Human Capital Intel - 8/13/25
The value of culture | 'Wild West' AI usage requires org rethink | 1-1s under fire | CEO notes create controversy
Welcome to the latest edition of Human Capital Intelligence, your weekly brief synthesizing over 250 leadership, HR, and people sources to filter out the noise. As always, we would love to hear from you at ken@reyvism.com with questions you’d like answered or topics covered.
Sent this by a friend? Sign up here to receive HCI in your inbox every week.
By Ken Stibler; Powered by Reyvism Analytics
Deep Dive: Putting a price tag on the right culture
In the middle of a frenzied AI talent war, Anthropic CEO Dario Amodei is holding his ground. While Meta has been offering $100 million signing packages to lure away star engineers from rivals, Amodei says many employees have rejected those offers, with some refusing even to take a meeting with Mark Zuckerberg. Rather than counter with higher pay, he has stuck to Anthropic’s strict, level-based compensation system, arguing that bending the rules for a few would undermine fairness, trust and the mission that holds the company together.
Amodei calls the decision a “unifying moment,” proof that people stay because they believe in the work, not just the money. Anthropic’s retention rates back him up, with 80% of hires over the past two years still on board, higher than OpenAI, Google DeepMind and Meta.
He and other tech leaders, including OpenAI’s Sam Altman, argue that Meta’s approach not only disrupts culture but also risks recruiting people who may not be the most mission-aligned. “They are trying to buy something that cannot be bought,” Amodei says.
The pattern is visible elsewhere. Microsoft’s AI head, Mustafa Suleyman, has successfully poached senior talent from DeepMind not by matching Meta’s biggest checks, but by promising smaller, faster-moving teams and less bureaucracy. Recruits say the appeal was a “low ego, high ambition” environment where they could have more impact, not the largest possible salary.
The lesson: money opens doors, but it the intangible culture that keeps the right people inside. Fair systems, a mission worth committing to, and an environment that lets top talent do their best work can compete with even the richest offers.
Leaving AI’s wild west era requires new organizational infrastructure
AI is spreading through companies from the bottom up, and old organizational models are struggling to keep pace. Employees are adopting tools on their own, often paying out of pocket, while most organizations have not invested in systematic training or governance. Many executives are still learning the basics (91% admit they have pretended to know more about AI at work than they actually do), leaving organizations chasing from behind and adopting via committee,
The bottleneck is organizational. On traditional org charts, responsibility for AI is scattered across IT, HR, and line-of-business leaders, with no single owner to connect the dots between risk, capability, and competitive advantage. This gap creates tension as executives push for outcomes without understanding trade-offs, while managers inherit pressure to deliver without the tools or guidance to do it well. Without a clear center of gravity, AI adoption becomes inefficient, exhausting, and failure-prone.
The fix is a defined AI leadership role. In some companies that will mean a CEO personally driving AI as a primary priority. In others, it could be an internal AI lead or a fractional executive who owns the mandate to educate leadership, guide employees, select high-impact use cases, choose tools wisely, and set guardrails that minimize risk.
Quote of the Week: Shooting the messenger
“When you start firing people for giving you bad news, then people are hesitant to give you bad news, and then bad things happen. We’ve got to work hard to get ground truth.”
—Rep Don Bacon (R-Neb.) on President Trump’s firing the commissioner of the Bureau of Labor Statistics following disappointing jobs data
Reading List:
The 1-to-1 is under fire
The once-sacred weekly seeing its managerial relevance questioned, Inc reports. Rooted in early 20th-century management theory and later popularized by tech leaders, the 1:1 evolved from command-and-control briefings to coaching and development sessions. Yet today’s faster, flatter, AI-enhanced workplaces make standing weekly check-ins feel like productivity tax rather than leadership. Many are unfocused, manager-dominated, or confused with therapy sessions. The emerging alternative is a more intentional mix of micro-interactions, data-driven insights, and targeted human moments, less ritual, more relevance.
A note about CEO notes
Every few months, a CEO memo goes viral for the wrong reasons. AT&T’s John Stankey is this week’s unfortunate winner, with a 2,500-word response to an employee survey that was both unusually candid and tone-deaf. John “ended” the old contract of loyalty for job security and spelled out exactly what employees should and should not expect, yet offered little to inspire commitment in return, framing the shift as non-negotiable and treating basics like a functional office as rewards. That is the tightrope with CEO notes: too soft and they read as PR, too hard and they fuel disengagement. The best strike a balance, pairing hard truths with meaningful commitments and writing as if every employee (and the entire internet) will read it.
Efficiency lessons from the Chinese Communist Party
Beijing has launched a nationwide push to cut meetings, shorten reports, and cap speeches at one hour in a 4,845-character directive that narrowly met its own new 5,000-character limit, the Financial Times reports. The stated goal is to curb “formalism and bureaucratism,” freeing officials to focus on real economic priorities. For HCI leaders, the parallel is hard to miss: if the communist party thinks its bureaucracy needs fewer, shorter, and sharper communications, there is probably room for most organizations to take a similarly unsentimental look at their own meeting culture and reporting habits.
Data Point: AI-y! Stop that.
57%
Number of employees who admitted to entering sensitive information in publicly available GenAI tools which 68% admit to using on company devices.
In Other News
By 2028, 1 in 4 candidate profiles will be fake, Gartner predicts. (HR Dive)
Why So Many Women Are Quitting the Workforce. (Time)
Why struggling companies are loading up on bitcoin. (Financial Times)
Trump Reminds CEOs Who the Ultimate Boss Is: President isn’t holding back in telling corporate bosses how to run their companies—or when to resign. (Wall Street Journal)
US companies move to tighten security after fatal shooting in New York. (Financial Times)
SEO Is Dead. Say Hello to GEO. Search-engine optimization now feels dated. Generative-engine optimization is all about trying to trick AI chatbots. (NY Mag)
Many employees with side hustles see them as insurance policies, Glassdoor says. (HR Dive)
Why HR should ditch employee engagement strategies for wellbeing. (HR Brew)
No, jobs are not just bundles of existing tasks to automate with AI. (AI and Workforce Predictions)