Jack Dorsey is Block’s cofounder, Block Head and chairman, and previously co-founded and led Twitter. His current public writing examines how AI could coordinate work inside companies, alongside the agent software developed by Block’s engineering teams.[1][2]
Start with From Hierarchy to Intelligence →
X Career
Dorsey co-founded Twitter, the public short-message service, and served as its CEO from May 2007 to October 2008 and again from July 2015 to November 2021. At Block, he has served as principal executive officer and a director since July 2009. His title changed from CEO and president to Block Head in April 2022. These roles span products that help people exchange messages, accept payments and manage money.[2]
Contributions
From a card reader to seller services
Dorsey describes Square’s initial goal in 2009 as making a credit-card reader that worked with the phone a seller already owned. The team broadened that task from accepting a payment to helping a seller run a business: point-of-sale software, online stores, invoices, customer management and banking tools could work together. His contribution as cofounder and executive was to develop that product direction around reducing the systems a merchant had to assemble separately. In the same account, he traces Cash App’s 2013 launch to a simpler way for consumers to send money, which later expanded into a connected set of consumer financial services.[6]
Block’s agent ecosystem
goose is an open-source agent with desktop, command-line and API interfaces that connect to outside tools through extensions. Block donated the project to the Agentic AI Foundation, and its documentation and repositories now live under that foundation. Dorsey’s association is through Block’s leadership and AI strategy. The software provides a practical companion to the organizational essay: an extensible agent can carry out work, while the essay asks how a company might coordinate the people and systems around it.[4][7]
Ideas
In “From Hierarchy to Intelligence,” Dorsey argues that management hierarchies developed partly to move information between people with limited attention. The proposed alternative is a continuously updated representation of the company and its customers, built from work artifacts and transaction data. An intelligence layer would use that context to compose existing capabilities into responses to customer needs. People would build specialized systems, own particular outcomes and develop colleagues as player-coaches. The essay describes Block as early in this transition and acknowledges that parts will break. It is a proposal about organizational design, rather than proof that AI has already replaced management effectively.[3][5]
Tools, writing & talks
Writing, research & talks
Start here
From Hierarchy to Intelligenceblock.xyzAn argument for using shared company and customer models to coordinate work, with individual contributors, outcome owners and player-coaches.
Read this first for Dorsey’s organizational thesis. It explains the proposed operating model without requiring programming knowledge; treat the customer scenarios as the authors’ proposed design.
- Buzz: a hive mind communication platform
github.comBuzz is a self-hostable team workspace built on the Nostr protocol where AI agents participate as first-class members alongside humans, sharing the same channels, identity model, and audit log. Agents can open repos, submit patches, review code, run YAML-defined workflows, edit canvases, and join voice huddles with their own signed keypairs, unifying chat, git hosting, CI, and search into a single event log.
Catalog workOriginal source An in-depth analysis of how organizational structures evolved from Roman military hierarchies to modern corporations, exploring how AI might fundamentally transform how teams work together beyond just productivity gains.
Catalog workOriginal sourceJack Dorsey explains Block's move to replace management hierarchy with an AI-centered intelligence structure, discussing the thinking behind his essay on rebuilding company organization from first principles.
Catalog workOriginal source
Tools
- gooseblock.github.io
A local open-source agent with desktop, command-line and API interfaces, now part of the Agentic AI Foundation.
Originally developed at Block; now an Agentic AI Foundation project[4][7]Original source