# Jack Dorsey: Every Company Can Now Be a Mini-AGI

Podcast: Long Strange Trip: CEO to CEO with Brian Halligan
Published: May 28, 2026
Reading time: 19 min
Canonical: https://podbrew.app/briefs/long-strange-trip-ceo-to-ceo-with-brian-halligan-jack-dorsey-every-company-can-n

Jack Dorsey, CEO of Block, and Roelof Botha, a Sequoia partner and Block board member, present a groundbreaking perspective on corporate design. They challenge the very foundation of traditional hierarchies, asserting that these long-standing structures are not just inefficient, but fundamentally obsolete in the modern business landscape.

Their discussion centers on Block's recent, dramatic restructuring, which involved cutting 40% of its workforce to rebuild around an AI "intelligence layer." They detail a bold vision for simplifying organizations into just three core roles, fundamentally replacing the traditional pyramid org chart with a circle, placing AI at the center and people at the edge. The aim is to create a company that functions as a highly efficient, real-time "mini-AGI."

This conversation is essential for any founder or leader seeking to navigate the transformative power of AI. It offers a provocative blueprint for reimagining company operations, emphasizing proactive, AI-driven transformation rather than incremental adjustments. The insights provide a pathway to build an organization truly fit for the future, enabling unprecedented agility, insight, and responsiveness to customer needs.

## Key takeaways

- Digital work artifacts (messages, code, recordings) can be aggregated to form an AI-driven "world model" of a company.

- This AI model allows all employees to directly access and understand company operations, transforming the organization into a more efficient "mini AGI."

- Block is integrating an AI 'intelligence layer' to provide real-time, queryable access to company data for board members, analysts, and employees, enhancing decision-making and information flow.

- The company is restructuring around three core roles: ICs (Individual Contributors/builders augmented by AI agents), DRIs (Directly Responsible Individuals owning customer outcomes and strategy), and Player Coaches (managers who teach and develop skills by doing the work alongside their teams).

- Block aims to significantly flatten its organizational hierarchy, reducing management layers from five to two or three, to improve company velocity and roadmap execution by facilitating direct communication and accountability.

- The CEO's role is evolving from traditional oversight to architecting the company as an intelligent system, continuously aligning it with customer outcomes.

- Traditional company roadmaps are becoming a limiting factor; the future demands the ability to serve customer-requested features not on existing plans.

- AI should be viewed as a catalyst for complete business transformation, requiring companies to rebuild their core operations around AI rather than merely augmenting existing processes.

- Companies that fail to fundamentally integrate AI and only treat it as a productivity tool will struggle to differentiate and survive in the new structural reality.

- The "Dorsey Mode" proposes an organizational structure where AI-powered systems provide "ground truth" signals, enabling flatter, more productive teams by replacing traditional hierarchical decision-making.

- Implementing an effective non-hierarchical structure requires an intelligent system capable of processing vast amounts of data to provide the necessary signals for informed decision-making, a capability previously missing for companies like HubSpot.

- Traditional product roadmaps hinder adaptability; companies should instead offer capabilities that empower customers to co-create, acting as partners in development.

- Making an entire company "legible" by collecting and centralizing all generated data into an intelligence system offers unprecedented understanding, surpassing human-dependent reporting.

- Block's leadership determined that the optimal size and structure of their company had changed due to the capabilities of new AI tools.

- The company conducted a data-driven exercise to identify the minimum staffing required for critical operations, regulatory compliance, and strategic growth, leading to a significant workforce reduction.

- Avoid dual CEO structures as they can lead to fragmented cultures and hinder the integration of distinct business units, as seen with Block's Square and Cash App.

- The ultimate success of AI-driven projects still relies on human creativity, taste, and judgment, which contribute the final and most critical 20% of the work.

- Successful leaders must be prepared to challenge conventional thinking and risk stakeholder credibility to pursue groundbreaking ideas, as demonstrated by Block's expansion into lending and Cash App.

- Timeless CEO qualities, summarized as ALE (Authenticity, Logic, Empathy), are crucial for effective human leadership, regardless of technological changes.

- In an AI-driven world, leaders must constantly reprogram their minds by questioning assumptions and challenging past decisions to adapt to rapid change.

## 00:00 - 06:04 Jack Dorsey proposes an AI-centric company structure to replace traditional hierarchies

Traditional company hierarchies have governed organizations for over 2000 years, primarily to manage information flow at a human scale. This age-old structure is now being questioned as fundamentally inefficient for modern businesses.

In today's remote-first work environments, every action generates digital artifacts like Slack messages, emails, pull requests, and recorded meetings. These artifacts contain rich data about how the company operates, builds, and even makes mistakes.

Jack Dorsey suggests leveraging these digital traces by building an "intelligence" or AI model on top of them. This model can create a "world model" of the company, moving beyond the traditional reliance on human managers to relay information up and down the chain.

This AI-driven approach allows anyone within the company to query and interact with the model to understand the company's state and capabilities. It effectively treats the organization as a "mini artificial general intelligence," optimizing information flow and overall efficiency.

> Treat the company as a mini AGI, an artificial general intelligence, because it really is.

## 06:04 - 12:05 Block Reorganizes its Structure with an AI Intelligence Layer and Three Core Roles

Block is implementing a new 'intelligence layer' to allow real-time querying of company data. This system aims to enable board members to directly ask questions and have conversations with the company's intelligence, shifting board meetings to focus more on creative and existential decisions rather than day-to-day issues. This capability can also scale to earnings calls, providing analysts with timely, Regulation FD-compliant information tailored to their questions, which was previously impossible.

The company is simplifying its organizational structure by normalizing down to three core roles. The first is an IC (Individual Contributor), which includes builders and operators like engineers, designers, and salespeople. These ICs are augmented by AI agents, allowing one person to potentially achieve the breadth of work previously requiring a team, with their durable human skills being judgment, taste, and creativity.

The second role is the DRI (Directly Responsible Individual), who owns customer outcomes by developing strategy and roadmaps, and assembling IC teams. Their core human skills are ownership and accountability for the results. The third role is the Player Coach, which is akin to current managers but focuses on building the capability and capacity of others by demonstrating the work themselves rather than just directing. Their durable skills involve coaching, empathy, and building human capacity.

The goal is to flatten the company's hierarchy significantly, aiming to reduce the maximum depth from the CEO to any individual from five layers down to two or three, ideally even fewer. Leaders at Block, including the CEO and their direct team, are expected to embody all three roles: building as an IC, strategizing as a DRI, and coaching as a Player Coach to foster a more integrated and agile organization.

> imagine if every single board member can just query the company and have a conversation with the company's intelligence in real time

## 12:05 - 16:06 The CEO as Architect of Company Intelligence and Customer-First Product Roadmaps

Jack Dorsey describes an evolution in the CEO's role. Initially, he focused on establishing principles, fostering team dynamics, making customer-centric decisions, and consistently raising execution standards. This traditional view emphasized internal alignment and capability building.

However, Dorsey now envisions the CEO's core function as architecting the company as a living intelligence. The goal is to constantly align this company intelligence toward optimal customer outcomes, with humans on the periphery guiding this central 'world model' of the company.

A critical shift is the belief that a company's own roadmap can be its ultimate limiting factor. Emerging technologies suggest that customers will increasingly expect to request features not on any existing roadmap and have them delivered instantly. This requires a new approach to building foundational capabilities that can be flexibly composed and delivered through various interfaces like Square or Cash App.

The future of product development involves proactive intelligence and a customer-driven roadmap. Leveraging deep insights from transactional data, which Dorsey calls the 'most honest signal in the world,' companies can proactively offer valuable services, like protecting customer cash flow. When customers make requests, these interactions directly highlight any deficiencies in capabilities, effectively defining the company's real-time roadmap based on genuine customer needs.

> A company's ultimate limiting factor is its own roadmap.

## 16:06 - 20:07 Jack Dorsey Argues AI Demands a Complete Business Transformation, Not Just Productivity Gains

Jack Dorsey asserts that artificial intelligence represents a fundamental structural shift for businesses, not merely a tool for productivity enhancement. He warns against viewing AI solely as a "copilot" or an augmented feature, urging companies to instead consider how to rebuild their entire operations with AI at the core.

He emphasizes the critical role of human judgment in this AI-centric world. Businesses must ensure that their AI implementations align with their intended values, taste, and uniqueness. This involves building companies that deeply understand human nature, providing a tangible signal that helps define their intelligence and differentiation.

Dorsey shares his personal experience, dedicating three hours daily to pushing AI tools to their limits, consistently finding surprising capabilities. He stresses that companies not undergoing this core transformation risk looking too similar to "Frontier Labs" and will struggle significantly to differentiate and survive in the evolving landscape.

> I don't think this is a productivity thing, I think it's a structural thing that needs to shift.

## 20:07 - 24:07 Dorsey Mode: Ground Truth Systems Over Hierarchy

A new organizational model, dubbed "Dorsey Mode," proposes an alternative to traditional hierarchical management or even founder-centric flat structures. This mode envisions an organization where intelligence derived from a "ground truth" system, often AI-driven, takes precedence over human-led decision-making, aiming to recapture the high productivity of early-stage companies.

This ground truth system provides clear signals, much like a capitalist market, dictating what products to build based on actual demand rather than internal power dynamics or an individual's persuasive ability. The goal is to eliminate layers of hierarchy and foster a highly transparent environment where the right signals inform collective work, leading to more productive and agile teams.

The core mechanism of Dorsey Mode is a customer-centric feedback loop. Instead of relying on traditional market research, interviews, or inferring needs, an AI-powered conversational interface with customers directly reveals what they care about and what they want. This high fidelity data effectively allows customers to shape the product roadmap, with human teams then deciding on strategic alignment.

Previous attempts to flatten organizations and eliminate hierarchy, such as HubSpot's nine-month experiment with no org charts or titles, often failed due to the absence of an intelligent system to provide these crucial decision-making signals. The advent of sophisticated AI makes this "ground truth" system viable, addressing a fundamental challenge in non-hierarchical structures.

> when your interface is a conversation with your customer instead of like this visual navigation, you suddenly get like this amazing fidelity of like, what do our customers actually care about? What do they actually want?

## 24:07 - 28:07 Removing the Roadmap for Customer Co-Creation

Companies are facing an existential challenge, grappling with questions about what a durable company structure looks like in a rapidly changing world. Jack Dorsey argues that a company's relevance and competitive advantage must stem directly from its ability to meet evolving customer expectations.

The example of OpenAI illustrates a fundamental shift: customers, even non-technical users, desire agency—the ability to tangibly control and customize intelligence for their specific needs. This trend significantly raises the baseline for what customers expect from products and services.

To adapt, companies must abandon traditional roadmaps, which limit innovation. Instead, they should provide a set of capabilities and an intuitive interface, allowing customers to actively build alongside them. The focus shifts to intelligent systems that can compose user interfaces in real time based on customer desires.

While building a company as an intelligent system from its inception offers immense potential for rapid prototyping and development, the challenge of distribution and capturing customer attention in a noisy market remains substantial.

> our limiting factor as a company is our roadmap. Like, we need to remove that from the equation. We need to ensure that our customers are truly building alongside us

## 28:07 - 32:09 Questioning hierarchy and embracing company legibility through data

Jack Dorsey suggests that even companies of around 100 people should critically examine the need for traditional hierarchies. He shares his experience at Disqwer, where they removed formal titles, normalizing everyone to a "lead" role, such as "Lead of X." This approach helps clarify individual responsibilities and avoids the rigid expectations often associated with corporate titles, which can be particularly relevant when interacting with larger, more hierarchical organizations like banks.

Dorsey also emphasizes the importance of making an entire company "legible." This means leveraging all information and data generated by daily work, feeding it into an intelligence system, and making it queryable. Such a system provides a far deeper understanding of the organization than relying on human reporting, which can be influenced by agendas, politics, or emotions.

The goal is to evolve from basic legibility to proactive insights and eventually predictive world models around the company and its customers. While determining causality and achieving predictability are still research challenges, Dorsey believes they are solvable. He notes that AI models have recently advanced significantly in understanding large and legacy codebases, moving beyond just prototyping to mature capabilities, hinting at the feasibility of this vision.

> Imagine if your company was entirely legible, like entirely legible, every aspect of it, and we're not far off from that from a data perspective, it's putting the intelligence on top of it and making it useful and then making it proactive.

## 32:09 - 36:11 Block Reduces Workforce Based on AI's Impact and Reimagined Company Structure

Block's leadership team fundamentally re-evaluated their company's structure after experiencing the capabilities of new AI tools. A key question posed was whether they would build Block the same way if these advanced tools were available from the start. The unanimous answer was that the company would not be its current size or structured in the same manner.

This realization prompted an immediate strategic exercise to redefine essential staffing levels. Block identified the minimum number of people required to maintain 100% service uptime, ensure full compliance with financial regulators, and meet growth commitments while rebuilding as an intelligence-focused organization. This data-driven analysis directly informed the subsequent significant reduction in the workforce, which was executed rapidly within three weeks.

Jack Dorsey emphasized the intention to be proactive rather than reactive. By making these difficult decisions ahead of market pressures, Block aimed to conduct the process with greater integrity and generosity towards departing employees, while also setting a tone of acting towards excellence. This forward-thinking approach allowed the company to reshape its operations on its own terms.

> Would you build the company this way if you had these tools today? Like, what would the company look like? And everyone around the table and my team just said like, it wouldn't look like this, it wouldn't be this size, it wouldn't be structured this way.

## 36:11 - 42:11 Building an Effective Board with Trust and Diverse Perspectives

When bringing on investor board members, it is crucial to treat this decision as a critical hire, not just a financing agreement. Founders should prioritize the individual's character and alignment over brand names, understanding that this relationship is long-term and that board members can ultimately replace the CEO. This perspective puts significant pressure on selecting the right people from the outset, focusing on those who can elevate conversations and execution.

The primary role of a board is to ensure the company has the correct CEO moving forward. To fulfill this, boards must be composed of individuals with diverse perspectives, open to challenging existing ideas and even embracing "wild ideas." This openness helps avoid stagnation and the demoralizing cycle of constant downsizing, fostering a more creative and inspiring environment. A board that can rationalize and document unconventional paths is essential for long-term growth.

It is advisable to integrate independent board members early, ideally within one to two years of a company's inception, or certainly by the time product-market fit is achieved. Unlike investor board members, independents can often provide a different, more purely mentorship-focused relationship with the founder. Many companies err by building out their boards too late or haphazardly, especially in the run-up to an IPO, resulting in members who lack context and history, which can be detrimental when facing significant challenges like short-seller reports or activist investors.

> I think the first financing, when you get an investor, if you're getting an investor board member, you should treat it as a recruiting decision. Not a financing decision, because it'll have a much bigger bearing on the ultimate outcome of your company.

## 42:11 - 46:13 Jack Dorsey warns against dual CEO structures and emphasizes continuous learning from regrets.

Jack Dorsey identifies having dual CEOs as an "anti-pattern" based on his experience at Block. He initially aimed for multiple CEOs across Square and Cash App but realized this created a holding company rather than fostering synergy to challenge the financial network. This structure ultimately led to differing cultures, values, and execution levels across the business units.

Dorsey's biggest regret is not learning fast enough from over-delegation. He states that while he embraces mistakes, failing to actively learn from them is his true regret. Specifically, he delegated too much within Block, which contributed to the fragmentation he observed across its various services.

To stay relevant, Dorsey stresses the necessity of continuous building and learning from what the company puts out. He foresees future regrets stemming from a failure to inject enough "entropy" or intent into the system to maintain relevance, highlighting a need for constant adaptation and growth.

The company has shifted its meeting culture, moving from presentations to prototypes. Teams now bring actual prototypes with simulated or real data, allowing for real-time modification and deeper conversations about what is being built. This change significantly broadens the exploration scope and aids in better judgment.

> My only regrets in life and, and also in our, in our businesses, are where I decided not to learn something.

## 46:13 - 50:13 Leveraging AI for Agile Exploration and the Value of Human Judgment

New development tools and AI significantly reduce the cost and time required to explore different product paths. What once took a long time and costly prototypes, especially in hardware, can now be achieved in an hour, making it feasible for organizations to try multiple ideas in parallel rather than being strictly focused on one path.

While these advanced tools can expedite about 80% of the development work, the final 20% remains dependent on human creativity, taste, and judgment. This crucial human element is where true innovation and the 'magic' of a product emerge, constantly pushing models beyond their initial perceived capabilities.

Pursuing these innovative but often initially unpopular ideas requires strong leadership. Leaders must be comfortable risking credibility with stakeholders to champion new ventures. Block, for instance, faced skepticism from its board when venturing into lending and from its own company when developing Cash App, but these initiatives ultimately proved critical to its success.

> these tools will build about eighty percent of what, where we need to go, and then that last twenty percent is gonna be How good our creativity, how good our taste is, how good our judgment is.

## 50:13 - 54:14 Lead with Conviction and Embrace Infinite Mentors

Jack Dorsey recounts his experience with Square Cash App, which initially faced strong resistance from investors and caused him to lose credibility internally. Despite the skepticism, he stood by the product, driven by its core principle. Cash App eventually grew to become over half of Square's business, illustrating the necessity for leaders to accept temporary credibility loss when pursuing bold, conviction-driven ideas.

Dorsey later re-evaluated his approach to mentorship after finding traditional CEO coaching unhelpful. He adopted an "infinite mentors" philosophy, choosing to see every person, problem, and encounter as a potential source of learning. This mindset encourages leaders to actively seek wisdom from all interactions, rather than depending on a singular mentor.

The practice involves deliberately deciding what to learn from each experience and reflecting on those lessons daily. This approach grants agency and ownership, transforming even negative feedback or setbacks into valuable teaching moments. Dorsey's biggest regrets stem from instances where he failed to extract a lesson from a situation, emphasizing the importance of this continuous learning process.

> Every single person I talk with, every single encounter I have, every single problem I face, that's my mentor.

## 54:14 - 56:14 Meditation Trains Focus, Observation, and Intentional Action

Meditation is often mischaracterized as a 'woo-woo' spiritual practice involving passive contemplation. However, it is a highly physical and rigorous discipline designed to train the mind, much like physical exercise trains the body. It involves intense, prolonged focus and observation.

During meditation retreats, practitioners might spend ten days in silence, dedicating the first three days to concentrating solely on the physical sensation of their breath on their upper lip. The subsequent seven days involve scanning the body for various sensations, including pain from sitting cross-legged for three hours without moving. This sustained practice sharpens one's ability to focus intently and observe phenomena without immediate emotional or intellectual reaction.

The core of this practice is to understand the impermanence of all sensations and thoughts. By observing pain and knowing it will pass, the mind learns not to suffer or attach itself to transient experiences. This principle extends beyond physical sensations to emotions and life encounters, providing a framework to see things as they truly are before consciously choosing how to respond, rather than instinctively reacting.

Engaging in this physical practice enhances focus, refines the power of observation, and reduces the impulsive instinct to react. This allows individuals, especially those in demanding roles like CEOs, to approach situations with greater clarity and intentionality, making it a valuable skill for decision-making and leadership.

> So I would recommend it only because it Sharpens your focus, it sharpens your power of observation, and it diminishes your instinct to immediately react to things and to actually see them for what they are and then choose how you wanna act with that information.

## 56:14 - 1:03:32 Timeless CEO Qualities and the Imperative for Constant Reprogramming

Roelof Botha identifies three timeless qualities for CEOs: Authenticity, Logic, and Empathy (ALE). He explains that authenticity means being genuine, logic involves being predictable and rational, and empathy requires caring deeply about the team and the business. These qualities remain fundamental for effective human leadership, even as technology transforms industries.

Jack Dorsey emphasizes a new critical quality for leaders in the current fast-paced, AI-driven environment: the ability to constantly reprogram one's mind and question assumptions. He believes it is increasingly valuable to challenge one's own requirements, past decisions, and company traditions, rather than rigidly adhering to them.

With the rise of AI, Dorsey notes that people are prone to going along with the momentum and offloading their intelligence to tools. He warns against viewing AI's suggestions as definitive outputs, advocating instead for seeing them as better inputs to create superior human-generated outputs.

Maintaining a unique, opinionated point of view is paramount for leaders. This involves the ability to decipher signal from noise and drive toward a specific vision. Founders, in particular, build something that didn't exist because they wanted to see it, rather than creating copies of existing ideas, which is becoming easier with AI tools.

> We're still in a mode where most people are seeing all what the intelligences do, these tools do as output rather than better input to create better output ourselves.

---

Get podcast briefs for shows you follow: https://podbrew.app/