Jack Dorsey: AI to Replace Middle Management at Block

Block (formerly Square) co-founder Jack Dorsey has stated that artificial intelligence (AI) should replace the role of the middle manager following the company’s recent workforce reduction. Approximately 4,000 employees, representing around 40% of Block’s 10,000+ workforce, were impacted by the cuts.

Restructuring Driven by AI Capabilities

Dorsey argues the restructuring wasn’t a cost-cutting measure, but a permanent shift to leverage AI for coordination, product decisions, and internal alignment. He identified a significant capability shift in AI tools, including Anthropic’s Opus 4.6 and OpenAI’s Codex 5.3, in December, noting their ability to effectively operate within large codebases.

The Role of AI in Organizational Structure

According to Dorsey, traditional corporate hierarchies were created to manage information flow within large organizations. He believes AI is now capable of addressing this challenge. In a recent essay, “From Hierarchy to Intelligence,” co-authored with Roelof Botha, managing partner at Sequoia Capital (an investor in Block), Dorsey detailed this vision.

AI as a Continuous Information Aggregator

The essay posits that AI can continuously aggregate internal data – including code, decisions, workflows, and performance metrics – to provide a real-time picture of company operations. This replaces the contextual understanding traditionally held by managers. This data feeds into what Block calls an “intelligence layer” which dynamically composes financial products based on market demand.

A New Organizational Model

Block plans to operate with three core roles: individual contributors focused on building the system, directly responsible individuals owning 90-day outcomes, and player-coaches who develop people while remaining hands-on. The company intends to break down its business into modular capabilities like payments, lending, and payroll.

System-Driven Product Development

Instead of relying on fixed roadmaps, Block aims to use the AI system to identify needs – such as a merchant facing a cash flow gap – and assemble solutions from existing capabilities. If a capability is missing, the system will prioritize its development, effectively generating a system-driven backlog.

Despite the potential, Dorsey acknowledges that roughly 95% of AI-generated code changes still require human review, and AI tools are not yet suitable for highly regulated areas like banking and money transfers.