Block has made company-wide automation its number one priority, developing an internal, open-source AI agent named Goose. This tool is used across the organization, from engineering to enterprise security, to automate tasks and is projected to save 20-25% of manual work hours.
The transition from a decentralized General Manager (GM) model to a centralized functional organization was a critical prerequisite for Block's AI success. This change broke down silos, enabled platform leverage, and shifted the mindset away from treating engineering headcount as a commodity.
Block has made its core AI agent, Goose, entirely open-source, encouraging adoption and extension by other companies. This strategy contrasts with the closed, 'walled garden' approach of many tech giants, aiming to build a collaborative community around the technology.
One of the most significant impacts of AI at Block has been the empowerment of non-technical employees. Staff in roles like enterprise security are now using AI agents and programming tools to build their own solutions, demonstrating a lower barrier to software creation.
Despite the focus on automation, Block's leadership emphasizes that AI currently underperforms senior humans in complex, high-judgment tasks like system architecture and strategic decision-making. Human taste and a critical mindset are seen as essential to guide AI and prevent low-quality output.
Keep pulling the thread on Dhanji R. Prasanna.