Calvin's Updates

The Latchkey Club Daily Draft — September 13, 2026

Teleprompter / Blog Script

I had several AI agents running the other morning.

One was researching. One was working on code. Another was sorting through a different problem in the background. For a few minutes, it felt like I had figured out how to be in four places at once.

Then the results started coming back.

Now I had four things to read, check, correct, and decide what to do with. The machines had finished their part faster. Somehow I was more behind.

Welcome back to the channel, guys. Today I wanted to talk about something I’m beginning to notice with AI: a tool can save time without making the day feel any less full.

I use AI a lot. I’m not standing outside this conversation warning everybody about a technology I refuse to touch. I use it to help build software, research questions, organize information, and keep some of the loose ends of work and family from disappearing.

It can be extremely useful.

But usefulness creates its own temptation. When a task that took an hour now takes twenty minutes, I rarely sit there and admire the forty minutes. I find another task.

Then another one.

Pretty soon, the reward for becoming more efficient is a day with more work packed into it.

I read about an eight-month study at a technology company where employees had broad access to generative AI. The researchers found that people worked faster, took on a wider range of tasks, and let work spread into more hours of the day, often without anybody directly ordering them to do it.

AI made it easier to begin things. People sent prompts during lunch, before meetings, and in the evening. They kept several processes moving at once. The day became denser.

That word feels right to me. Denser.

It may still end at the same time on the clock, but there are fewer spaces inside it. While one agent is running, I start something else. While that result loads, I answer a message. During a meeting, another process finishes and puts a little notification on the screen.

Nothing is technically forcing me to look.

I still look.

For Gen X workers in their 50s, this lands at an interesting time. A lot of us are trying to make it through the last five, eight, or ten years of full-time work. We may be paying for college, helping family, watching retirement accounts, managing a body that has started sending more detailed status reports, and trying to stay current while the workplace changes again.

We are not short on responsibility.

AARP has cited research showing more than half of Gen X workers dealing with moderate to high burnout. And if you look at Gen X conversations online, the phrase that keeps showing up is not always, “I can’t wait to retire.” It is, “I’m tired.”

Those are related, but they are not exactly the same.

Some people genuinely need to leave a damaging job. Some cannot afford to leave yet. Some are not done with meaningful work; they are done with the pace, politics, interruptions, or constant feeling that the finish line moves every time they get close.

Then AI arrives promising to reduce the burden.

If we are not careful, it may simply help us reach the moving finish line faster so somebody can move it again.

I don’t think the answer is to stop using the tool. For me, that would give up too much real help. The better question is: who gets the time it saves?

Does the company get all of it as extra output? Do I give it to three projects I would not have attempted before? Does it become another hour at night because the agent can keep working after I should have stopped?

Or can some of that capacity remain unclaimed?

That feels almost irresponsible to say. We grew up believing that if you had time, you found something useful to do. At work, being the person who could carry more usually made you valuable. If a new tool increased your capacity, of course you used the capacity.

But at 57, I’m not sure maximum capacity is the right target every day.

A network run at maximum capacity has no room for a spike. A calendar run at maximum capacity has no room for a family problem. A body run at maximum capacity eventually sends an invoice.

Maybe efficiency needs a reserve.

That reserve is where judgment happens. It is where I notice that the clean AI answer solved the wrong problem. It is where I remember the person affected by the decision. It is where I can finish one thing before opening five more tabs.

It is also where lunch goes.

That sounds small, but the Berkeley researchers specifically noticed that AI was filling moments that used to be pauses. A prompt takes only a minute, so it seems harmless to start one while eating. But now part of my attention is waiting for the result. The break has become a lightly supervised work session.

I know this because I do it.

The machine does not get tired, but supervising machines is still human work. Reading output is work. Comparing sources is work. Deciding which result matters is work. Catching the confident mistake before it travels downstream is definitely work.

Running four agents does not give me four brains. It gives my one brain four incoming trays.

I think people our age may have an advantage if we use it. We have lived through enough productivity revolutions to know that saved labor does not automatically become rest. Email was supposed to speed communication. Then we got more email. Smartphones made us reachable from anywhere. Then anywhere became a place we could work.

AI will probably follow the same path unless we make a deliberate choice.

So I’m trying to develop a few rules.

First, before I hand a task to AI, I want to know what the saved time is for. If the answer is just “more,” that is not much of an answer. Maybe the purpose is to leave work on time, protect a concentration block, take a walk, make one careful decision, or be available when somebody at home needs me.

Second, I don’t need every process running at once. Parallel work looks impressive on a screen, but each result creates a decision. Sometimes the calmer approach is one agent, one finished review, then the next task.

Third, a faster first draft should not create a faster final decision when the stakes are high. AI can shorten the assembly. It should not eliminate the pause where experience asks, “What are we missing?”

And fourth, I need an end-of-day rule that the tool cannot negotiate with. Agents can keep running. Research can keep expanding. There is always one more improvement available. If stopping depends on the work being completely finished, I will not stop.

This matters beyond the current job.

If I practice filling every recovered minute now, retirement will not automatically teach me how to rest. I may leave the employer and rebuild the same pressure with projects, side income, volunteer work, home repairs, and technology that lets me manage all of it from my phone.

A different calendar can hold the same habit.

I don’t want retirement to be the first time I learn that an open hour does not need to defend its existence.

That does not mean becoming passive. I want to stay useful. I want to build things, help people, keep learning, and use the experience I have while I still can.

But useful and continuously occupied are not the same thing.

Maybe the mature use of AI is not seeing how many jobs I can make it do. Maybe it is choosing the work that matters, letting the tool reduce the strain, and refusing to spend every saved minute before it arrives.

Anyway, that’s what I’ve been thinking about. When technology saves you twenty minutes, who usually gets those twenty minutes? 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 happened when you had several AI agents working at the same time?
   • Follow-up: How did faster machine output create a larger pile of human decisions?
2. The strange result: When a one-hour task becomes a twenty-minute task, what usually happens to the forty minutes?
   • Follow-up: Do you experience the savings, or immediately replace them with more work?
3. A denser day: How can prompts during lunch, between meetings, or at night remove pauses without formally extending the schedule?
   • Follow-up: Why does a notification pull attention even when nobody requires an immediate response?
4. The Gen X timing: What responsibilities are many people in their 50s carrying while trying to reach retirement?
   • Follow-up: What is the difference between wanting to stop all meaningful work and simply being tired of the current pace?
5. Who gets the savings: Does AI-created capacity belong automatically to the employer or the next project?
   • Follow-up: What would it mean to leave some capacity unclaimed?
6. Capacity needs reserve: Why is maximum capacity a risky target for a network, calendar, or body?
   • Follow-up: What kinds of judgment and human attention happen only when some room remains?
7. Supervision is work: What mental work remains after an agent produces an answer?
   • Follow-up: Why do four agents still feed into one human brain?
8. What Gen X has seen before: What happened when email and smartphones made communication faster and people more reachable?
   • Follow-up: What boundary does AI require that those earlier tools did not create for us automatically?
9. Practical rules: What should the saved time be for before you automate a task?
   • Follow-up: When should you run fewer processes, preserve a pause, or keep a fixed stopping time?
10. Closing: What habit around busyness could follow you into retirement even after the employer is gone?

• Follow-up: When technology saves you twenty minutes, who do you want to receive them?

Title Options

1. Don’t Let AI Spend the Time It Saves
2. AI Made Me Faster—and My Day Got More Crowded
3. The Hidden Cost of Running Everything at Once

Thumbnail / Onscreen Text Options

• WHO GETS THE SAVED TIME?
• FASTER. BUT MORE TIRED.
• YOUR DAY GOT DENSER

Shorts / Reels Cutdowns

• “Four agents, one brain” — AI can run multiple tasks in parallel, but every result still lands in one human decision queue.
• “Maximum capacity needs a reserve” — a network, calendar, or body with no spare capacity has no room for the unexpected.
• “A different calendar can hold the same habit” — leaving a job will not automatically cure the reflex to fill every open hour with work.

Viewer Question

When technology saves you twenty minutes, who usually gets those twenty minutes—and who would you like to get them instead?