‘No One Works Here’ urges firms to replace human bottlenecks with AI efficiency

Knowledge workers in 2026 are grappling with what MIT Sloan School of Management senior lecturer Paul Cheek calls a “split-screen reality.” On one side are the “AI-abled elite” who can move quickly, free from the administrative burdens of legacy organizations. 

On the other side, legions of competent, diligent professionals are still operating under the burden of a 20th century decision-making hierarchy while their career paths become less certain with each step. 

In the following excerpt from his new book, “No One Works Here: How AI-Driven Enterprises Are Raising the Bar for Business, Leadership, and Competition,” Cheek makes a case for intentionally taking humans out of the loop when machines can make better, faster decisions.

The human bottleneck

With a title like “No One Works Here,” my book could have been the story of a company with a bunch of employees who are all slackers, a company where no one works: the least efficient company. My book is the opposite. It’s a story about the most efficient companies — those where humans don’t “work” in the traditional sense.

The companies that are most likely to fall behind their peers are the ones that insist on keeping a human in the loop for decisions that a machine could have made yesterday. That doesn’t mean all decisions, just those that might not require a human in the loop in the era of the AI-driven enterprise.

It reminds me of the central tension in the 1983 fictional film “WarGames,” which showcases the extreme example of what happens when you propose removing the human from the loop. There is a pivotal scene where military brass debate the implementation of the WOPR (War Operation Plan Response) supercomputer. 

The argument for the machine is cold and logical: Human operators in the missile silos were a liability because, faced with the gravity of their actions, they hesitated. They couldn’t “turn the key” to launch the nuclear missiles. The proposal is, instead, to route the missile launch codes directly to the WOPR, effectively taking the human out of the loop to ensure instant execution. In the context of nuclear warfare, that logic is terrifying: It could lead to accidental Armageddon.

However, we no longer have to look to fiction to see the benefit of removing the human from the loop. We see it every day on our highways. And it saves lives: A modern motor vehicle equipped with an intelligent emergency braking system [reacts] before a human driver even perceives a hazard or has the time to react. By the time the human brain processes the visual cue of a car stopping two vehicles ahead, a car often obscured from the driver’s line of sight, the AI has already calculated the physics of the impending collision and initiated a stop. In this moment, the human’s biological latency is a liability; the machine’s autonomous agency is the lifesaver.

Stepping back from the WOPR and the emergency braking system examples, in the context of the global economy, the logic is inescapable. In most industries, we are not debating nuclear launch codes; we are debating pricing adjustments, supply chain routing, and customer support queries. In these domains, the “hesitation” of the human operator — the need to schedule a meeting or build consensus — is not a safety feature; it is a bottleneck. It could be an existential threat to the organization.

Entrepreneurship Development Program

In person at MIT Sloan

Register Now

The reliability tax

However, we must acknowledge the “reliability tax.” When you first deploy an agent, latency often spikes. You will spend more time auditing the agent’s output than you would have spent doing the work yourself. This is the latency J-curve: Things get slower and riskier before they become instant. The winners are the organizations that push through this “messy middle” of validation to reach the other side: trust-based autonomy.

Moving forward, organizations must prioritize human bandwidth for those situations in which the unique capabilities we possess are a necessity and the consequences of delegating to AI, even with guardrails, could be detrimental.

Structural debt versus technical debt

As software engineers, we talk a lot about “technical debt.” But the companies on the Fortune 500 list, like most established businesses, are suffering from structural debt.

In the last century, we built organizations to solve for the limitations of human communication. We created hierarchies because one person couldn’t manage a thousand people. We created departments to specialize knowledge because one brain couldn’t know everything. We created middle management to route information because we didn’t have networks.

This structure creates latency. When a legacy company wants to make a decision, the data has to travel up the chain of command. It gets aggregated, summarized, PowerPointed, and debated. By the time the decision is made and the communication comes back down, the market has moved. 

The fear of a market shifting before a product can launch used to be a uniquely deep-tech nightmare. That 10-year deep-tech threat has now come for everyone. Today, a digital product can have its core value proposition commoditized mere hours after launch by a single Big Tech release.

Excerpted with permission from the publisher, Wiley, from “No One Works Here: How AI-Driven Enterprises Are Raising the Bar for Business, Leadership, and Competition,” by Paul Cheek. Copyright © 2026 by John Wiley & Sons, Inc. All rights reserved.

Paul Cheek is a senior lecturer at the MIT Sloan School of Management, senior adviser for entrepreneurship and artificial intelligence at the Martin Trust Center for MIT Entrepreneurship, co-founder and executive director of the AI-Driven Enterprise Institute, and founder of Entonomy. In addition to “No One Works Here,” he is the author of “Disciplined Entrepreneurship: Startup Tactics” and the recipient of MIT’s Monosson Prize for his impact on entrepreneurship education.