The Operating System Nobody's Building

TLDR: A culture initiative won't fix this. 85% of employees are withholding critical information from their managers right now. Not because they don't know it. Because they don't feel safe saying it. Meanwhile, thousands of CEOs are admitting their AI investments aren't producing results. Billions spent. Almost nothing to show for it. Economists have a name for this. Solow's productivity paradox. The tool arrives. The results don't follow. Everyone treats it like a mystery. It isn't. It's a misdiagnosis. The most tech-native generation in the workforce is quietly sabotaging AI adoption. Not because they can't use it. Because their nervous systems have correctly identified it as a threat to their survival. That's not a Gen Z problem. That's a leadership problem. And it's the same problem, at every level, in every organization still wondering why the technology isn't delivering. The data driving your decisions, your strategy, your AI, was produced by people optimizing for self-preservation, not truth. And no platform can fix that. Last week I made the case that you set the temperature of your organization biologically, not mechanically. That the trust account isn't just a relationship metric. It's a biological risk mitigator. If you're still thinking about that, good. Because I'm not done. Every system that processes information, coordinates responses, and adapts to its environment runs on the same basic architecture: inputs, signals, responses, and regulation. The internet. A city's traffic grid. Financial markets. Your immune system. Your organization. Not like a nervous system. Is one. And most of them are chronically dysregulated. The Missing Infrastructure Marc Benioff calls Salesforce "the operating system for the agentic enterprise." As someone who has built her career in the Salesforce ecosystem, I believe in the vision. But it stops just short of being truly transformative. The operating system he's describing runs on data. And the quality, honesty, and completeness of that data depends entirely on the operating system he's not talking about. The one that determines whether your people will give the platform, whatever platform that ends up being, the data it needs. Honestly. Accurately. Without filtering it through self-protection first. You can't buy it from any technology vendor. Fortunately, just like most infrastructure problems, this one is upgradeable. Companies are spending, or about to spend, millions of dollars rebuilding the wrong thing in-house. The operating system most organizations need to rebuild isn't their CRM. It's the human one. Because you can't access innovation from survival mode. The nervous system won't allow it. A team running on threat detection isn't innovating. It's surviving. And survival mode only has one roadmap: don't be the one who gets blamed. Which means your AI inherited the same fear your culture runs on. That's not a technology gap. That's a nervous system gap. No amount of AI spend closes it. The Receipts In February of this year, McKinsey surveyed 10,000 senior executives across 15 countries and found that 3 out of 4 organizations are failing to build high-performance cultures. Less than 1 in 4 achieves lasting impact. (McKinsey State of Organizations, February 2026) That number has held roughly constant across decades of organizational research. The tools change. The consultants change. The failure rate doesn't. Here's what makes it worse. 88% of those same leaders confirmed their organizations are deploying AI. And 86% said they don't feel properly prepared to embed it into daily operations. Everyone bought the tool. Almost nobody built the foundation underneath it. McKinsey's own research found that for every dollar spent on AI model development, organizations should expect to spend three dollars on change management. (McKinsey, "Upgrading Software Business Models to Thrive in the AI Era," 2025) Three to one. Most are spending the inverse. And even McKinsey stops short of saying what change management actually requires to work. They name the investment. They don't name the mechanism. Because buy-in from a nervous system in survival mode isn't buy-in. It's compliance. And here's the part nobody is saying loudly enough. More pressure produces less output. McKinsey found that employees in high-pressured organizations are actually less willing to step up and meet greater demands than those in lower-pressured ones. (McKinsey, State of Organizations, 2026) Push harder, get less. This isn't about whether people can handle pressure. We absolutely can. What we can't run on is pressure plus fear. The same deadline from a leader whose nervous system signals safety lands completely differently than the same deadline from one whose nervous system signals threat. Same ask. Different nervous system experience. Different output. That combination doesn't produce performance. It activates an internal parking brake. You can floor the accelerator all you want. If the nervous system believes survival is on the line, the brake stays on. McKinsey also found that companies giving equal weight to people and performance are more than 4x more likely to sustain top-tier financial results. During the pandemic, those organizations were 59% less likely to show signs of financial distress. (McKinsey, Performance Through People, 2023; McKinsey Organizational Health Index) But McKinsey doesn't ask what developing your people actually requires underneath. You can run the same programs, check the same boxes, and still have a team in survival mode. Because it's not what you do. It's why you do it. The nervous system knows the difference. It always knows. This isn't soft. This is the ROI of the human operating system. Everyone calls it a culture problem. And they're not wrong. But culture describes what you observe. The nervous system is the mechanism that produces it. Saying "that's a culture problem" is like saying "that's just weather" when someone explains atmospheric pressure. You're not wrong. You just haven't explained anything. Culture initiatives fail at the rate they do because they attempt to treat the output without addressing the mechanism. They reach the leader's calendar. Not the leader's nervous system. The Manager I'll Never Forget I had a manager once, more than twelve years ago, who was proving this point before anyone had language for it. It wasn't a one-time grace under pressure. It was just how he operated. Every time something went wrong, he absorbed the news instead of deflecting it back at the person who brought it. He didn't make people pay for telling the truth. He thanked them. He sat with it. Then used it to fix the system that allowed the mistake in the first place. Every time. He's still my favorite manager I've ever had. I've thought about why that story still lives in me and I think it's this: when a leader shows you that honesty is survivable, your nervous system files it as evidence. Evidence compounds (hello, marbles). The team that watched him absorb that news learned something no training program could teach. They learned this room is safe enough for the truth. And he did it without a framework, a workshop, or a culture initiative. He just never made people pay for telling the truth. The Raw Data Amy Edmondson at Harvard has been studying psychological safety for over 25 years. Google spent two years studying 180 of their own teams, conducting over 200 interviews and analyzing 250 team attributes through Project Aristotle, trying to figure out what made their best teams different. The answer, across decades and millions of dollars of research, wasn't talent density or strategy clarity or compensation or tech stack. It was safety. Specifically: the shared belief that you can speak up, take risks, and tell the truth without fear of punishment or humiliation. (Edmondson, Administrative Science Quarterly, 1999) That 85% figure from the opening isn't an outlier. It's Edmondson's research with James Detert, confirmed across industries. (Edmondson & Detert) The withholding doesn't stay contained. It shows up in missed warning signs, failed projects, and the AI integrations that look great on paper and go nowhere in practice. Edmondson named the condition and identified leader behavior as the primary input. What the research doesn't say clearly enough is what psychological safety actually is at a biological level. She named the output. The mechanism underneath it is what comes next. The Mechanism In 1994, neuroscientist Stephen Porges introduced Polyvagal Theory, a framework for understanding autonomic regulation, how the autonomic nervous system continuously organizes your physiological state in response to cues of safety or threat. (Porges, 1994) One of its core principles is neuroception: the nervous system's continuous evaluation of safety and danger, operating largely outside of conscious awareness. You don't decide to feel unsafe in a room. Your nervous system decides first. It reads the body language, the tone, the presence, before a single word lands. Another is co-regulation. Two nervous systems in proximity are not operating independently. They're in a continuous feedback loop. A calm nervous system calms others. A dysregulated one doesn't. I talked about this last week at the individual level. But co-regulation doesn't stop at two people. It scales. First, let's define what a regulated nervous system actually is: your body's ability to return to a calm baseline after stress. The return to equilibrium. When your body learns that a particular environment is consistently safe, it stops spending energy on threat detection and redirects it toward thinking, connecting, and creating. As a leader, you are not in a two-person loop with one employee. You are the single nervous system that every other nervous system in your organization is reading simultaneously. Which means psychological safety isn't a program. It's not a policy. It's not a workshop. It's what happens when every nervous system in your organization feels safe enough to come out of survival mode. And you are the primary signal that makes that possible. The Pattern That Travels Leaders don't just affect the people currently in their rooms. They travel forward in time through everyone who ever worked for them. The regulated leader becomes the standard someone spends their career trying to recreate. Not every leader leaves that kind of mark. Some leave a different one entirely. I've lived both sides of this. Two different companies. Two different cities. Two leaders who had no way of knowing each other. Running the exact same playbook. The tactics were identical, which tells you everything you need to know about how this kind of leadership travels. Both of them would do the same thing. They'd bring up a mistake we'd already discussed and I'd already corrected, weeks later, in front of others, at exactly the moment it would land hardest. Not to fix anything. To remind me where I stood. They'd also hand me goals that sounded clear until you tried to execute them. When I asked for clarity they were either unavailable or turned it back on me, as if my inability to read their mind was evidence of my own deficiency. The ambiguity was never accidental. Unclear goals keep people dependent. And they give a leader with no real answer somewhere to hide. If you never defined success clearly, you can never be wrong about it. And when things go sideways, everyone absorbs the blame but the person who never defined the target. The mistake gets corrected. The trust doesn't. Once you've learned that your mistakes are currency in someone else's pocket, you stop being honest about them. You start managing the optics instead of solving the problem. You get very good at looking fine. And while they were holding my history over me in the rooms I was in, they were rewriting the story in the rooms I wasn't. My wins became theirs. The exposure became mine. That's not just a bad leader. That's a dysregulated nervous system that learned, somewhere along the way, that control was the only available form of safety. And then replicated it onto everyone underneath them. The fear-based leader becomes either the cautionary tale someone swears they'll never become, or the unconscious blueprint someone replicates without ever questioning it, because that's what leadership looked like when they were learning what leadership was. That second one is the most dangerous. Not the leader who consciously chooses fear. The leader who genuinely believes fear-based leadership is just what leadership is, because nobody ever showed them anything different. That's not a character flaw. That's a transmission problem. Fear-based leadership replicates itself through the people it produces. Down the org chart. Across careers. Into every organization those people ever touch. Until someone breaks the pattern. Which is why one regulated leader matters more than most organizations realize. Not because of what they produce directly. Because of what they make possible in every person who ever works for them. The weak link is never the human. It's always what the human was never given. I will die on this hill. The Upgrade A culture initiative won't fix this. It never has. What your organization needs is for its leaders to become regulated nervous systems that other nervous systems can synchronize with. One regulated leader breaks the pattern. A leadership team of them transforms the organization. Your nervous system doesn't clock in and clock out. Neither does its impact on the people around you. Now you know what the operating system is. Next, we talk about how to build it. Sources: McKinsey State of Organizations, February 2026; McKinsey, "Upgrading Software Business Models to Thrive in the AI Era," 2025; McKinsey, "Performance Through People," 2023; McKinsey Organizational Health Index; Amy Edmondson, "Psychological Safety and Learning Behavior in Work Teams," Administrative Science Quarterly, 1999; Edmondson & Detert, research on employee voice; Google, Project Aristotle; Stephen Porges, Polyvagal Theory, 1994; Fortune / NBER, AI productivity paradox research, 2026; Marc Benioff, Salesforce Q3 FY2026 earnings call, February 2026.