How to Reorg After AI Changes Everything | Block's Owen Jennings on the a16z Show
Summary
- Jennings says the first week of December broke the decades-old link between headcount and company output. Opus 4.6 and Codex 5.3 suddenly became capable inside complex existing codebases, letting one or two engineers become “10, 20, 100 x more productive.” After a Q1 review, Block cut slightly more than 40% of its workforce.
- The cut’s composition is Block’s rebuttal to the claim that this was merely a 2021 overhiring cleanup. Reductions were far larger in development, while outbound sales and account management were barely touched; Block also protected compliance and compliance technology. Jennings’s categorical line: “We’re not writing code by hand anymore. That’s over. That’s done.”
- Block replaced feature-team economics with small teams supervising abundant machine labor. Money Bot moved from roughly 15 people to four plus $2,000 of tokens, while Jennings says he now context-switches among as many as 14 agents producing parallel PRs. Meetings fell 70-80%, development layers fell roughly 50-60%, and squads now contain one to six people.
- The operating leverage extends beyond software development into deterministic business workflows. Block’s chatbots and AI phone support automate a majority of inquiries. Block keeps humans in the loop for risk and compliance today, but Jennings expects systems eventually to outperform “a thousand humans” processing those queues.
- Block’s product bet is that static financial-app interfaces will start disappearing within six months. Goose, its model-agnostic harness with access to probably 120 models, underpins Cash App’s Money Bot and Square’s Manager Bot. Generative UI can build customer-specific charts or even a restaurant scheduling app on demand, creating engagement upside alongside a “potentially a nightmare” QA problem across tens of millions of users.
- Near-term defensibility remains distribution, regulation, network effects, and hardware; long-term defensibility becomes proprietary understanding. Anyone might build peer-to-peer software in a week, but not “vibe code” 50-60 million monthly active users. Block’s intended moat is a rapid feedback loop around its distinctive signal—how buyers and sellers participate in the economy—because companies unable to name what they uniquely understand “maybe could get vibe coded away.”
- The AI transformation has not yet resolved Block’s public-market disconnect. Haber notes that the business and gross profit per employee grew while the stock stayed roughly flat for six or seven years; Jennings concedes that the roughly $260 price in 2021 was “a little bit irrational.” His answer is deliberately long-term: markets are voting machines now, weighing machines later, so “just focus on building.”
Deep dive
1. December broke the headcount-output equation
Jennings’s chronology begins with Jack being “generally right and generally early. Sometimes very early.” Block launched Goose—what Jennings calls the first agent harness he knows of—in early 2024, then spent 2024 and 2025 building tooling around it.
Late November and early December brought the discontinuity: Opus 4.6 and Codex 5.3 crossed from strong greenfield coding into complex existing codebases. With one or two tool-using engineers producing 10, 20, or 100 times more, Jennings says the old headcount-output correlation “basically broke.”
Haber’s pushback—worth keeping—is that the reduction might reflect 2021 overhiring. Jennings counters that Block sat around the peer-group middle on gross profit per employee from 2019 through 2024, perhaps in the second quintile last year, with Nvidia and Meta basically ahead; a cleanup of “cruft and bloat” would have landed in operations, not disproportionately in development.
2. Block rebuilt the organization around three non-negotiables
Because Block was coming from strength on profitability and operating income, management did not begin with a CFO-mandated percentage cut. It asked what the organization should look like given current tools and expected progress over the coming quarters.
The rebuild centered on reliability, customer trust and regulatory compliance, then durable growth. Compliance and compliance technology were essentially untouched; known roadmap work continued, but a feature might now require “a squad of three people instead of a feature team of 14.”
Jennings emphasizes the human execution: generous severance, no immediate technology lockout, and a companywide explanation from Jack and the executive team. After the Thursday announcement and a shocked weekend, meetings dropped 70-80%; weekly Monday all-hands and fewer layers helped make it feel “back to building.”
3. Work shifted from linear production to agent supervision
Jennings contrasts Block’s single large move with repeated 15% cuts that leave another layoff “looming over your shoulder.” The larger move also became a “massive forcing function” for changing workflows.
In Jennings’s Money Bot example, a team of roughly 15 becomes four people plus $2,000 of tokens, with unlimited token access and fast mode in Claude Code. Instead of sequential PR work, Jennings describes his own workflow: 14 agents building PRs in parallel while he checks, nudges and commits their output.
The structure follows the workflow: flexible squads now contain one to six people, and Jennings estimates development layers fell by about 50-60%. He says his product organization has only two layers, perhaps three in a couple of places; designers and product managers are shipping PRs.
Internal-only G2 lets anyone automate deterministic workflows. Builder Bot can autonomously merge PRs and occasionally complete complex features end to end; 85-90% completion is more typical. AI chat and phone support automate a majority of inquiries, while Block currently keeps humans in the loop as it handles risk, compliance and its relationships with partners and regulators.
4. Goose makes Block’s ecosystem an AI product surface
Block abandoned separate Square and Cash App business-unit hierarchies about 18 months ago, functionalizing engineering, design and product across Square, Cash App and Afterpay. Jennings estimates Cash App now contributes about 60% of gross profit, while the strategy increasingly thinks across the three businesses as an ecosystem.
Goose is the common substrate: a model-agnostic harness that can run Anthropic, OpenAI or open-source models; Jennings says Block has probably 120 models available. Cash App’s proactive Money Bot—“a CFO in your pocket”—and Square’s Manager Bot both sit on top of it.
Jennings expects static UI to change fundamentally within six months. Manager Bot could generate a custom scheduling app for a multi-location restaurant, including employee messages over WhatsApp or Signal, without that interface existing in the App Store source code. The upside is personalization, engagement and discovery; QA for nondeterministic outputs could be “potentially a nightmare” across tens of millions of customers, so Block is investing in proactive intelligence for customers who may not know what to prompt.
5. The moat migrates from code to hard-won understanding
Jennings sees near-term moats including distribution, network effects, licenses, regulatory posture and hardware. Anyone can create peer-to-peer functionality quickly, but cannot “vibe code” 50-60 million monthly active users or a piece of Square hardware.
Longer term, Block is building toward an intelligent system with world models of customers and itself. Its distinctive signal is how sellers and buyers participate in the economy; a company-level Markdown file can encode values and metrics, while Builder Bot or Claude Coder repeatedly turns that understanding into products.
That loop has already compressed feature delivery from months to perhaps one or two weeks. Jennings expects it might eventually run hundreds or thousands of times daily, with humans possibly acting more like editors.
Fewer engineers, designers and PMs per roadmap does not necessarily mean fewer globally: Jennings invokes Jevons paradox, imagining many more products, 50 or 100 more technology companies, and software development spreading into new sectors. His sharper warning is firm-specific: if a company cannot say what it uniquely understands, “you maybe could get vibe coded away.”