1. The Essay and Its Central Argument
Weeks after Block announced one of the most significant workforce reductions in fintech history, co-founder and CEO Jack Dorsey published a philosophical essay alongside Sequoia Capital managing partner Roelof Botha — an investor in Block — titled "From Hierarchy to Intelligence." The piece is not a defence of the layoffs in the conventional sense of justifying a difficult business decision. It is a first-principles argument about why the corporate management layer exists at all, and why AI has now made it structurally redundant. The essay's central thesis is that hierarchy — the layers of managers, directors, and vice presidents that characterise large organisations — was never a natural or necessary feature of corporate life. It was a solution to a specific information-routing problem that arose when organisations became too large for any single person to maintain situational awareness. Managers aggregate context from below, transmit direction from above, and maintain alignment across teams. AI, Dorsey and Botha argue, now performs each of those functions continuously and at a scale that no human management layer can match.
2. The Information Problem That Hierarchy Was Built to Solve
The essay's most intellectually substantive contribution is its analysis of why hierarchy emerged in the first place. In any organisation larger than a handful of people, there is more information than any individual can process — information about what is being built, what customers are experiencing, what decisions are being made downstream, what problems are emerging in production, what competitors are doing. The human response to this information overload has been consistent throughout industrial history: create specialised roles to collect, aggregate, and transmit that information up and down the organisational ladder. Managers become information intermediaries — their core function is not to do technical work but to know what is happening across their area of responsibility and to ensure that knowledge reaches the people who need it. That function, Dorsey and Botha argue, is precisely what large language model systems, continuous data aggregation, and agentic AI workflows now do automatically, without the lag, distortion, and political friction that characterises human information chains.
3. Block's Two AI World Models
In place of the management layer, Dorsey and Botha describe two AI-driven "world models" that Block is building as the core of its restructured operating architecture. The first aggregates internal data from code commits, product decisions, workflow completions, and performance metrics to create a continuously updated picture of company operations — replacing the contextual awareness that managers traditionally maintained by attending meetings, reading status updates, and asking questions of their direct reports. The second maps customer and merchant behaviour using transaction data from Cash App and Square, creating a real-time model of how the company's products are being used and where demand is shifting. Together, these models feed what Block calls an "intelligence layer" that is designed to compose financial products dynamically in response to market conditions — translating the customer intelligence into product decisions without requiring human intermediation at each step of that translation.
4. The Scale and Timing of the Block Cuts
Block's workforce reduction, announced in late February 2026, cut approximately 4,000 positions from a total headcount of more than 10,000 — a reduction of approximately 40% that brought the company down to just under 6,000 employees. The cuts were concentrated in middle management, consistent with the essay's argument that the management layer is the specific function being automated. Dorsey told shareholders in a letter accompanying the announcement that the business remains strong — Block's gross profit was growing and the company was serving more customers than before — explicitly framing the cuts as being driven by capability change rather than financial necessity. Block's share price rose more than 24% on the announcement, with investors interpreting the reduction as an acceleration of efficiency gains rather than a signal of business difficulty.
5. The December AI Capability Shift That Triggered the Decision
Dorsey has been specific about what changed his assessment of AI's organisational implications. In an interview with Wired in early March, he identified December 2025 as the month in which he observed a qualitative shift in the capability of AI tools — specifically Anthropic's Claude Opus 4.6 and OpenAI's Codex 5.3, which he described as becoming capable of operating effectively in large, complex codebases for the first time. The ability to navigate and modify production-scale codebases — rather than generating isolated code snippets in a clean environment — represented the threshold at which AI moved from a developer productivity tool to an autonomous engineering participant. That shift, in Dorsey's framing, changed the fundamental assumption about how many engineers and engineering managers a company of Block's scale needs to maintain and grow its technical systems.
6. The Counter-Argument: Employees' Experience on the Ground
The essay's claims about AI's current operational capability have not gone unchallenged by people with direct experience of the tools within Block. Current and former employees who spoke with The Guardian described a more complex reality: approximately 95% of AI-generated code changes still require human modification before they can be merged, and AI tools cannot yet lead in regulated domains like banking and money transfers, where compliance requirements, error consequences, and regulatory examination create constraints that current AI systems cannot navigate autonomously. These accounts suggest a gap between the theoretical capability of the models Dorsey observed in December and the operational readiness of AI-driven workflows in production environments at the scale Block operates. Whether that gap closes over the next year — as Dorsey predicts — or proves more durable than the essay implies is the empirical question on which the thesis ultimately depends.
7. The "AI-Washing" Critique
The essay arrives in a context where scepticism about AI-driven layoff narratives has been building. When Block first announced the cuts, Bloomberg characterised the move as arousing "suspicions of AI-washing" — the practice of using AI as a legitimising framework for restructuring decisions that are primarily driven by conventional cost or overhiring correction considerations. Block's headcount had grown from 3,835 at the end of 2019 to more than 10,000 during the pandemic years — a more-than-2.5x increase that several commentators have attributed to the broad tech overhiring of that period rather than to genuine operational need. Dorsey acknowledged in a social media response that Block had over-hired during the pandemic, attributing some of the excess to structuring Cash App and Square as two separate companies rather than one, which was corrected in mid-2024. The essay's frame — that the cuts represent a forward-looking restructuring rather than a correction of past excess — exists in tension with that acknowledgement.
8. The Prediction: Most Companies Will Follow
One of the essay's more provocative claims is its forward-looking projection about how rapidly the restructuring Dorsey is undertaking will propagate across the broader corporate landscape. In his shareholder letter, Dorsey stated that within the next year, he believes the majority of companies will reach the same conclusion and make similar structural changes. That projection — which implies a wave of AI-driven management reduction across the technology and financial services sectors — has been both cited as evidence of genuine industrial transformation and dismissed as characteristic CEO hyperbole. What is not in dispute is that Block is not alone: Amazon, Meta, Salesforce, Pinterest, CrowdStrike, and Chegg have all made workforce reductions that they have attributed at least partially to AI-driven efficiency improvements, with each company's framing varying between genuine capability change and pandemic-era overhiring correction.
9. The Implications for Cash App, Square, and Crypto
Block's restructuring has specific implications for its product portfolio, which includes Cash App — the peer-to-peer payment network with significant crypto functionality including bitcoin buying, selling, and the Lightning Network — and Square, the merchant payment platform. Cash App's bitcoin feature set has been one of the most significant vectors for retail BTC adoption, providing a regulated, accessible on-ramp for millions of users. If the management reduction and AI-driven operating model successfully improves product velocity and reduces the time from insight to product change, crypto-adjacent features within Cash App could benefit from faster iteration. If the operational disruption of a 40% headcount reduction creates instability in the core payment infrastructure, the near-term effect could be the opposite. Block's stated goal of achieving $2 million in gross profit per employee — 4x the pre-pandemic efficiency ratio — implies significant operational output from a much smaller team.
10. The Deeper Question: Is This the Beginning or an Anomaly?
The Dorsey-Botha essay represents one of the most fully articulated executive-level arguments for AI's replacement of management function rather than its supplementation of management capability. The distinction the essay draws — between companies giving everyone a copilot that makes existing structure work slightly better, and companies replacing what hierarchy does — is a genuinely meaningful conceptual division. Whether that distinction proves durable in practice depends on questions the essay does not fully answer: Can AI reliably make production-quality decisions in regulated financial services without human oversight? Can an intelligence layer replicate the contextual judgment that experienced managers accumulate through years of specific organisational experience? And when AI systems fail — and they will fail — who bears the accountability that management hierarchy currently provides? Block's restructuring is a high-stakes test of these propositions in a real company with real customers, real regulators, and real consequences for both groups if the answer proves to be no.