Calvin's Updates

The Latchkey Club Daily Draft — September 7, 2026

Teleprompter / Blog Script

I was looking at some of the work I’ve collected over the years, and I had a slightly uncomfortable thought.

If I left tomorrow, there are probably plenty of files somebody could open.

I’m not sure there are enough people who would know why the files say what they say.

Welcome back to the channel, guys. Today I wanted to talk about retirement, mentoring, and the difference between documenting your work and actually preparing somebody else to carry it.

I’m 57, so retirement is not an abstract idea anymore. The exact date can still move, and there are financial and family questions to work through, but I can see the runway. That has changed the way I think about my job.

For a long time, being useful meant being the person who could solve the problem. Somebody called, something was not working, and I knew where to look. That feels good. It also makes the day move faster because explaining everything can take longer than fixing it yourself.

But if I keep being the only person who fixes it, I may be solving today’s problem while creating a larger one for later.

A procedure can tell you which button to press. Experience tells you when not to press it.

A diagram can show how something is connected. It may not tell you that the layout changed after a customer complaint, or that one location has an exception because of a decision made eight years ago, or that the person you need to call will answer much faster if you explain the history first.

That information usually does not live in the official file. It lives in stories, relationships, scars, and small warnings that begin with, “Before you do that, you should know…”

I read a Deloitte article recently about the amount of knowledge expected to leave workplaces as older employees retire. One finding stood out: 92 percent of the organizations in the survey were not consistently capturing knowledge from people who were approaching retirement.

That is a corporate statistic, but the reason behind it feels very ordinary.

Most of us are busy doing the work. Documentation gets postponed because the problem in front of us is louder. Then somebody announces a retirement date, a folder gets created, and everyone starts trying to download thirty years of judgment in the last six weeks.

That is probably too late.

And I don’t think the answer is documenting everything. Nobody wants a thousand-page archive called “Everything Jay Knows.” I would not want to read that, and I’m Jay.

The better question is: where would the team become stuck if I disappeared for a month?

Which customers depend on a relationship that only I have? Which decisions require history that is not written down? Which recurring problems look simple until the normal solution does not work? Which part of the job does everybody assume I will handle because I always have?

That is the knowledge at risk.

AI can help with some of this. I can record a conversation, have the tool organize it, pull out the decisions, draft a procedure, create a checklist, or identify places where I used a term that a new person would not understand.

That can remove a lot of the friction from the first draft.

But AI cannot decide by itself which detail is actually important. It can make a clean document out of a bad explanation. It can summarize the official process while missing the exception that causes the real trouble. And if the source material is scattered, outdated, or wrong, AI can organize the confusion faster.

So I think the tool belongs in the middle of the handoff, not at the center of it.

The center still has to be two people doing real work together.

I’ve been thinking about a simple test: before I retire, somebody else should do the job while I’m still available to watch.

Not watch me do it one more time. They do it.

They lead the customer conversation. They make the recommendation. They build the document. They respond when the first plan fails. I sit close enough to stop a serious mistake, but far enough back that they have to think.

That will probably be slower. It may be uncomfortable for both of us. I’ll notice a shortcut and want to jump in. They may know I’m sitting there and keep looking at me for confirmation.

But that moment is where the transfer happens.

When they ask, “Why did you choose that?” I have to explain the judgment, not only the step. When they choose something different, I may find out their way is better. That matters too. A handoff should not turn the next person into a copy of me.

People coming behind us know things we don’t. They may understand a newer tool, see an old workaround that is no longer needed, or ask a question we stopped asking years ago. Mentoring should move both directions. I can give them history and judgment. They can show me where history has become habit.

That may be one of the better uses of AI in a mixed-generation team.

The experienced person can talk through the messy background. The newer person can challenge assumptions and use current tools to turn that conversation into something searchable and usable. Then both people review it against the real work.

The result is not a digital museum of how we used to do things. It is a living guide that the next person has already tested.

I also think relationships have to be transferred before responsibilities are.

If I have been the main contact for somebody for years, an email announcing my replacement is not much of a handoff. The next person needs time in the room. They need to hear the concern directly, understand the history, and become a familiar voice before I disappear.

Trust cannot be uploaded with the project files.

So part of my job now is making sure other people are included earlier. Let them answer. Let them lead. Let the customer learn that the team is capable, not only the person who has been around the longest.

There is some humility in that.

Being needed can feel a lot like being valuable, but they are not always the same thing. If everything stops when I leave, that may prove I was important. It may also prove I did not prepare anyone.

I would rather leave behind people who are confident than a phone that keeps ringing after I retire.

For anybody our age who can see the end of the runway, maybe the first step is not writing a giant manual. Pick one responsibility that depends too much on you.

Choose someone. Bring them into the next real situation. Let them try. Ask what was unclear. Use AI to capture the explanation and turn it into a first draft. Then have them use that draft without you and tell you where it fails.

Do that while there is still time for a second attempt.

Because the goal is not to preserve every sentence we ever said at work. The goal is for good judgment, useful history, and strong relationships to continue helping people after our name comes off the calendar.

Anyway, that’s what I’ve been thinking about. If you left your job for a month, what is the one thing people would still have to call you about? 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: What did you notice when you looked at the files and systems you have built over the years?
   • Follow-up: How much of the “why” would still be missing if somebody only had the documents?
2. Retirement runway: How has seeing retirement as a real horizon changed the way you think about being useful at work?
   • Follow-up: Why is solving every problem yourself efficient today but risky later?
3. Steps versus judgment: What can a procedure capture, and what usually remains inside experience?
   • Follow-up: Which exceptions, relationships, or old decisions are easy to miss?
4. Find the knowledge at risk: Where would your team get stuck if you disappeared for a month?
   • Follow-up: Which work depends too heavily on you because everyone assumes you will handle it?
5. AI’s proper role: How can AI turn conversations into drafts, checklists, and searchable notes?
   • Follow-up: Why can it also organize a bad or incomplete explanation too confidently?
6. Let somebody else try: Why does the successor need to perform the real job while the experienced person is still present?
   • Follow-up: What makes it hard to watch someone take a slower or different route?
7. Two-way mentoring: What history and judgment can the experienced worker give?
   • Follow-up: What can a newer worker teach about current tools or habits that no longer make sense?
8. Transfer relationships: Why is introducing a replacement by email not enough?
   • Follow-up: How can another person earn trust before the handoff becomes official?
9. Needed versus valuable: Why can being indispensable feel good but reveal a failure to prepare others?
   • Follow-up: What would it look like to leave behind confident people instead of permanent dependence?
10. Closing: If you left for a month, what is the one responsibility people would still have to call you about?

Title Options

1. Before You Retire, Let Somebody Else Try
2. Your Work Files Don’t Contain What You Really Know
3. The Real Retirement Handoff Starts Years Before You Leave

Thumbnail / Onscreen Text Options

• COULD THEY DO IT WITHOUT YOU?
• YOUR FILES AREN’T THE HANDOFF
• LET SOMEBODY ELSE TRY

Shorts / Reels Cutdowns

• “A procedure tells you which button to press” — experience tells you when not to press it, which is why the reasons and exceptions matter more than another giant manual.
• “Don’t watch me do it again” — real knowledge transfer starts when the next person leads the work while the experienced person is still close enough to coach.
• “Trust cannot be uploaded” — project files can move instantly, but customer relationships need to be transferred through shared conversations before retirement.

Viewer Question

If you left your job for a month, what is the one thing people would still have to call you about—and who should be learning it now?