AI Didn’t Kill Work. It Exposed the Lie.

There’s a tax inside most companies that we are only now starting to quantify. It’s the cost of all the people, meetings, software, approvals, and reporting layers that exist solely to move information from one side of the room to the other. We used to call this “collaboration” but lot of it was just motion. Status meetings, escalation chains, follow-up loops, internal handoffs.

We built entire organizations around the transportation of work instead of the execution of it and now it feels like AI is forcing a level of honesty most companies have avoided for decades. It's like the mask is coming off and what’s underneath is a little bit uncomfortable. A surprising amount of what we called leadership, process, and productivity was just expensive routing.

This week, that reality kept showing up everywhere I looked.

Let’s get into it.

RIP middle management.

The dirtiest secret in corporate America? A shocking amount of middle management was never management. It was just workflow routing with a salary.

Was this ever really management? Or did we convince ourselves it was leadership because we needed to justify it?

Either way, AI has forced the mask off. Meta and Block, Inc. are done pretending. Meta is replacing middle managers with “org leads,” and Block is turning them into “player-coaches.”

Don’t get distracted by the rebrand, this feel more like an extinction event. The companies moving fastest are realizing you do not need five layers of people translating information from one team to another.

You need:

One great operator with AI can now do what used to require:

It feels like a lot of people aren’t being replaced by AI. They’re being replaced by the realization that their job never should have existed in the first place.

Is the future org chart flatter, meaner, and brutally accountable? It sounds like that’s already happening.

Getting rid of layers is the easy part. The hard part is convincing the remaining team to help you burn the rest of the busywork to the ground. You have to change the rules of the game.

Want AI adoption? Don’t ask your employees to "experiment" and start paying them to build AI Agents that do the work.

Leadership says “use AI.” Employees hear: “build the thing that replaces you.” Then everyone acts surprised when adoption stalls.

Of course it stalls. You’re asking people to automate work while their compensation, status, and job security are still tied to preserving that work.

The incentives are broken. If you actually want adoption, reward the behavior.

That’s exactly what we’re doing. This month, we’re paying the team to build autonomous AI agents that create real business value. We’re calling it April Madness :) Big cash prize to the person who builds the most creative, effective autonomous agent, one that wakes up and does real work.

If you want real AI adoption, stop asking employees to “experiment” and start paying them to delete work.

When the incentives are aligned, the focus shifts from "how do we look busy" to "what tools actually produce the result." Here is how we are connecting the pieces to bridge the gap between research and revenue.

I keep hearing “Will Claude Code replace Clay?” I don't think so.

They solve different parts of the same problem, and most teams are still struggling to connect those parts in a way that actually produces pipeline.

Here’s how we’ve started to think about it:

Claude for depth

Clay for structure

Atonom for execution

What’s been interesting is realizing that most GTM stacks today are very good at generating insight and organizing it. They’re much less consistent at actually carrying that work through to an outcome.

And that gap, more than any individual tool, is what seems to determine whether these systems create real pipeline or just better inputs. I don’t think this is really a Claude vs Clay conversation.

It’s about whether your system is designed all the way through to execution, or whether it still depends on a human to close the loop every time.

Building these execution loops yourself leads to a radical question: if we can build everything we need, why are we still paying monthly for software?

SaaS is dead. Why pay for software when Claude Code can build anything? But here’s the catch…

Most people do not want to spend their weekend becoming the support team for the software they were supposed to buy.

Look, the truth is convenience has always been the product. The software was just the wrapper. Not because they can’t do it themselves. Because they don’t want to. They want the damn thing to work. Open the app, get the outcome, move on.

That’s what great software has always sold. It was never code, it was certainty.

So yeah, SaaS is dead on LinkedIn but in the real world, people will keep paying for whatever removes work from their lives.

Which brings me to something that’s been bothering me. The people supposedly most excited about AI are starting to resent it.

Nearly 1 in 3 Gen Z respondents say AI makes them feel angry???

Weird, right? According to a new Gallup study, anger toward AI among Gen Z made a huge jump from last year. The generation everyone thought would love AI is getting sick of it. Why?

They’re being forced to use it

It's funny, it's like it's not optional curiosity anymore.

So the emotion isn’t “wow, this is exciting.” It’s, "I have to use this to keep up." That’s how people start to hate a tool fast.

They can feel themselves getting stupider

A lot of younger workers understand that AI can make them faster right now while making them dumber over time.

Cool. But if the machine is doing all the reps, when do I actually get better? I think people can totally feel that.

The hype doesn’t match reality

Every week there’s another post saying AI will transform everything. Then people go use it and hit:

I think a lot of frustration starts there. People hate being sold a revolution and handed a mediocre copilot.

You know, maybe the anger isn’t about AI at all. Maybe it’s the realization that we’re trading capability for convenience. Faster work today but weaker judgment tomorrow. And if the next generation already feels that tension, what happens when we raise an entire workforce that no longer knows how to think without it...

Whether it is the middle manager, the routing software, or the agency that doesn't actually produce a result, the buffer zones are being deleted. We are being forced into a new level of honesty about what creates value and what is just friction with a title.

We have spent years normalizing the act of looking busy. We mistook full calendars for impact and long email threads for collaboration. But the tools we have now are too efficient to allow that charade to continue. The goal was never to work harder. The goal was to remove the layers until all that was left was the outcome.

Own the execution, or get lost in the motion.

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.