Calvin's Updates

The Latchkey Club Daily Draft — August 5, 2026

Teleprompter / Blog Script

I was watching an AI demonstration recently, and the person presenting it was moving pretty fast.

Open this window. Connect that account. Change this setting. Type the prompt this way. Then the answer appears, everybody nods, and we move to the next feature.

If you already knew the tool, it probably looked simple.

If you missed one click about thirty seconds earlier, you were now watching somebody perform a card trick.

Welcome back to the channel, guys.

Today I wanted to talk about something I think is happening to a lot of people in their fifties and sixties at work. We are being introduced to AI at the same time many companies are deciding what experience is still worth, and those two things are getting mixed together.

Somebody can be new to an AI tool and suddenly look like they are new to the work.

Those are not the same thing.

I’m 57, and I use AI every day. I use it for research, organizing information, building software, documenting processes, and turning ideas into things I can actually test. I’m not standing outside the technology complaining that somebody moved the door.

But I also know what it feels like to open a new system and not know where anything is. The buttons have unfamiliar names. The person teaching it assumes you understand three other tools first. Everybody else seems to be moving faster, although half of them may just be better at looking confident on a video call.

That beginner feeling can be uncomfortable when you have spent thirty years becoming competent.

At 25, nobody is surprised when you ask where the setting is. At 57, you may wonder if the question will be treated as evidence that you cannot keep up.

So people get quiet. They stop the presenter less often. They write down a few words and tell themselves they will figure it out later. Then the company says the older workers are not adopting the technology.

Maybe some are resisting it. That happens.

But maybe the training was built for people who already knew the language. Maybe the demonstration showed a stunt instead of a real job. Maybe nobody explained which part of the person’s existing knowledge still matters after the tool arrives.

A recent analysis from the Center for Retirement Research looked at workers 55 and older in jobs with different levels of AI exposure. The early result was that people in more exposed occupations had larger relative increases in leaving work than people in less exposed occupations.

The author was careful about what that means. AI can remove a job task, but it can also make somebody more productive and possibly extend a career. We are still early in this.

But the possibility of experienced people leaving work because the technology changed around them should get our attention.

Not because every older worker has to stay forever. Some people are ready to retire, and they have earned the right to make that decision.

The problem is when somebody leaves because they were made to feel obsolete before anybody gave them a fair way to become useful with the new tool.

I keep thinking about something my own team told me when we were talking about training. They wanted the foundation. Start from zero. Do not assume everybody already understands the system just because they have been around the work.

That was good feedback for me because experience creates blind spots in the teacher too. I can skip six steps without realizing they were steps. Then I mistake the learner’s confusion for a lack of ability when the problem is that I started in the middle.

AI training often starts in the middle.

It starts with prompts, agents, models, automations, and a screen full of options. But the older worker may be asking a more basic and more important question: What problem is this supposed to help me solve?

That question is not resistance. That may be judgment.

People who have lived through enough technology rollouts know that a good demonstration and a good working system are different things. We have seen software that promised to save time create three new approval steps. We have watched a simple process become a dashboard nobody checks. We know the person who clicks the button may still own the mistake after the vendor leaves.

So caution is not always fear. Sometimes it is memory.

At the same time, experience cannot become an excuse to refuse every new tool. I have to be honest about that too.

Knowing the old process does not give me permanent authority over the future. If a new tool can do part of the work better, I need to learn enough to understand it. I cannot keep saying, “That is not how we do it,” when what I really mean is, “I do not want to feel like a beginner again.”

Humility has to run in both directions.

The person who knows AI should not assume speed with the interface equals wisdom about the work. And the experienced person should not assume years in the job make every old habit worth preserving.

The useful part happens when those two people can work on one real problem together.

Not a generic exercise about writing a poem or planning a pretend vacation. Pick something that actually causes friction on Tuesday afternoon.

Take a messy report that has to be organized every week. Take a customer question that requires information from three systems. Take a repeated documentation task where the first draft takes two hours. Use approved data, keep the risk low, and let the experienced person explain what a correct result needs to contain.

Then let the person who knows the tool show what AI can do.

Now both sides have something to contribute.

The newer user can say, “That answer looks polished, but it missed the exception that causes the real problem.”

The faster user can say, “We can build that exception into the instructions and test it against three old cases.”

That is a different kind of training. Nobody is pretending the tool is magic, and nobody is pretending experience makes learning unnecessary.

I think people over 55 may have a hidden advantage if we approach AI this way. We have enough history to test the output against consequences. We may remember why a rule exists, which customer promise cannot be broken, or what happened the last time somebody optimized the number without understanding the system around it.

But that advantage only becomes visible when we get our hands on the tool.

Experience trapped outside the new workflow looks like resistance. New technology without experience inside it produces mistakes faster.

So if I were learning AI at work right now, I would keep it small.

One real task. One measure of whether it helped. One explanation of what I corrected and why. Then do it again until I can teach somebody else.

And if I were teaching it, I would slow down enough to let people ask the question they think they should already know. I would explain the purpose before the buttons. I would stop using the fastest person in the room as the standard for everybody else.

Because the goal is not to make an experienced worker look young.

The goal is to help a capable person stay capable as the tools change.

There is a difference.

Anyway, that’s what I’ve been thinking about.

Have you ever been treated like you did not understand the work just because you were new to the tool? Or have you seen an older coworker become much more useful once somebody taught the technology around a real problem?

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 the fast AI demonstration feel like once you missed one early step?
   • Follow-up: Why can a simple demo look like a card trick to a new user?
2. The distinction: Why is being new to an AI interface different from being new to the work?
   • Follow-up: Why does asking a basic question feel riskier at 57 than it did at 25?
3. What silence looks like: Why might experienced workers stop asking questions during training?
   • Follow-up: How can that silence be misread as resistance or inability?
4. The research signal: What did the Center for Retirement Research find about AI exposure and work exits among people 55+?
   • Follow-up: Why should we avoid treating that early association as proof of one simple cause?
5. Starting from zero: What did your team teach you by asking for foundational training?
   • Follow-up: How can an experienced teacher accidentally start in the middle?
6. Caution versus fear: What has Gen X learned from earlier technology rollouts, dashboards, and vendor promises?
   • Follow-up: When is caution actually memory of consequences?
7. Humility in both directions: What does the AI-fluent person need from the experienced worker, and what does the experienced worker need to admit?
   • Follow-up: How can both kinds of pride damage the work?
8. One real task: What recurring Tuesday-afternoon problem would make a better training exercise than a flashy AI demo?
   • Follow-up: How would you measure whether the tool actually helped?
9. Close: What would change if the goal were not to make older workers look young, but to help capable people remain capable as tools change?

Title Options

1. Don’t Confuse Being New to AI With Having Nothing Left to Offer
2. The Most Experienced Person Can Still Be an AI Beginner
3. Stop Teaching AI Like Everybody Already Knows the Language

Thumbnail / Onscreen Text Options

• NEW TO AI ≠ NEW TO THE WORK
• START FROM ZERO
• EXPERIENCE STILL COUNTS

Shorts / Reels Cutdowns

• “You missed one click, and now it’s a card trick.” Use the opening demo to show why fast training can make a capable person feel lost.
• “Caution is sometimes memory.” Cut the section about failed technology rollouts, extra approval steps, and owning the mistake after the vendor leaves.
• “One real task.” Explain the practical learning method: choose one recurring problem, define a correct result, test AI on old cases, and teach the lesson to somebody else.

Viewer Question

Have you ever been treated like you did not understand the work simply because you were new to the tool?