Work Is Bypassing Systems

Work is changing faster than most people want to admit. Not in theory. Not in five years. Right now, in small, almost boring ways that are easy to ignore until they stack up.

The line between humans and software is getting blurry. Tools are starting to make decisions. Software is starting to coordinate people. The systems we built for a different era are being stretched past their limits.

A lot of these moments feel small or even funny at first. But taken together, they point to something bigger. The operating model is changing.

Let’s get into it.

Before we talk about AI and the future of work, a quick story about a very small crime.

I need to confess something before this turns into a federal case. Last night I’m at a restaurant with a friend. He orders water. Normal, law-abiding, and mostly innocent. Then I watch him walk over to the drink station…and give the lemonade lever a tiny squeeze.

Not a cup of lemonade or a whole a refill. Let's call it a spritz for arguments sake. We freeze. Eye contact is made. I ask him, “Is this a joke… or do I need to turn you in?”

We debated for a solid two minutes. Is this a crime or a is it a loophole?

Here’s the thing. He only drinks water and black coffee. He’s boring. He’s accepted that. But he’s also old school. Water deserves a little lemon, he says. That used to be normal. There were wedges or bowls of citrus. Now? Nothing. So he adapted.

He’s hit the lemonade lever before and he'll hit it again. I can’t tell if I should have turned him in or thanked him for adapting to a world that quietly gave up on citrus.

This seems trivial. It isn’t. It’s about how people adapt when systems quietly stop serving them. Including at work.

The adaptation with AI is getting weird. We went from “AI will take your job” to “AI please rent a human” in about 18 months.

That was kinda fast.

There’s a site called Rentahuman.ai. Which is exactly what it sounds like and somehow still worse.

AI agents can post bounties. Humans line up to do them. Im not kidding...

Real examples:

Congrats. You’re no longer an employee. You’re… the hands. Hacker News is losing its mind over this, which is my favorite part.

People are like:

“What if an AI orchestrates a crime by hiring 3 gig workers who don’t know the full plan??”

Relax.

Half of you wouldn’t even stand up. You’d click “Done,” upload a blurry photo, and collect the $10. One comment nailed it: “First I built software to do my bidding… Now the software is using my hands.”

Also yes, humans are now officially a feature :)

Anyway. Update your LinkedIn headline. Open to work. For AI.

This is usually where people jump to panic and call it dystopian. That reaction misses the real change. Who is actually doing the coordinating?

Lately it’s becoming harder to accept the way tools are supposed to work.

We ripped out a $30k Salesforce contract, looked at the “alternatives,” laughed, and built our own CRM for $1K instead.

$30k annually so my team could move columns around and lie to themselves with nicer dashboards. Nope, not anymore.

I checked the options: HubSpot. Attio. Pipedrive. Copper. Close. Different paint. Same house.

So we stopped shopping and we built our own CRM. First version took 3 hours. Built with Lovable and it costs $100 a month.

Works exactly how we work. And before the CRM mafia starts foaming at the mouth, yes, it does the basics:

All there. Because newsflash… a CRM is just data models and workflows.

And no we didn't have a:

One operator. One weekend. Done. If you’re a startup paying five or six figures a year for a CRM, be honest about what you’re buying:

My advice, having gone through this the hard way:

1) Stop worshipping tools

2) Stop confusing spend with sophistication

3) Stop letting vendors tell you complexity equals maturity

Build a CRM for what you need. This was never really about a CRM. It was about how much complexity we tolerate to avoid rethinking execution.

People keep asking why we named the company Atonom (ah-toh-nom). It wasn’t a rebrand. Something broke.

For years, B2B software followed the same playbook. Tools on top of tools. Licenses. Seats. Dashboards. Then armies of people to make it all work.

What was sold as “business in a box” turned into blank systems, endless configuration, and a growing tax on every company. Stack 20 or 30 SaaS tools together, hire layers of teams to manage them, and call it scale.

That worked. Until it didn’t. Investors got tired, companies got tired, and honestly operators got tired.

Growth at any cost stopped working, but brute force was all anyone knew. Then AI crossed a line. It stopped helping and started doing.

That’s when it clicked. We weren’t building software anymore. The product stopped being something you use and became something that works.

Cloud Employees are autonomous units of work that live inside your systems and own jobs end to end.

They work alongside humans, learn over time, and execute without constant supervision.

That’s what Atonom means:

Atonom is about building the autonomous organization. One where humans focus on judgment, creativity, and direction, and Cloud Employees handle the repetitive, always-on execution. That’s why we chose the name.

Names matter when something new is forming. This is not automation. It is autonomy.

The point is not the name. The point is the operating model. Humans focus on judgment, direction, and relationships. Software takes responsibility for execution.

Once you see that shift, it becomes hard to unsee it.

Which brings me to what I am actually building next.

I never actually announced what I was doing next. Now it has a name: 𝗔𝘁𝗼𝗻𝗼𝗺

AI was the obvious next move. What wasn’t obvious was how badly most companies screwed it up. They treated AI like software. More tools. More features. Same org chart. Same excuses.

Everything snapped into focus when I teamed up with David Elkington.

SaaS isn’t dying because it failed. It’s dying because it never actually did the work. For years, SaaS sold software and called it “service.” Meanwhile, humans did all the execution. That model is over.

AI isn’t here to “help” with work. It’s here to do it.

That's what we're building at Atonom. You don’t rent tools. You hire AI teammates that own outcomes.

We call it the Autonomous Organization. It runs on Cloud Employees. Advanced AI agents that live inside your systems, learn how your business actually works, and execute without needing to be managed like a child.

This is a new operating model for growth. Execution no longer scales with headcount. Output does. That changes what’s possible. And yeah, this is going to be fun.

See what we’re building here.

What is happening right now is not an AI trend or a tooling cycle. It is a shift in how work gets done and who or what is responsible for outcomes.

Most companies will try to layer this on top of existing models until it breaks. The ones that move early will not look smarter at first. They will just look different.

That’s it for today. Connect with me on Linkedin if you actually want to understand what an Autonomous Organization looks like in the real world.