The PodKnowledge|← All Episodes
Based on Lenny's Podcast data
Lenny's Knowledge Sketch

How Block Became the Most
AI-Native Enterprise

Dhanji R. Prasanna
CTO, Block (formerly Square)
OCT 26 2025
The Bet

Block Rewrote Itself
for the AI Era

AI INTEGRATION DEPTH
"I think we should do this. I think we should do it centrally and it's important for us to be ahead of the game and be an AI native company."
  • Dhanji's AI manifesto: a memo he wrote to Jack Dorsey while a part-time engineer, arguing Block should push AI centrally
  • Jack came to Sydney, walked with him for two days, then offered him the CTO role
  • The transformation wasn't AI features. It was moving from a GM structure to a functional org, one head of engineering, one head of design
  • Dhanji's number one priority now: "automate Block"
Framework

Block's AI-Native Architecture

GM STRUCTUREFUNCTIONAL ORGAUTOMATE BLOCK
8-10 hrs
saved per week on AI-forward teams
20-25%
of manual hours saved company-wide
50%
of test runs cut
  • The org shift: GM structure (Square, Cash App, Afterpay, TIDAL as portfolio companies) to one functional org, all engineers under one head, all designers under one head
  • Conway's Law: "you ship your org structure" — your operating model matters a lot to what you build
  • Goose: Block's open-source AI agent, built on the Model Context Protocol (MCP)
  • Most surprising use: non-technical teams (enterprise risk management) building their own tools in hours, not quarters
Dhanji's number one priority"Automate Block. Getting AI and getting AI forms of automation through our entire company."
Enterprise AI Reality

What Block Learned

  • Works: code generation, test automation, non-technical teams building their own software, background CI checks analysing vulnerabilities and drafting patches while engineers sleep
  • Where AI still lags: "architecture and design and race conditions, orchestration"
  • Unexpected: non-technical people using AI to build things "are really showing the most impact from these tools"
  • Adoption skews to both ends: "the more senior and the more junior they are, the more comfortable or the more eager they are to adopt these AI tools"
The non-technical surprise

Enterprise risk built its own self-service system in hours instead of waiting a quarter for an internal apps team. Dhanji calls this "one of the most surprising and energizing uses of Goose within Block."

Where senior humans still win

Senior engineers thinking about architecture, design, race conditions, and orchestration are where AI "isn't quite there." Throwing AI at giant legacy code bases and hoping "is not how it's playing out."

Playbook

Build an AI-Native Company

  • Use the tools yourself. Jack, Dhanji, and the executive team all use Goose every single day, alongside other AI programming tools
  • Restructure before you re-tool. Moving from a GM structure to a functional org was "the key to making our transformation into being more of an AI-native company"
  • Start small. Goose began as one engineer's side project; Cash App and Block's Bitcoin product both grew from hack week ideas
  • Hire for a learning mindset. Jack wants Block to be "a learning first company"
Dhanji's number one priority"Our number one priority is to automate Block, which means getting AI and getting AI forms of automation through our entire company."
Contrarian

What Dhanji Learned That Goes Against the Grain

✗Code quality is critical to a successful productINSTEAD →✓ "The two have nothing to do with each other." YouTube stored videos as blobs in MySQL and became one of Google's biggest products.
✗The fastest path to productivity is more AI toolsINSTEAD →✓ "It's not wrong in some ways." For Block, moving from a GM to a functional org structure changed how work happened more than any tool.
✗Engineers benefit most from AI toolsINSTEAD →✓ The non-technical people using AI agents to build things "are really showing the most impact from these tools."
✗Never rewrite from scratchINSTEAD →✓ Dhanji pushes his teams to imagine "what would our world look like if every single release rm -rf'd the entire app and rebuilt it from scratch."
Based on Dhanji R. Prasanna's episode on Lenny's Podcast. All ideas on this page are from the episode.Watch on YouTube
GO DEEPER IN THE EPISODE
05:26 The AI manifesto: convincing Jack Dorsey07:33 Transforming into a more AI-native company12:05 How engineering teams work differently today15:24 Goose: Block’s open-source AI agent20:18 Measuring AI productivity gains across teams21:38 What Goose is and how it works32:15 The future of AI in engineering and productivity37:42 The importance of human taste40:10 Building vs. buying software44:08 How AI is changing hiring and team structure53:45 The importance of using AI tools yourself before deploying them55:13 How Goose helped solve a personal problem with receipts58:01 What makes Goose unique59:57 What Dhanji wishes he knew before becoming CTO01:01:49 Counterintuitive lessons in product development01:04:56 Why controlled chaos can be good for engineering teams01:08:07 Core leadership lessons01:13:36 Failure corner01:15:50 Lightning round and final thoughts
Every sketch here began as a transcript.

Send us yours and we’ll draw it. $1 a transcript.

Sketch mine →
𝕏︎ X / Twitterin LinkedIn📸 Instagram🔗 Copy link
0:00