.

That kind of learning feels different because the reward is immediate. I’m not collecting a certificate for a future that may or may not come. I’m making Tuesday’s work a little clearer.

And I think this is where people over 50 may have an advantage, even if we’re slower to try the newest thing.

We usually have a better inventory of actual problems.

We know which report nobody reads. We know which missing detail causes the field visit to go sideways. We know which customer question sounds simple but is really pointing to something else. We know where the process breaks when the clean example in the training manual meets an actual person.

That doesn’t automatically make us good at AI. Experience is not a password that unlocks every new tool. We still have to practice, and we still have to be willing to look awkward while learning something.

But we don’t have to begin with a blank page. We can bring a worthwhile problem to the tool.

That may be enough of a late-career learning plan: keep one real problem in front of you, learn only as much as needed to improve it, and make sure another person can use what you learned.

The last part matters to me because I don’t want these five years to be only about staying employed. I want them to count for the people who will still be there after I leave.

If I use AI to make myself faster but everything still depends on me, I may have improved my output without improving the handoff. But if I use it to make a confusing process visible, turn repeated judgment into better questions, or give somebody a cleaner starting point, then the learning keeps working after I’m gone.

I also think there has to be a boundary.

Not every new system deserves my time. Some workplace technology is badly introduced. Some training is mostly a company checking a box. Some AI features create more review work than they remove. Being willing to learn does not mean pretending every rollout is wise.

At 57, maybe the useful skill is knowing when to lean in and when to let something pass.

Does this tool help with a problem I actually have? Will I use it often enough for the learning to stick? Does it help me make a better decision, explain something more clearly, or reduce work that never needed my judgment in the first place?

If the answer is no, I probably don’t need it.

If the answer is yes, five years is plenty of time for it to matter.

I’m trying to stop saying “only five years” as if those years are already spoken for. A child can go from middle school to graduating high school in five years. A person can get stronger, lose mobility, build a skill, repair a relationship, or drift a long way in five years.

It is a meaningful amount of life.

So I still want a retirement date. I still want to protect my health, prepare financially, and make the handoff real. But I don’t want the countdown to become permission to coast before I arrive.

Maybe the goal isn’t to keep up with everything. Maybe it’s to stay awake to the work that is still mine to do, solve the next useful problem, and keep learning until the season actually changes.

Anyway, that’s what I’ve been thinking about. If you’re within a few years of retirement, has that made you more selective about learning, or has it made you want to check out early? Leave me a note in the comments. Thanks for listening.

Video Prompt Script — Questions to Answer Without Reading

Use these as prompts. Don't read them on camera; answer them naturally.

1. Opening: When did you start thinking of age 57 as the beginning of a five-year countdown?
   • Follow-up: When did you notice that “five more years” could sound like a plan or like you were already leaving?
2. The real tension: How much should someone keep learning when retirement is visible but not immediate?
   • Follow-up: Why is chasing every new tool a bad use of limited energy?
3. The risk of checking out: What can change before a planned retirement date arrives?
   • Follow-up: How can mentally retiring early reduce your choices later?
4. What the research raised: Why might an older worker decide that learning AI is not worth the effort?
   • Follow-up: Why is the long-term effect of AI on older workers still uncertain?
5. Your practical learning pattern: How did one broken Python script lead you into using AI and agents for larger work?
   • Follow-up: Why did beginning with a real problem work better than trying to “learn AI” in general?
6. The over-50 advantage: What kinds of real workplace problems become easier to recognize after decades on the job?
   • Follow-up: Why does experience still require practice and humility with a new tool?
7. The handoff: How can late-career learning help another person instead of only making you faster?
   • Follow-up: What would it mean for the learning to keep working after you leave?
8. Boundaries: What questions help you decide whether a new tool deserves your attention?
   • Follow-up: When is refusing a tool good judgment rather than resistance to change?
9. Closing: What do you want the next five working years to contain besides simply staying employed?

Title Options

1. Five Years Is Still a Career
2. Don’t Retire in Your Head Before You Retire for Real
3. How Much Should You Keep Learning Before Retirement?

Thumbnail / Onscreen Text Options

• 5 YEARS IS NOT NOTHING
• DON’T CHECK OUT EARLY
• STILL LEARNING AT 57

Shorts / Reels Cutdowns

• “Five years is not a long weekend” — the opening contrast between a responsible retirement countdown and mentally standing at the exit.
• “Don’t learn AI. Bring it a real problem.” — Jay’s path from one broken Python script to learning tools through useful work.
• “Selectivity is not disengagement” — three practical questions for deciding whether a new workplace tool deserves limited late-career attention.

Viewer Question

If you are within five years of retirement, what is one thing you still want to learn, build, or pass on before you leave?