On March 31, 2026, Oracle laid off 30,000 people via a 6 a.m. email. No warning from HR. No conversation with a direct manager. Just a termination notice and locked accounts by sunrise. It was the largest layoff in the company's 49-year history.

Oracle is not alone. Bayer has eliminated over 12,000 positions since 2023 under a restructuring explicitly designed to strip out management layers and reduce bureaucracy. Block, Amazon, Meta, C.H. Robinson, and dozens of others have made similar cuts, many citing AI directly. According to Challenger, Gray & Christmas, AI led all stated reasons for layoff announcements in March 2026, cited in 25% of that month's job cuts, up from 10% in February and just 5% for all of 2025. This is not a tech story. It is an organizational one.

Jack Dorsey, co-founder of Twitter and CEO of Block (the parent company of Square and Cash App), traces the lineage of corporate hierarchy in a recent essay, "From Hierarchy to Intelligence," co-authored with Sequoia Capital's Roelof Botha. When Dorsey talks about organizational structure, people listen: he's built and scaled two of the most consequential technology companies of the past two decades. His argument starts with the Roman army's span-of-control model, runs through the Prussian General Staff, and arrives at a provocation: the architecture of the large organization hasn't fundamentally changed in two thousand years. AI is changing it now. And the layoffs are the warning shot, not the main event.

The main event is what comes after. AI is compressing the coordination layer: the middle management that routes information, aligns priorities, and makes decisions across large organizations. When that layer thins or disappears, the corporation gets leaner at executing its core business. But something else disappears with it, and this is the hidden cost: whatever fragile internal capacity existed for building adjacent and transformational businesses internally. That's the innovation crisis and growth gap that AI layoffs are creating. The question is not whether AI will reshape your org chart. It will. The real question is where will transformational growth come from when it does?

The Innovation Gap Gets Worse, Not Better

Middle management has never been popular. But it served a function beyond coordination. It was where innovation programs lived. Where cross-functional projects got shepherded through budget cycles. Where someone with political capital could protect a breakthrough innovation or business idea long enough for it to prove itself.

That protection was already failing. Elliott Parker, CEO of Alloy Partners, named the problem in his book The Illusion of Innovation: corporations engage in activity that feels like innovation but is optimized for safety and predictability. They become "vetocracies," where any manager can kill a threat to the status quo, and the relentless pursuit of capital efficiency crowds out the experimentation that transformational innovation requires. Elliott and I both built our careers at Innosight, the growth and innovation consulting firm co-founded by late Harvard Business School Professor Clay Christensen, where we saw the Innovator's Dilemma play out inside corporation after corporation.

That shared experience inspired the creation of Alloy Partners: a venture builder designed to overcome the innovator's dilemma by co-creating external ventures that combine the speed and agility of startups with the assets, expertise, and capabilities of large corporations. Parker's conclusion, grounded in that work, is clear: the structure of a large corporation is hostile to the uncertainty, speed, and tolerance for failure that new ventures demand.

AI may make the structure leaner. It doesn't make it less hostile. The pressure to demonstrate ROI on AI investments will only intensify the focus on efficiency and predictability, the very forces that suppress transformational innovation. The internal corporate innovation team that already struggled to survive a budget cycle will find even less oxygen in an organization optimizing every function for measurable output.

An obvious question: if middle management was already failing at internal innovation, why is losing it a crisis? Because bad innovation capacity is still capacity. Zero is worse than bad. The crisis isn't that we lost a working system. It's that we lost a failing system without replacing it with anything, and the relentless pressure to optimize will likely ensure nothing replaces it internally.

This is the innovation gap. The same AI that makes the existing business faster makes new business creation harder inside the organization. The Innovator's Dilemma doesn't get resolved by AI. It gets sharper.

Dorsey sees this. His essay separates two corporate responses to AI implementation: give everyone a copilot and cut headcount, which he calls making "the existing structure work slightly better without changing it," or rebuild the company as an intelligence with a capability layer that composes solutions for customers without predetermined roadmaps. Most companies are choosing the first path. Dorsey says it plainly: "If the answer is nothing," referring to what a company uniquely understands, "AI is just a cost optimization story. You cut headcount, improve margins for a few quarters, and eventually get absorbed by something smarter."

The real question isn't how many people you can cut. It's what new value you can create with fewer people, different structures, and new vehicles for growth.

Why the Answer Is External (And What Dorsey's Framework Reveals)

In 1937, Ronald Coase published "The Nature of the Firm" and asked a question that still matters: if markets are so efficient, why do firms exist at all? His answer was transaction costs. When the cost of coordinating an activity inside a firm, through management, hierarchy, and internal processes, is lower than the cost of buying that activity on the open market through contracts and negotiation, the firm grows. When internal coordination costs exceed external ones, the firm shrinks. The boundary of the firm is not fixed. It moves with the relative cost of coordination inside versus outside.

AI shifts that balance. It lowers coordination costs everywhere, but not evenly. A ten-person team using AI tools approaches zero coordination cost. A 10,000-person organization using the same tools still has approval chains, cross-functional alignment meetings, and budget cycles. These are structural features of hierarchy, not speed problems. AI can make each step faster. It cannot eliminate the steps without eliminating the structure that requires them. Coase's logic would suggest that the right size for building new ventures is getting smaller, and the case for building them inside a large corporation is getting weaker.

If the optimal unit is smaller, what does it look like? Dorsey's essay provides a blueprint. He describes two elements that matter most here. First, capabilities: the atomic primitives of the business, hard-to-acquire building blocks like data, regulatory permissions, or distribution. Second, an intelligence layer: the system that composes those capabilities into solutions for specific customers at specific moments, without predetermined roadmaps. He also describes three roles that replace traditional management: deep specialists who build the capabilities, Directly Responsible Individuals (DRIs) who own cross-cutting problems with full authority, and player-coaches who develop people while continuing to build.

This is Dorsey's vision for what an organization looks like after middle management is gone. Small teams. Flat authority. Capabilities composed into solutions. No permanent bureaucracy.

The Model Already Exists

Dorsey designed his framework for internal transformation. But it describes something that already exists externally.

There is a model - the venture studio - that has been systemically building external ventures alongside corporations for years, using exactly the structure Dorsey envisions: small AI-native teams, experienced co-founder entrepreneurs with full authority, specialists who build capabilities, coaches who develop people, and no permanent hierarchy. It works precisely because the businesses it creates are built outside the organization, where the hierarchy and antibodies can't reach them. And it gives corporations a low-risk way to test what Dorsey's post-hierarchy future feels like in practice, one venture at a time, without betting the entire organization to find out.

The innovation gap AI is creating is real and venture studios can be a great testing ground to prepare for the post-hierarchy future of corporations. In Part 2, we will explain what it is and lay out the strategies corporations need to make it work: sharing capabilities as building blocks, co-creating ventures with real ownership at favorable economics, and attracting the entrepreneurial talent that only startups can reach.