Key insights:
Jack Dorsey just made one of the boldest calls in modern business history. He cut 40% of Block's workforce and rebuilt the company around an AI intelligence layer.
His thesis is simple. The pyramid org chart we have used for 2,000 years is no longer the most efficient way to move information through a company.
In a recent Sequoia Capital conversation with Roelof Botha and Brian Halligan, Jack laid out a vision where AI sits at the center of the organisation and humans operate at the edges. This blog breaks down what that means for you, your company, and your career.
The hierarchy exists for one reason. Information flow at human scale.
Managers relay information up and down a chain. Decisions get filtered. Context gets lost. Politics creep in. By the time a customer signal reaches the CEO, it has been distorted by five layers of interpretation.
Jack pointed out that every action inside a modern company creates an artifact. A Slack message. An email. A pull request. A Google Doc. A recorded meeting. These are all signals.
The hierarchy was built to manage these signals through humans. But you no longer need humans to do that job.
Put an intelligence layer on top of all those artifacts. Now anyone in the company can query the company itself.
You can ask the company how it is doing. You can ask what is broken. You can ask what customers want next. The CEO does not have a monopoly on context anymore.
This is what Jack means when he calls the company a mini-AGI. Every business is already a form of intelligence. It just has not been structured to be queried like one.
The technology is finally good enough to model the whole company. That changes everything about how decisions get made.
If you are sitting in a middle management role whose primary job is to relay information, you should be paying attention.
The jobs that survive this shift are the ones that involve building, owning outcomes, and coaching humans. If you cannot do at least one of those things at a high level, you are exposed.
The good news is you can position yourself on the right side of this shift. By learning to build with AI, automation, and code, you become the person companies need more of, not less.
This is exactly why the Complete RPA Bootcamp exists. You go from beginner to building real automations, agents, and computer-use systems. You become the builder, not the person being replaced.
Block is collapsing every job title into just three roles. This is the most concrete part of Jack's manifesto, and it is worth studying carefully if you lead a team.
The first role is the IC. The builder or operator.
This is the engineer, the designer, the salesperson, the product person. The person actually working with the tools to build the product or operate the company.
ICs are heavily augmented. They have access to agents. One person can now do the work that used to take a team of ten.
The durable human skill here is judgment, taste, and creativity. The AI can produce 80% of the output. The last 20% is taste. That is where your value lives.
The second role is the DRI. The directly responsible individual.
This person owns customer outcomes. They put strategy together. They figure out what road map will actually solve customer problems. They assemble teams of ICs to get it done.
The durable human skill here is ownership and accountability. The DRI does not hide behind committees or process. They own the result.
This is one role per outcome. No shared ownership. No diffusion of responsibility. If it fails, you know exactly who is accountable.
The third role is the player coach.
This is what we used to call a manager. But there is a critical difference. The player coach builds capability in others by doing the work, not by telling people how to do it.
They might also be an IC or a DRI. Coaching is an assignment, not a reporting line. You get assigned to help certain ICs master their craft.
The durable human skills here are empathy, coaching ability, and building human capacity. The soft skills great managers have always been known for.
What is removed is the requirement that a manager also be strategic or hold political power over their reports.
The behind-the-scenes story of how Jack made the call to cut 40% is one of the most useful parts of the conversation. It is a masterclass in decisive leadership.
December 2024. The coding models took a noticeable jump. They went from being good at greenfield prototypes to actually understanding large legacy codebases.
The hallucination problem was largely solved for code. Tool harnesses like Claude Code and Block's internal Goose matured rapidly.
Everyone went home for the holidays. Everyone played with the tools. Everyone came back stunned.
Jack sat down with his leadership team and asked one question. If you were building this company today with these tools, would it look anything like this?
The answer was a unanimous no.
They worked through a structured calculation:
They added a buffer for mistakes. That gave them the number.
The exploration to execution took under three weeks. The board, led in part by Roelof Botha, was aligned quickly because Jack had written a detailed, principled memo laying out the logic.
This is the part most CEOs get wrong.
Jack did not want to react. He did not want to wait until the company had its back against the wall. Public companies that wait end up doing annual 10% layoffs that demoralise everyone.
Acting early lets you be generous with the people leaving. It lets you set a tone of excellence for the people staying. You are building toward something, not retreating from something.
If you are leading a company right now, this is the question to sit with. Are you reacting or are you acting?
The shift Jack is describing is not just a story about Block. It is a preview of what every knowledge-work company will look like within a few years.
Most companies are treating AI as a co-pilot. Something you bolt onto existing workflows to make people 10x more productive.
Jack thinks that is the wrong frame entirely. He believes this is structural. You do not add AI to your org chart. You rebuild the org chart around AI.
If you only use AI as a productivity tool, you end up looking like everyone else. You end up rhyming too closely with the frontier labs. And then you cannot differentiate.
The companies that win will be the ones that rebuild from first principles with AI at the core.
If the three durable roles are building, owning outcomes, and coaching, then the skills you need are clear:
The middle layer is going away. You either build, own, or coach. Pick one and get exceptional at it.
The fastest path is to learn the tools that companies like Block are actually using to rebuild themselves.
That means understanding Robotic Process Automation, agentic automation, coded automation, and computer-use agents. These are the building blocks of the intelligence layer Jack is describing.
The Complete RPA Bootcamp takes you from beginner to pro across all of these areas. You finish the program ready to step into a role as an automation developer, the kind of IC Jack describes as the largest part of the workforce going forward.
Instead of letting AI and automation replace you, you become the person who builds the AI and automation. That is the switch that future-proofs your career.
The conversation also went deep on CEO craft. Boards, second acts, conviction, and learning. There is gold here for anyone leading a company.
Jack and Roelof both emphasised treating your first investor as a recruiting decision, not a financing decision.
You can never fire them. They can fire you. So optimise for the person, not the brand name on the fund.
Add an independent board member within a year or two, ideally once you hit product market fit. That person can be a different kind of mental sparring partner for the founder than an investor board member can.
And do not wait until the run-up to an IPO to build out the board. Boards built in a rush have no chemistry, no history, and no shared values. They fall apart under pressure.
Jack is the only person to have been simultaneously CEO of two public companies. His advice? Do not do it.
He called it an anti-pattern for public companies. Private companies, maybe. But running two public companies pulls focus and creates structural problems you cannot easily fix.
His biggest regret at Block was delegating too much in the early days. He let Square and Cash App run as separate fiefdoms with separate CEOs. The result was differing cultures, values, and execution levels. A holding company instead of an integrated financial network.
Cash App is the case study. The team of eight people building it was hated internally for two to three years. The board did not understand it. Investors did not understand it. Even Keith Rabois, a PayPal founding team member on Block's leadership, said no.
Jack defended it anyway. He knew he was losing credibility every day. He also knew he could earn it back if it worked.
Cash App is now over half of Block's business.
The lesson is simple. Every leader has to be comfortable losing credibility with stakeholders. If you have a clear principle for why something needs to exist, you can stake part of your reputation on it. If it works, you earn the credibility back with interest.
Roelof and Jack closed with their views on what makes a great CEO. Some things change. Some things never do.
Roelof's framework is ALE. Authenticity, logic, and empathy.
These have not changed. They will not change. The fundamentals of leading humans stay the same even as the tools change.
Speed. The pace of change is so fast that the cost of dithering has gone up dramatically.
Jack's decision to restructure Block took three weeks from exploration to execution. A year ago that would have been a six month process. In two years it might need to be a one week process.
You have to be able to reprogram your own mind and assumptions constantly. Question your past decisions. Question your sacred cows. Question what the competition is doing.
When the tools can produce 80% of any output, what separates winners from losers is the last 20%.
That last 20% is taste. It is having a point of view. An opinionated drive to push something into the world that did not exist before.
Right now there are a lot of companies that are just copies of copies of copies. They are easy to make because the tools are easy to use. What is missing is a perspective.
Your perspective. Your judgment. Your willingness to build something uncomfortable because you believe in it.
If you want to see Jack walk through this thinking in his own words, including the part where he describes meditation as a physical practice for sharpening your mind, watch the full conversation embedded below from the Sequoia Capital YouTube channel. It is one of the most thought-provoking discussions of the year on where companies are heading and what kind of builder you need to become to thrive in that future.