# From Hierarchy to Intelligence *~3,500-word essay by [[Jack Dorsey]], posted to X on April 8, 2026, with companion page at [block.xyz](https://block.xyz/inside/from-hierarchy-to-intelligence). Endorsed by [[Sequoia Capital]] in the opening paragraph.* A manifesto arguing that AI makes corporate hierarchy obsolete — and that Block is building the replacement. Published weeks after Dorsey laid off 40% of Block's workforce (~4,000 people), the piece provides the intellectual framework justifying those cuts: middle management existed to route information through organizations, and AI can now do that job. Sequoia Capital's endorsement in the opening paragraph elevates this from a CEO's self-justification to an industry-defining organizational thesis backed by Silicon Valley's most prestigious VC firm. ## The Argument ### Historical Arc: Why Hierarchy Exists The post traces corporate hierarchy through five stages, arguing that every form of organizational structure since antiquity has been an information-routing protocol constrained by human cognitive limits: 1. **Roman Army** (~200 BC): The contubernium (8 soldiers → decanus) through to legion (5,000 → legate). The structure 8 → 80 → 480 → 5,000 was an information routing protocol built around the human limitation that a leader can effectively manage 3–8 people. This is "span of control." 2. **Prussian General Staff** (post-1806): After Napoleon destroyed the Prussian army at Jena, reformers created a dedicated officer class whose job was not to fight but to plan, process information, and coordinate. Dorsey calls this "middle management before the term existed" — professionals whose purpose was to "support incompetent Generals, providing the talents that might otherwise be wanting among leaders and commanders." 3. **American Railroads** (1840s–1850s): West Point-trained engineers brought military organizational thinking to private companies. Daniel McCallum of the New York and Erie Railroad created the world's first org chart (mid-1850s) to manage a 500-mile system. Train collisions were killing people. The chart formalized the same hierarchical logic the Romans had used. 4. **Scientific Management → Matrix Organizations** (1900s–1970s): [[Frederick Winslow Taylor]] optimized tasks within hierarchy. The Manhattan Project tested cross-functional coordination under Oppenheimer. McKinsey's matrix organization (1959) combined functional specialties with divisional units. The McKinsey 7-S framework (late 1970s) acknowledged that structure alone was insufficient. 5. **Tech Company Experiments** (2000s–2020s): Spotify squads, Zappos Holacracy, Valve's flat structure. All failed at scale. Organizations revert to hierarchy because "no alternative information routing mechanism has been powerful enough to replace it." ### The Claim: AI Changes the Constraint The core argument: for 2,000 years, organizations needed human middle managers to route information. AI can now maintain a continuously updated model of an entire business and coordinate work that previously required humans relaying information through layers. This is the first time a technology has been capable of actually performing the coordination functions that hierarchy exists to provide. ### Block's Proposed Architecture Dorsey proposes replacing the traditional corporate structure with four components: 1. **Capabilities** — Atomic financial primitives (payments, lending, card issuance, banking, BNPL, payroll). Not products. Building blocks with no UIs of their own. 2. **A [[World Model]]** — Two sides: - *Company world model*: How the company understands its own operations, replacing information that flowed through management layers. Dorsey notes that Block is remote-first and "everything we do creates artifacts" — decisions, discussions, code, designs, plans, problems, and progress all exist as recorded actions. This becomes the raw material for the model. - *Customer world model*: Per-customer, per-merchant representation built from proprietary transaction data. Dorsey argues "money is the most honest signal in the world" — when people spend, save, send, borrow, or repay, that is the truth. Block sees both sides of millions of transactions daily (buyer through Cash App, seller through Square). 3. **An Intelligence Layer** — Composes capabilities into solutions for specific customers at specific moments, proactively. Dorsey's example: a restaurant's cash flow tightens ahead of a seasonal dip the model has seen before; the intelligence layer composes a short-term loan, adjusts repayment, and surfaces it before the merchant even looks for financing. No product manager decided to build that solution. The traditional roadmap — where product managers hypothesize about what to build next — is described as "any company's ultimate limiting factor." 4. **Interfaces** — Square, Cash App, Afterpay, TIDAL, Bitkey, Proto. Delivery surfaces. "They are important, but they are not where the value is created. The value is in the model and the intelligence." ### The New Org Structure: Three Roles Dorsey proposes normalizing the entire company to three roles, eliminating permanent middle management: - **Individual Contributors (ICs)** — Deep specialists who build and operate capabilities, the model, the intelligence layer, and interfaces. The world model provides the context a manager used to provide, so ICs make decisions about their layer without waiting to be told. - **Directly Responsible Individuals (DRIs)** — Own specific cross-cutting problems or customer outcomes for defined periods (e.g., 90 days). Full authority to pull resources across teams. - **Player-coaches** — Combine building with developing people. Replace managers whose primary job was information routing. Still write code or build models. Handle craft and people development. "They don't spend their days in status meetings, alignment sessions, and priority negotiations. The world model handles alignment." "There is no need for a permanent middle management layer." ### The Closing Argument The post ends with a competitive framing and a universalist claim: "If the answer is nothing, AI is just a cost optimization story. You cut headcount, improve margins for a few quarters, and eventually get absorbed by something smarter. If the answer is deep, AI doesn't augment your company. It reveals what your company actually is." And: "Block is far enough along to show the idea is more than theory." And: "Companies move fast or slow based on information flow. Hierarchy and middle management impede information flow. For two thousand years... we have had no real alternative... They aren't anymore. Block is building what comes next." ## Critical Analysis ### What It Actually Justifies This manifesto was published weeks after Block cut 4,000 jobs. It provides an intellectual architecture for that decision and for future cuts. The argument: middle management existed to route information → AI routes information better → middle management is obsolete → we eliminated it → this is progress, not downsizing. Former Block executives have disputed this narrative, arguing the layoffs were primarily about correcting bloated post-COVID hiring and failed expansions into lending, banking, and BNPL — not a visionary AI-driven reorganization. ### The Sequoia Endorsement The post opens with Sequoia Capital's voice: "At Sequoia, we see that speed is the best predictor of start-up success." This is significant. Sequoia is the most prestigious VC firm in Silicon Valley. Their endorsement transforms what could be read as a CEO justifying mass layoffs into an industry-defining organizational theory backed by institutional authority. It signals to other companies: this is the direction. It also positions Block favorably for investor confidence at a moment when the company's stock could be vulnerable after a 40% workforce reduction. ### Historical Cherry-Picking The historical narrative is selective. The Roman army's hierarchy worked because it was a military organization with clear chains of command, not a creative enterprise requiring judgment and innovation. The Prussian General Staff's middle management layer was specifically designed to compensate for mediocre leaders — which could be read as an argument *for* middle management, not against it. The Manhattan Project succeeded under Oppenheimer's singular genius, which Dorsey acknowledges was "a wartime exception" — yet he implies Block can routinize that exception through AI. Every historical example is a military organization or a response to military problems; the essay does not grapple with what organizational research says about knowledge work, creativity, and employee wellbeing. ### The "World Model" as Corporate Panopticon The "company world model" — an AI that maintains a continuously updated picture of what every employee is doing, what's blocked, where resources are allocated — is, from a labor perspective, a total surveillance system. When Dorsey says "everything we do creates artifacts" and "decisions, discussions, code, designs, plans, problems, and progress all exist as recorded actions," this is describing a remote-work panopticon where every action is legible to the system and presumably to senior leadership. The essay frames this as empowering ("ICs can make decisions about their layer without waiting to be told"), but the power to observe everything flows upward. ### "People on the Edge" The phrase "the intelligence lives in the system, the people are on the edge" inverts the traditional understanding of human agency in organizations. In this model, the AI system holds organizational intelligence, and humans perform the tasks the AI cannot yet perform. This is a transitional framing — it implies that as AI improves, the "edge" tasks humans perform will shrink. Dorsey acknowledges humans are needed for "intuition, opinionated direction, cultural context, trust dynamics, the feeling in a room" and "ethical decisions, novel situations, and high-stakes moments" — but the direction of travel is clear: the system expands, the edge contracts. ### The Scientific Management Parallel The essay invokes Taylor but doesn't reckon with the critique of [[Scientific Management]]: that breaking work into machine-legible components, optimizing for efficiency, and centralizing intelligence in "the system" (Taylor's engineers, Dorsey's AI) produces alienated workers, fragile organizations, and systems that optimize for measurability over meaning. Dorsey's "capabilities as atomic primitives" and "intelligence layer that composes solutions" is Scientific Management updated for the AI era — with the AI playing the role Taylor assigned to the engineer-manager. ### Connection to Broader AI-Labor Narrative Dorsey's explicit claim — "we believe the pattern behind this... will reshape how companies of all kinds operate over the coming years" — positions Block's restructuring not as a one-company decision but as an inevitability for all large organizations. This is the same rhetorical move documented across tech industry AI narratives: framing labor elimination as technological destiny rather than a corporate choice. ## Relationship to KB Themes ### [[The Network State]] — Parallel Architecture Where [[Balaji Srinivasan]]'s *The Network State* proposes replacing the nation-state with crypto-enabled digital communities, Dorsey's "From Hierarchy to Intelligence" proposes replacing the corporate hierarchy with AI-enabled coordination. Both share the same core move: traditional institutions (government/corporation) are information-routing systems constrained by human limits → technology removes those limits → the institution can be fundamentally restructured → the people who build the technology should govern the restructuring. Both are products of the same Silicon Valley ideology of institutional exit and technocratic replacement. ### [[DOGE Ecosystem]] — Government Parallel Dorsey's argument about corporate middle management mirrors [[Elon Musk]]'s argument about government bureaucracy through [[DOGE Ecosystem]]: both frame human intermediaries as information bottlenecks that technology can replace, justifying mass workforce elimination. The difference is scale — Dorsey is restructuring a 10,000-person company, Musk targeted the entire federal government. Both use the same rhetorical structure: invoke historical inefficiency → claim technology now solves it → fire people → declare progress. ### [[Crypto-Political Machine]] — Ideological Alignment Dorsey's vision of Block as "an intelligence rather than a hierarchy" fits within the broader Silicon Valley ideology that human institutions are inefficient systems waiting to be replaced by technology — the same ideology that drives crypto's replacement of banks, defense tech's replacement of legacy contractors, and DOGE's replacement of civil servants. ### AI-as-Justification Pattern This manifesto is the most intellectually developed version of a pattern seen across the KB: tech executives using AI as the stated justification for corporate restructuring that serves shareholder interests. The 3,500-word historical arc from Roman legions to Block's "intelligence layer" provides scholarly cover for what critics describe as a straightforward cost-cutting exercise wrapped in organizational theory. ### [[Effective Accelerationism]] Resonance The essay's closing argument — that companies either build deep AI understanding or "get absorbed by something smarter" — echoes the accelerationist framing that resistance to AI transformation is futile and adaptation is survival. There is no option presented for "keep humans in management roles because human judgment matters." The only options are: transform or die. ## See Also - [[Jack Dorsey]] - [[The Network State]] - [[DOGE Ecosystem]] - [[Crypto-Political Machine]] - [[Elon Musk]] - [[Dark Enlightenment]] - [[Effective Accelerationism]] - [[Scientific Management]] — the 1911 doctrine the essay updates; see also [[Technocracy Incorporated]] for where the engineer's claim to rule went next - [[PayPal Mafia to Pentagon Pipeline]]