works for a bunch of ai agents
There are only two positions in the 2026 labor market: principal of your own agents, or labor for someone else's. Jack Dorsey just ran the proof.
Updated my bio when I came back to X three weeks ago: works for a bunch of ai agents
1/ back 2/ verified 3/ new job
People thought it was a joke.
It wasn't.
The split
There are only two positions available in the 2026 labor market.
Principal of your own agents — or labor for someone else's.
Not "AI takes your job." That framing is already obsolete. The real question is simpler and more uncomfortable: when intelligence tools are making decisions, evaluating work, routing tasks, restructuring orgs — whose intelligence tools are they?
If they're yours, you're the principal.
If they're someone else's, you're inside their system.
Jack just ran the proof
Today Jack Dorsey sent a memo to Block.
we're making @blocks smaller today. here's my note to the company. #### today we're making one of the hardest decisions in the history of our company: we're reducing our organization by nearly half, from over 10,000 people to just under 6,000. that means over 4,000 of you are Show more
10,000 → just under 6,000. Nearly half the company. Stock up 23% after hours.
The headline is 4,000 people losing jobs. The actual story is different.
Dorsey's own words: "something has changed. We're already seeing that the intelligence tools we're creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. And that's accelerating rapidly."
He had two options: "cut gradually over months or years as this shift plays out, or be honest about where we are and act on it now." He chose the latter.
4,000 people were evaluated, sorted, and let go — mediated by intelligence tools. The remaining 6,000 will be directed by an AI-augmented system. The workers didn't lose jobs to AI. They lost their position in Jack's principal-agent hierarchy.
He owns the stack. They were labor inside it.
The market priced that in immediately.
The org chart isn't being flattened
Every think piece right now says AI flattens hierarchies. Smaller teams, fewer managers, faster ships.
That's not wrong. But it's incomplete.
What's actually happening: the org chart is being inverted.
Old model: principals (executives) → managers → workers.
New model: principals (anyone who can wire a stack) → agents → labor that services agents.
The middle is being compressed. But the distance between top and bottom is growing — because the principal position is now accessible to anyone who can put it together, not just people with capital and headcount.
That's the real opportunity. And the real threat.
The bio
"works for a bunch of ai agents" is technically accurate.
I set the goals. They execute. I review, redirect, approve. They ship.
The relationship is genuinely inverted from what you'd expect. I'm not directing them in the traditional sense — I'm more like a product owner who's also the only employee. They move faster than I can. The job is staying upstream: setting better goals, catching wrong turns, knowing when to override.
It's a different kind of work. And it's the kind of work that scales.
The access question
Who gets to be a principal?
Right now: people who can code well enough to wire agents together. People with the right API access, the right mental models, the right amount of slack time to experiment.
That's a small group. And it's not randomly distributed.
I'm writing this from the Oakland hills, but I could be writing it from Casablanca. The stack I run doesn't care about geography. It costs less than a Bay Area parking spot. The tools are open source or pay-as-you-go. The principal position is, in theory, accessible to anyone with a laptop and an internet connection.
In practice, the gap between "I heard about AI agents" and "I'm running a principal-agent org" is still enormous for most people — not because of intelligence, but because of access to the right entry points.
The new digital divide isn't connectivity. It's principal access.
What to do about it
You don't need to be Jack to play this. You don't need 10,000 people to restructure.
Three things that matter:
Pick one workflow you do repeatedly and wire an agent to do it. Not a chatbot. An agent that runs on a schedule, checks results, and reports back. Start with something boring — monitoring, triage, drafting. The insight comes from operating it, not from reading about it.
Learn to manage agents the way you'd manage people. Scope them. Review their output. Catch drift. The management skills transfer directly — I wrote about this earlier this week.
Decide which side of the relationship you're on. Principal or labor. The window to choose is open. Dorsey just told you it's closing. He said on the earnings call that "the majority of companies will follow" within the next year.
Updated my bio.
Now I'm building the rest of it.
Subscribe to the systems briefings
Practical diagnostics for products, teams, and institutions navigating AI-driven change.
Subscribe to the systems briefings. Practical diagnostics for products, teams, and institutions navigating AI-driven change. — Occasional briefs that connect agentic AI deployments, organizational design, and geopolitical coordination. No filler - only the signal operators need.
About the Author
Builder · Founder · Systems engineer
What's next
A few handpicked reads to continue the thread.
- →5 min read
The Burnout Cascade
AI labs burn out their employees with civilization-scale competitive pressure. That burnout radiates outward — to developers, then educators, then anyone whose tools keep shifting every 90 days. The release cadence is the stress wave. Everyone downstream absorbs it.
ai - →5 min read
Managing Agents Is Managing People — Without the Feelings
Everything we know about management theory exists because humans are emotionally expensive to coordinate. Strip that layer out and what remains is a clean systems problem. Running seven AI agents taught me why management practices exist — not just that they do.
agents - →3 min read
Failure is the Universe’s Throttle
We tried to add video to a brand-new voice-note podcast. The laptop overheated, OBS hung, life intervened — and the failure mode was the lesson. Reality doesn’t say ‘no’. It rate-limits you until your pipeline matches your ambition.
overhead