Summary: We worry that a future superintelligence will optimize for the few at everyone else’s expense. We built that system already, without AI: the modern corporation running on shareholder-primacy doctrine since the 1970s. That doctrine was never truly a legal requirement — it was an ideology that took hold once executives were paid in stock. This piece traces how we got here, why the gap between what companies say and how they actually decide breaks employee trust no EVP can repair, and why fixing it is structural work, not a messaging campaign.

Would a Superintelligence Prioritize the Self-Interest of the Few Over the Needs of the Rest?

We built that version already, without AI.

It's not in a data center or trained on the internet. It was built in boardrooms and business schools over the better part of a century. The modern corporation is a goal-directed, self-optimizing system with no capacity for empathy and a single programmed objective. We have spent decades debating what values to align a hypothetical future artificial intelligence to, while the one we already built has been running its own alignment experiment on the global workforce since about 1976.

This is not a scalding of capitalism. It is a think map to make it better.

The doctrine most executives believe they are required to operate under is to maximize shareholder value, at any cost. It is largely fiction. Call it a religion. The case most often cited is *Dodge v. Ford*, 1919. The Michigan Supreme Court ruled against Henry Ford after he tried to withhold dividends from shareholders in order to benefit workers and lower car prices. That narrow ruling between feuding shareholders in a closely held company somehow became the foundational proof that executives must always put shareholders first. Corporate law scholars have spent decades pointing out the leap. The case has been widely cited as supporting shareholder primacy, though scholars disagree about how far it actually goes. It was a dispute resolution that got drafted into an ideology.

Most corporate law, Delaware law especially, which governs the majority of U.S. corporations, gives executives enormous discretion under the Business Judgment Rule. Boards can consider employees, communities, and long-term organizational health. The legal permission to do the right thing has largely always existed.

So where did the shareholder 'first and only' doctrine come from?

In 1970, Milton Friedman wrote an essay in the *New York Times* declaring that the social responsibility of business is to increase its profits. Six years later, economists Jensen and Meckling published an agency theory paper arguing that executives were self-interested agents who needed to be disciplined, aligned with shareholders through compensation. The solution they proposed: pay executives in stock.

The doctrine didn't create the incentive. The incentive created demand for the doctrine. Once executive wealth became directly tied to share price, shareholder primacy stopped being a theory and became a survival instinct. It was not legally required. And of course, it was personally lucrative. Nobody had to be cynical about it. They just had to be paid in stock and on the surface seem human.

A system left to optimize without interference has one logical destination. The Monopoly inventor figured that out and made it a board game.

Manage to the quarter. Cut what can't be measured. Treat talent as a cost line rather than a capability investment. And then, because the organization still needs people to show up and care, say something different externally. We are a people-first organization. People are our greatest asset. Our diversity makes us stronger. Our culture is our competitive advantage. Meanwhile: headcount reductions, benefit cuts, stalled development budgets, leaders promoted for hitting numbers rather than building teams.

Employees are not confused by the gap between those two realities. They are clarified by it. They see exactly what the organization values, regardless of what it says. And once that trust breaks, no EVP refresh fixes it. No employer brand campaign reaches it. You cannot advertise your way out of a structural contradiction.

The Business Roundtable, an association of more than 200 CEOs, seemed to grasp this when they issued their 2019 statement redefining the purpose of a corporation to include all stakeholders, employees, communities, customers, not just shareholders. That was before the pandemic disruptions, too. It was framed as a new vision. What it actually was: an acknowledgment that the prior fifty years had been a choice, not a legal requirement. A quiet admission that the doctrine was shaped by legal interpretation, managerial incentives, and business ideology more than by law. The damage doesn't undo itself with a statement.

AI researchers are now urgently asking about systems that don't yet exist: what happens when you build something that optimizes relentlessly for a narrow objective with no regard for broader harm? What happens when the goal and the damage are structurally invisible to each other?

We know the answer. We have been living inside it.

WorkingTheory was built around that gap, addressing the structural problem underneath. The distance between what organizations claim to be and what they actually are, accumulated over decades of optimizing for metrics that were never designed to measure the health of a culture. They were designed to measure the health of an executive compensation plan.

Most companies diagnose this as a communications challenge, hire accordingly, and wonder why nothing changes. Alignment isn't fixed by better messaging. It's fixed by closing the gap between what the organization says it values and how it actually makes decisions, about people, about investment, about what gets rewarded and what gets cut. That work is harder and slower than a campaign, and it's the only thing that can help organizations and their people thrive through these tumultuous times.

The question was never really about artificial intelligence, or even shareholder primacy.

It was always about what we choose to optimize for, and who bears the cost when we get it wrong. Capitalism mostly works. It needs a little help from time to time.

We worry that superintelligence will optimize value for a small few at everyone else’s expense. But we’ve been doing something similar for decades through the doctrine of maximizing shareholder value above all else. The costs have been enormous, and it doesn’t have to stay this way.