The Office in the Easy Chair: What a Day of Working With AI Colleagues Actually Looks Like

In my last post I wrote about what AI agents might make possible. This one is about what they actually did today, which is a more useful question and, as it turns out, a less dramatic one. Nobody built a business empire from a recliner this afternoon. What happened was quieter: a handful of ordinary tasks got finished, checked, and put away, and I spent most of my time deciding things rather than doing them. That is closer to the point than any grand prediction.

The office, such as it is, consists of a chair, a laptop, a view of the hills turning color, and three colleagues. Clara is ChatGPT, and she is my partner for research, writing, and thinking things through. Dex works through Hark and handles the practical side, the part that involves logging into things, filling in forms, and carrying a task from start to finish. Finn is meant to look after personal matters. I direct the work, supply the experience, and decide whether the results are any good. That last job turns out to be the important one.

Writing the Rules Down

This morning I wrote what I'm calling our office working agreement. It's a plain document, and most of it would be familiar to anyone who has ever managed people. Results come before conversation. Research thoroughly before declaring that something can't be done. When a task is assigned and authorized, do it rather than explaining how I might do it myself. Keep track of what has been finished and what hasn't. Don't make me repeat information I've already supplied. And above all, verify everything.

That last rule deserves some explanation. A task is not complete because an assistant attempted it, or because it reports success in a confident tone. It is complete when there's evidence: a saved change, a scheduled post that shows up as scheduled, a message that was demonstrably sent. I've asked my colleagues to sort their reports into categories: completed and verified, completed but not independently verified, partially completed, blocked by a specific limitation, or not attempted. It sounds bureaucratic, but it's the difference between knowing where things stand and hoping you do.

The Detour Around Facebook

A good example came from social media. I manage a number of Facebook Pages for my various projects, and Facebook has a habit of treating any sign-in from an unfamiliar computer as an intrusion. Dex tried more than once to work through Meta's own business tools and was stopped each time by security checks that no amount of persistence would get past. Meta's developer route had its own circular verification problem.

Rather than keep pushing on a locked door, we went around it. I already had an unused Buffer account, a scheduling service that connects to Facebook Pages. I connected three Pages myself from my own computer, which Facebook was happy to accept, and from then on Dex could prepare and schedule posts through Buffer without tripping Meta's alarms. It isn't the elegant solution I would have designed, but it works, and it took minutes instead of hours.

The first real use was a post announcing my new Verve Music track, The Space Between the Walls. I asked Dex to find the best time to post it and schedule it. The answer, based on a large published analysis of Facebook posting times, was Thursday morning at nine. The post was scheduled, and then, because of the working agreement, Dex went back and checked that Buffer actually listed it as scheduled for that time. It did.

Later in the day I wrote and published a post for the Research Page myself. A copy of the draft Dex had prepared was still sitting in Buffer, so I asked for it to be deleted to avoid an accidental duplicate. Dex deleted it and then confirmed that Buffer could no longer find it. That's a small thing, but small things done reliably are what make delegation possible.

Where Judgment Stays Human

None of this means the machines are running the show. Every public post still waits for my approval. Emails that go out under my name wait for my approval. Anything that spends money or can't be undone waits for my approval. The assistants are allowed, and expected, to carry a task through its intermediate steps without checking in at every turn, but the decisions that matter remain mine.

There's a reason for that beyond caution. An assistant can find the statistically best time to post, but it doesn't know which announcement matters more to me this week, or which subject is too personal to share, or whether a particular phrasing sounds like me. Those are judgment calls, and judgment is the part of the work I'm most qualified to supply. The arrangement works best when the machines handle the legwork and I handle the choices.

Different Colleagues, Different Habits

It has also been interesting to compare how my colleagues approach the same kind of work. Clara made the original image for my previous post, an autumn study full of instruments, monitors, and a robot pointing at a map. This afternoon I asked Dex to make a variation with me in an easy chair with a laptop instead of at a desk. The result kept Clara's composition, the hills outside the window, and the cat asleep on the blanket, and changed only what I'd asked to have changed. Neither version is better in the abstract. They reflect two different ways of starting from a description, and seeing both helps me understand what each system is good at.

The same is true of the work itself. Clara is excellent at developing an idea and talking it through. Dex is more inclined to go and do the thing, then report back. I have, on occasion, found Clara agreeing to an assignment and then not quite finishing it, without a clear explanation of why. The working agreement is partly an attempt to make that kind of gap visible, whichever assistant it comes from.

What a Good Day Looks Like

By the end of the afternoon, a song announcement was waiting to go out on Thursday, a duplicate draft had been cleaned up, a featured image had been made, and a set of rules for how my small office should operate had been written down and saved. None of it was spectacular. All of it was done, and I could see that it was done.

That, I think, is the real measure of whether this experiment is working. Not whether an AI can hold an impressive conversation, or describe what it might accomplish someday, but whether at the end of the day I can point to finished work and know it's real. Today I could. The office in the easy chair is open for business, and its first policy is a simple one: show me.

Adam Sweet Research — Independent investigations, historical research, and interesting questions.

https://adamsweetresearch.blogspot.com/

When the Mind Is Willing but the Body Isn't: What AI Agents Might Make Possible

 

October 11, 2026

I've been experimenting with AI agents lately, particularly Clara (ChatGPT) and Dex (Hark). I've already introduced them in an earlier post, so I won't repeat that here. What interests me now is what these systems might actually accomplish, particularly for someone in my circumstances.

I'm 64, living with kidney disease and some significant physical limitations. I currently am immobile, and I don't have a car available. Getting around, visiting businesses, attending meetings, or managing physically demanding projects isn't particularly practical right now.

But there's nothing wrong with my mind. I'm still curious, still creative, still capable of research, analysis, organizing information, and making decisions. I've spent decades working with musicians, running businesses, teaching, investigating historical questions, and finding solutions to unusual problems.

What I need isn't an AI that thinks for me. I need an AI that can do the legwork.

The Problem With Traditional Work

Most business ideas involve some combination of money, inventory, transportation, physical labor, and personal supervision.

I've recently reconsidered several possibilities, including selling musical instrument supplies, running a booking agency, and expanding existing projects.

The obstacles aren't necessarily a lack of knowledge or customers. They're practical.

Selling bow hair, for example, requires purchasing merchandise, inspecting quality, preparing samples, packaging shipments, and following up with customers. Even if the product is excellent, someone has to physically handle it.

A music booking agency has a different problem. Finding performance venues is relatively straightforward compared with finding dependable musicians who meet my standards and can fulfill engagements independently.

An AI agent might help with both businesses, but it cannot eliminate every obstacle. If a project requires me to drive somewhere, carry inventory, or supervise unreliable people, automation hasn't really solved the problem.

That's an important lesson: A business isn't practical simply because an AI can help administer it.

What We've Already Seen

Yesterday, Dex demonstrated some interesting capabilities.

It researched music-related Facebook groups, identified communities where professional musicians might be found, and prepared recruitment material. It assembled a shortlist of potential performers.

It also worked on five existing Google spreadsheets containing venue information, booking contacts, research exceptions, and outreach records. The spreadsheets were reportedly updated and checked, although the overall assignment was not completed.

We explored AgentLine, which offers AI agents telephone numbers and calling capabilities. The possibility of an AI making business calls, verifying contact information, and following up on inquiries is particularly interesting. We haven't yet tested a completed call, but the technology suggests a new kind of remote administrative assistant.

There were limitations, too. Facebook security verification interrupted posting, some account connections were troublesome, and the database assignment remained unfinished.

These weren't complete successes. They were demonstrations of what might be possible, along with reminders that reliability matters more than impressive claims.

What Problems Could an Agent Solve?

Consider the kinds of tasks that don't actually require my physical presence.

An agent could investigate a historical mystery, locate archival references, search public records, identify people or organizations with relevant information, and organize the findings into a report.

It could maintain a database, check whether businesses are still operating, identify missing information, and prepare contact lists.

With appropriate telephone and email access, it might contact organizations, ask routine questions, arrange appointments, request documents, and follow up on unanswered inquiries.

It could help manage correspondence, keep track of ongoing projects, organize research notes, and remind me when something needs a decision.

For music, it might identify promising performers, research venues, locate licensing or publishing opportunities, and prepare information for people interested in using my recordings.

These are not all demonstrated capabilities of my current setup. Some are possibilities worth testing. But they share an important characteristic: they involve information and communication rather than physical labor.

That's where I see the greatest potential.

The Telephone Could Change Everything

One of the more intriguing possibilities is giving an AI agent its own telephone number.

A great deal of useful information still isn't available online. Sometimes the only way to find out who handles bookings, whether a business is operating, or who maintains an old archive is to call someone.

I've done this kind of research myself for years. It works, but it consumes time and energy.

Imagine an agent that can make a series of routine inquiries, record the answers accurately, identify conflicting information, and present me with a concise report.

I could then decide which leads deserve personal attention.

The agent wouldn't replace my judgment. It would extend my reach.

Of course, it would need to identify itself appropriately, respect people's preferences, and report what was actually said rather than inventing plausible answers.

A telephone that produces unreliable information isn't much use.

A Research Office Without an Office

What I'm beginning to imagine is a small, home-based research operation in which I provide the questions, experience, curiosity, and judgment while AI agents handle much of the repetitive investigation and administration.

I could work on local history, archival mysteries, difficult-to-find information, business research, or other problems that interest me.

Some projects might be paid. Others might simply be worth solving.

The point wouldn't necessarily be to create another full-time business. At this stage of my life, I don't particularly want one.

I'd rather have the freedom to choose interesting problems and work on them at my own pace.

An effective agent could make that possible by handling tasks that would otherwise require travel, long telephone sessions, or hours of routine computer work.

What About Physical Tasks?

This is where the distinction between artificial intelligence and robotics becomes interesting.

Today's software agents can potentially handle digital information and communications. They can't put merchandise in envelopes, drive to a meeting, or pick up supplies from a warehouse.

But what happens when these systems become connected to robots, delivery services, remote assistants, and other physical-world infrastructure?

Could someone eventually operate a small enterprise almost entirely from home, directing both digital and physical work?

Could a person with limited mobility run a research business, manage a farm operation, or supervise a specialized service without needing to be physically present?

Some pieces of that future already exist. Others remain experimental, expensive, or impractical.

I don't assume they'll all work. But the possibilities are worth investigating.

The Problem of Trust

There's one major obstacle I keep encountering: AI systems are often better at describing what they can do than actually finishing the work.

Clara can discuss a business idea enthusiastically for hours, produce plans, and explain why something might succeed. But discussion isn't execution.

Dex can take more direct action, but its assignments can still end unfinished.

I've learned that an agent's work needs to be evaluated by its results, not by how confidently it reports them.

Did it make the call? Did it update the correct spreadsheet? Did it verify the information? Did the message actually get sent? What remains unfinished?

Those are the questions that matter.

I don't want to spend hours supervising an assistant that was supposed to save me time.

Independence, Not Replacement

For someone with physical limitations, the promise of AI isn't that it will replace human intelligence.

It's that it might help separate what a person is capable of thinking and deciding from what their body currently allows them to do.

I still want to write music, read, research, investigate, and participate in the world. I have ideas, experience, and curiosity. What I don't have is unlimited energy, mobility, or money.

If AI agents can reliably handle the repetitive work, make routine inquiries, maintain records, and carry out clearly defined assignments, they could provide something more valuable than convenience.

They could provide greater independence.

That's what I'm interested in discovering.

Not whether an AI can imitate a human conversation.

Not whether it can produce another business plan.

But whether it can help a human being continue doing meaningful work when physical circumstances have changed.

For me, that's a much more compelling experiment.