> ## Content Index
> Fetch the complete content index at: https://operators-ai.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# Will AI replace your business?
- URL: https://operators-ai.ghost.io/will-ai-replace-your-business/
- Published: 2026-08-14T15:49:28.000Z
- Updated: 2026-08-14T15:49:28.000Z
- Description: A viral article asked whether AI will take your job. Nobody answered it for the people who own the job. Let's fix that.
- Author: Justin Zsimovan
- Tags: lessons

Yesterday I made a decent chunk of reseller X mad at me.

I posted that AI is going to eliminate basic online arbitrage flipping, a style of reselling where you find discounted products on retail websites and resell them on Amazon. A lot of people run it as a side business or a full income, and the replies came in hot. One guy told me he'd start worrying "when I see AI pushing a cart full of toys in store." Another wrote a long, genuinely thoughtful reply that boiled down to: these retail sites are already scraped to death by software, so what exactly does AI change?

Hold that question, because the answer applies to a lot more businesses than reselling, and it deserved more than the 280 characters I had to work with.

The same morning, [blocmates published a long research piece](https://x.com/blocmates/status/2088260925501440500?ref=operators-ai.ghost.io) asking "Will AI Replace Your Job?" It's one of the better roundups I've read on the topic. They pulled together layoff numbers, adoption studies, and cost data to answer the question everyone with a W-2 is quietly asking. It's worth your time.

Here's the thing though. That article, like almost everything written on this subject, is aimed at employees and venture-backed founders. Nobody writes the version for the person who owns a small business. You don't have a boss who decides your fate in a quarterly planning meeting. You ARE the quarterly planning meeting. So today I want to distill what that article found, and then translate it for us.

## The problem

Start with the scary numbers, because they're real. Tech layoffs crossed 123,000 roles this year by Forbes' count. Jack Dorsey said out loud what most CEOs won't: Block's 4,000 cuts weren't about Covid over-hiring or financial trouble, they were about AI doing the work. Microsoft, Amazon, and Oracle have all taken big swings.

Now the part that doesn't fit the doom narrative. blocmates cites a Ramp and Revelio study showing AI is NOT the primary driver of the current layoff wave. Companies adopting AI most intensely actually increased hiring by 12 percent. Some companies that cut staff over AI are quietly re-hiring. And only 2 percent of businesses have AI integrated deeply enough that it runs critical workflows.

So both crowds on X are wrong. The "AI is taking everything" doomers are early. The "nothing ever changes" crowd is worse than early. They're reading today's snapshot as a forecast.

The line in the article that matters most is this one: AI doesn't replace professions. It replaces tasks first. The jobs built around those tasks are what become vulnerable. The research calls these "connector roles," jobs that mostly move information from one place to another, follow established processes, and produce standardized outputs.

Employees get to stop there and ask whether their job title is on the list. You and I don't have that luxury, because here's the uncomfortable translation for owners: a small business can be a connector role too. If the thing your customers pay you for is mostly moving information from one place to another, comparing, checking, forwarding, and formatting at a screen, then your whole company sits where the most exposed jobs sit. It just has an LLC wrapped around it.

That describes more businesses than most owners want to admit. The bookkeeping practice that's mostly categorizing transactions and producing monthly statements. The freight broker matching loads to carriers. The travel agent assembling itineraries from the same searches the customer could run. The content shop churning out product descriptions. The lead-gen service that scrapes, filters, and forwards lists. And yes, basic online arbitrage: pull a list of discounted products, compare against Amazon prices, check the sales rank, buy if the spread clears your threshold, ship, reprice, repeat. That's the business I poked at yesterday, and it's the one I know best, because I run a reselling operation myself.

## The honest answer to the best objection

Back to that reply, because the objection generalizes too. "These sites are already scraped to death by software, so what changes?" Swap in your own industry: quoting software already exists, load boards already exist, comparison sites already exist. Software has been "coming for" every information business for twenty years, and the operators are all still here. So what changes?

Here's what changes. Every tool in that twenty-year wave had a human bottleneck in the middle. The scraper finds the candidate deal, then a person evaluates it, decides, and commits the money. The load board surfaces the load, then a broker works the phones. The quoting software drafts the number, then someone reviews it and hits send. The software sped up one task. A human still closed the loop, and human speed, human hours, and human hesitation are why margins survived. Your competitors had the same tools you did, and they were just as slow as you were.

An agent doesn't hesitate, doesn't sleep, and doesn't check a few times a day. It closes the whole loop, evaluation and decision included, continuously. In my industry, that means the deal windows that used to survive hours because 200 humans checked their software a few times a day will get eaten in seconds by whoever runs the machine. In yours, it means the version of your service that's really "software plus a human closing the loop" gets done by the loop alone, at a price you can't match by working harder.

I'm not speculating from the sidelines. My own systems already sweep supplier sites on a schedule and surface deals with the math done before I've had coffee. The person that cooks isn't me. It's the operator doing the same sweep by hand, three hours a night, who doesn't know the race changed.

And to the gentleman with the cart full of toys: you're right, and you're accidentally making my favorite point. Nothing is pushing a cart through a store clearance aisle any time soon. The parts of a business that require a body in a room are exactly the parts that hold. That's where we're headed next.

## How to identify your exposure

Forget "will AI replace my business" as a yes-or-no. Wrong question. Here's the owner's version of the task audit the researchers run on jobs.

Write down your money loop, every step from "opportunity exists" to "cash in my account." Not your org chart, not your service list. The actual sequence of things that happen, every time, to produce revenue. Then mark each step with one of two labels:

**AI-shaped:** repetitive, digital, rules you could write down, done at a screen, same steps every time.

**Human-shaped:** requires being physically somewhere, a relationship someone extends to you specifically, capital risk plus conviction, a license and the liability that comes with it, or judgment where the rules can't be written down.

Count the labels. If your loop is 80 percent AI-shaped, your business is a connector role and the clock is running. If the steps that actually produce your margin are human-shaped, you have a moat, and AI is leverage on top of it rather than a threat underneath it.

Run it on any business and the moats show themselves fast. A landscaping company's loop is overwhelmingly human-shaped: crews, equipment, dirt. Its AI-shaped steps, quoting, scheduling, follow-up, invoicing, are overhead to automate, not a business to lose. A restaurant, same story. A bookkeeping practice leans the other way: the categorizing and reconciling that fill the billable day are deeply AI-shaped, and the moat, to the extent there is one, is the advisory conversation and the trust, which is exactly the part most practices treat as an afterthought.

When I ran this on my industry yesterday, it produced the list that made everyone mad, and now you can see the logic behind it. Retail arbitrage holds because the edge is your body in a store finding what software can't see. Supplier and wholesale relationships hold because an account someone extends to you personally can't be scraped. Speculative buy-and-hold plays hold because the edge is conviction and capital risk, not information speed. The flip styles built purely on screen-speed and spreadsheet rules don't hold. Same audit, no sentiment, and it works identically on a brokerage, an agency, or a med spa.

None of the moats are permanently safe. They're defensible, which is the most anyone gets.

## How to fix it

If your audit came back mostly AI-shaped, you have two moves, and they're not mutually exclusive.

**Move one: become the machine operator.** The blocmates piece has a sharp line about employees at cost-cutting companies: in a company cutting payroll to fund compute, you want to be on the compute side of the trade. The owner's version: if agents are going to compress your niche, be the one running them. The tools are not enterprise-priced anymore. The same article's cost research found that $200-a-month subscriptions from the big AI labs currently deliver several times their price in equivalent usage, because the labs are subsidizing power users in a land grab. That subsidy is a window. Operators who build their systems during it get the compressed margins AND the volume. Everyone else just gets the compressed margins.

**Move two: migrate your weight up the loop.** Take your human-shaped steps and deliberately shift your time and capital toward them, while handing the AI-shaped steps to software before a competitor uses software to hand you a price you can't meet. For the bookkeeper, that's fewer hours categorizing and more hours in advisory conversations no tool can have. For the broker, it's the carrier relationships. For a flipper, it might be finally building the wholesale relationship you've been putting off. The research on employees found that entry-level roles exposed to AI now demand judgment and senior-level skills. The owner's version is the same sentence with the title changed: the surviving version of your business runs on judgment, relationships, and presence, with the repetitive layer delegated to software.

Here's your thirty minutes today. Write the money loop. Label every step. Count. If you don't like the count, pick ONE human-shaped step to invest in this month and ONE AI-shaped step to hand to a tool. That's it. You now know more about your actual exposure than most of the people yelling at each other about this online.

The people who got mad at me yesterday weren't mad because I was wrong. They were mad because the audit is uncomfortable when most of your loop comes back AI-shaped. I've run it on my own operations and had the same flinch. Run it anyway. The discomfort is information.

Hit reply and tell me what your loop looks like. I read every one.