Owen Jennings serves as Executive Officer and Business Lead at Block, overseeing product operations and customer support across Square, Cash App, and Afterpay. Previously CEO of Cash App during its critical scaling period, he has led Block through a comprehensive AI transformation that resulted in a 40% workforce reduction.

The conversation explores Block's radical restructuring around AI capabilities, examining the decision-making process behind the massive workforce reduction and how the company rebuilt itself around small squads working alongside AI agents. Jennings discusses the technical infrastructure powering this transformation, including BuilderBot for autonomous feature development and the Goose agent harness supporting 120+ AI models.

The discussion covers both internal operational changes and customer-facing product innovations like MoneyBot and ManagerBot, which generate custom interfaces dynamically. Jennings also addresses the broader implications for the tech industry and how companies can maintain defensibility in an AI-driven landscape.

The December Breakthrough That Changed Everything

Block identified a "binary change" in AI capabilities in late November/early December 2025, when models like Opus 4.6 and Codex 5.3 became capable of working with complex existing codebases, not just greenfield projects.

"There's been this correlation between the number of folks at a company and the output from the company for decades and decades. I think that basically broke" - Owen on the fundamental shift in the headcount-to-output relationship.

The company spent Q1 2025 as an executive team working through the implications of AI capabilities for both product development and company operations before executing the restructuring.

Executing a 40% Workforce Reduction Around AI

Block's reduction was strategically concentrated on development teams where AI showed the most impact, while areas like outbound sales and account management saw minimal cuts.

The company operated from a position of financial strength, focusing on organizational optimization rather than cost-cutting: "We said, what should the org look like given how these AI tools are flowing now."

Core principles included maintaining reliability (no outages), preserving customer trust and regulatory compliance, and continuing durable growth with smaller teams.

The execution included generous severance packages, maintaining technology access, and transparent communication through company-wide meetings with leadership.

Rebuilding Around Small Squads and AI Agents

Block restructured from traditional hierarchical teams of 8+ engineers to small squads of 1-6 people with significantly more flexibility to move between products and projects.

The company reduced organizational layers by 50-60% on the development side, with product having only 2-3 layers total to enable faster information flow.

"I have 14 agents who are building PRs on my behalf right now, and I'm going to context switch between all of those" - Owen describing the shift from linear to parallel AI-assisted workflows.

Meeting volume decreased by 70-80%, with weekly company-wide all-hands replacing frequent smaller meetings to maintain alignment.

BuilderBot and Autonomous Feature Development

BuilderBot autonomously merges pull requests and builds features to production, typically achieving 85-90% completion without human intervention on complex features.

All designers and product managers now ship code directly, fundamentally changing traditional role boundaries in software development teams.

The time from idea to customer deployment has been "compressed massively since December," enabling rapid iteration and feature delivery.

Block provides unlimited access to AI tokens and fast-mode Claude Code, removing resource constraints on AI-assisted development.

Goose Agent Platform and Cross-Company Automation

Goose serves as Block's model-agnostic agent harness supporting approximately 120 different AI models, allowing optimal model selection based on specific tasks.

The G2 agentic operating system enables anyone at the company to automate deterministic workflows, extending AI capabilities beyond just development teams.

Customer support automation handles the majority of inquiries through AI chatbots and phone support, with human-in-the-loop oversight for complex cases.

"Generally, the models and the agents are going to do a better job than humans" for product operations, risk operations, and compliance decisioning workflows.

MoneyBot, ManagerBot, and Generative User Interfaces

MoneyBot functions as a "CFO in your pocket" for Cash App users, proactively generating personalized financial insights and taking actions on behalf of users.

ManagerBot creates custom applications on-demand for Square merchants, such as multi-location scheduling apps with automated employee communication via WhatsApp or Signal.

Generative UI technology creates personalized interfaces that aren't stored in source code: "Your Cash App should look really different from mine" based on individual usage patterns and preferences.

The shift from static to dynamic interfaces presents new challenges: "It's potentially a nightmare from a QA perspective" to test non-deterministic outputs for tens of millions of users.

Future Defensibility and the Jevons Paradox Effect

Owen referenced the classic Jevons paradox concept, suggesting that while individual companies may need fewer engineers, the total number of tech companies and development opportunities could increase dramatically.

"The biggest moat is going to be which companies understand something that's super hard for other people to understand. And if your answer to that is I don't know, then you maybe could get vibe coded away."

Block's defensible insight centers on "how sellers and buyers participate in the economy," combined with their ability to iterate rapidly on that understanding using AI tools.

The company envisions operating as an "intelligent system" that continuously improves its world models through feedback loops between proprietary data insights and AI-powered development capabilities.