The People Building AI Just Asked for More Time
For years, the dominant assumption around AI has been that faster is better: faster models, faster adoption, flatter organizations, fewer layers between technology and execution.
Three stories this week complicate that assumption.
Anthropic CEO Dario Amodei called on the AI industry to deliberately pace frontier development—and two of his biggest rivals publicly agreed. Gallup found that workers closest to AI are actually the most anxious about what it means for their jobs, unless strong management changes the equation. And at a gathering of CEOs in Washington, one of the strongest celebrations of leadership wasn’t about technical fluency at all. It was about a CEO’s ability to inspire loyalty, excellence, and respect.
The pattern is worth paying attention to. As AI becomes more capable, the human systems surrounding it are not becoming less important. They are becoming the constraint.
The next phase of AI leadership may therefore be less about removing friction everywhere and more about knowing which friction—oversight, management, judgment, trust—is actually doing useful work.
🛑 AI’s Biggest Rivals Just Agreed on Something They Rarely Agree On: Dario Amodei called for pacing frontier AI development. Sam Altman and Elon Musk publicly backed the direction.
🧠 AI Doesn’t Make Workers Anxious About Their Jobs. Bad Management Does: Frequent AI users report substantially more displacement anxiety—but strong management sharply changes the outcome.
👥 CEOs Are Still Betting on the Leadership Trait AI Can’t Replace: At the Yale CEO Caucus, the leadership quality drawing attention wasn’t technical prowess. It was the ability to inspire people.
Read time: 6 minutes 👇
Enrollment Open AI Leadership Program for Executives: Inside Silicon Valley — Spring Cohort
April 5–9, 2027 | Silicon Valley, CA — view the program and apply
Some leadership questions are difficult to answer from a distance.
What is AI actually changing inside leading companies? Where are executives moving faster than the headlines suggest? And which ideas are worth bringing back into your own organization?
Those are the questions behind SVEA’s AI Leadership Program for Executives: Inside Silicon Valley.
Registration for our January 2027 cohort is now closed, and applications are now open for the Spring Cohort, April 5–9, 2027.
This isn’t five days of sitting through presentations about AI.
It’s five days inside the Silicon Valley ecosystem—meeting the people building, deploying, and making decisions around AI, with exposure to companies such as Tesla, NVIDIA, Airbnb, and Amazon, alongside a carefully selected group of senior executives, founders, and board members from around the world.
During the week, participants will:
Engage directly with AI builders, operators, and leadership teams across established technology companies and AI-native startups
Use proprietary frameworks to turn what they see and hear into an AI leadership playbook relevant to their organization
Compare approaches and challenges with a global peer group facing many of the same strategic questions
Build and refine an AI transformation playbook throughout the five-day experience
The goal isn’t to leave Silicon Valley knowing more about AI.
It’s to leave knowing what deserves your attention—and what you’re going to do about it.
The Spring Cohort is limited to 20 participants.
👉 Book a discovery call and explore the full program to join us April 5–9, 2027.
AI’s Biggest Rivals Just Agreed on Something They Rarely Agree On
Rival AI labs agreeing publicly is unusual. Agreeing that the industry may need to slow the pace of capability development is something else entirely.
Anthropic CEO Dario Amodei published an essay this month titled We Must Pace the Frontier, arguing that AI capabilities are advancing quickly enough that safety, alignment, and independent evaluation need more time to catch up. His proposal does not call for stopping model development. It calls for deliberately pacing capability gains while strengthening the systems designed to evaluate and control increasingly powerful models.
The first commitment is concrete: Anthropic says it will give independent third-party evaluators ongoing, employee-like access to its systems so they can verify whether safety commitments are actually being followed.
Then came the unusual part.
Elon Musk publicly backed Amodei’s argument. OpenAI CEO Sam Altman also agreed with the need to pace the frontier and said OpenAI would match Anthropic’s commitment to independent evaluators.
Agreement, however, is not coordination. There is no industry-wide enforcement mechanism, and the broader debate remains divided over whether slowing development reduces risk or instead creates competitive and geopolitical disadvantages.
Why it matters: The important signal isn’t that frontier AI development is about to slow. There is no guarantee that it will. The signal is that some of the people with the clearest view of the capability curve are publicly arguing that the systems surrounding AI may not be keeping pace with the technology itself. For executives building multi-year AI roadmaps, capability growth is no longer the only variable worth modeling. Governance, evaluation, regulation, and the possibility of deliberate pacing belong in the scenario plan too.
AI Doesn’t Make Workers Less Anxious. Good Managers Do
Most organizations have treated AI anxiety as an AI problem.
The response has therefore been predictable: more training, more communication about the technology, and more reassurance that employees should learn to work alongside it.
Gallup’s latest workforce research suggests the more important lever may be somewhere else entirely.
Across four waves of nationally representative U.S. workforce data collected between 2023 and early 2026, employees who use AI daily or several times a week were more than twice as likely to fear their jobs would be eliminated within five years as less frequent users. More exposure to the technology did not reassure them.
Management did.
Employees who felt respected and supported by their organizations reported substantially lower displacement concerns. Gallup found that the relationship between frequent AI use and job-loss anxiety was 6.8 percentage points smaller among workers reporting high respect and 11.1 points smaller among those who felt their organization cared about their wellbeing.
The consequences extend beyond anxiety. When the same workers became more concerned about AI displacement over time, Gallup observed lower engagement and job satisfaction, higher burnout, and greater likelihood of looking for another job.
The technology hadn’t changed between those employees.
Their experience of the organization had.
Why it matters: If your AI rollout invests heavily in tools and training but leaves frontline management untouched, part of the implementation infrastructure is missing. Managers are becoming the translation layer between what AI can technically do and what employees believe that capability means for their future. Treat manager quality as part of your AI strategy, not as a separate HR initiative.
CEOs Are Still Betting on the Leadership Trait AI Can’t Replace
At a moment when flattening organizations and removing management layers have become prominent themes in corporate restructuring, a gathering of CEOs in Washington offered a useful counterpoint.
At September’s Yale CEO Caucus, Corning CEO Wendell Weeks received the Yale Legend in Leadership Award. Fortune executive editorial director Diane Brady noted that the praise surrounding Weeks focused less on technical prowess than on his ability to inspire “excellence, loyalty, and respect” through his leadership of the 175-year-old company.
That distinction matters because it lands in the middle of a much larger management debate.
AI can absorb coordination work. Agents can automate reporting, information gathering, scheduling, analysis, and parts of decision support. Some management roles built primarily around moving information from one layer of an organization to another may therefore become harder to justify.
But that is different from saying management itself has lost value.
The ability to create clarity when people are uncertain, earn trust through change, develop talent, challenge a team without losing it, and get people to commit to something difficult remains a different category of capability.
AI may change how many managers an organization needs.
It doesn’t automatically change what a great one is worth.
Why it matters: Flattening an organization and improving it are not necessarily the same decision. Leaders redesigning management structures should distinguish between layers that exist because information once moved slowly and managers who materially improve the performance of the people around them. AI gives companies an opportunity to remove unnecessary management overhead. Used indiscriminately, it can also remove some of the human capability they will need most during the transition.
Three stories, one increasingly important tension: the more capable AI becomes, the more consequential the human systems around it become.
The frontier labs are debating whether technological capability is advancing faster than the mechanisms designed to govern it. Gallup’s data suggests employees experiencing AI most directly are looking to managers to make sense of what it means for them. And executives are still recognizing leaders for something no model benchmark captures particularly well: the ability to make other people want to do exceptional work.
That does not make this an argument against speed, automation, or flatter organizations.
It makes it an argument for being much more precise about what we automate away.
The next competitive advantage may not come from adopting AI faster than everyone else. It may come from building the leadership, judgment, trust, and organizational discipline required to absorb that speed without losing control of what makes the organization work.
At SVEA, that is the leadership challenge we believe matters now: not simply understanding what AI can do next, but building the judgment to decide what your organization should do with it.