Culture At Scale: Lessons From 10 Tech CEOs
Reading Time: 8 Minutes
Image created by Superhumxn team.
Differentiation
When AI becomes equally accessible to all, and technology is no longer an advantage, what’s left? Ravi Kumar S. and Rohan Murty recently wrote an article for the World Economic Forum arguing that a company’s collective ethos is the new competitive frontier. Culture is the primary competitive differentiator, and it’s something that can’t be easily copied.
It’s no secret that company culture matters in modern business. Organisations with high-trust cultures make decisions 9.3 times faster than those without, because people closest to a problem can act on it directly instead of sending it up a chain. McKinsey found that agile organisations, ones with clear values and fast decision-making, have a 70% higher chance of ranking in the top quartile for both organisational health and market performance. Despite the data, culture rarely receives sufficient airtime from boards and investors.
These 10 technology CEOs understand that scaling a company without aligned communication, values and ways of working is a losing game. Nvidia, OpenAI, Canva, Databricks, and the others on this list place company culture at the forefront of their strategy. Here’s how they think about culture and what we can learn from them.
What can I expect from a subscription?
Real-world strategies for navigating the future of work.
AI AGENTS AT WORK
Everything you need to know about how AI agents are changing teams and ways of working.
SUPERHUMXN IRL
Real insights from investors and operators.
MEMOS ON PEOPLE, CULTURE & AI AT WORK
The latest use cases and strategies for change and transformation.
1. Jensen Huang - CEO, Nvidia
"Innovation requires experimentation. Experimentation requires exploration. Exploration results in failure. Unless you have a tolerance for failure, you will never experiment." —Jensen Huang
Image Source: Marketwatch
In 1996, Nvidia almost went bankrupt. Its first chip, the NV1, had bet on a technical approach the rest of the industry didn't adopt, and the company burned through most of its cash before finding a buyer for its remaining stock. Huang has talked about that period as formative, the moment that taught Nvidia to test different perspectives rather than commit rigidly to one bet.
He's since spoken about the importance of experimentation as a cornerstone of Nvidia’s company culture: "Unless you have a tolerance for failure, you will never experiment. If you don't experiment, you won't innovate. If you don't innovate, you don't succeed."
When Huang talks about failure, he isn’t encouraging recklessness, but rather an environment where failure can be tolerated short-term so that winning in the long-term is inevitable. Nvidia's near-death in 1996 wasn't fatal because the team adjusted fast enough. That's the tolerance he's talking about, not an absence of consequences, but the ability to recover and try again before the mistake becomes terminal.
2. Mira Murati - CTO, OpenAI
"It's tough to build these technologies in a vacuum, without contact with the real world."—Mira Murati
Image Source: Mira Murati LinkedIn
In February 2025, Mira Murati made a bet when she launched her latest start-up venture. Thinking Machines Labs predicts AI that reasons and collaborates better with humans will outperform LLMs that simply focus on maximising outputs.
Mira brought in OpenAI co-founder John Schulman as Chief Scientist and recruited around 30 researchers from OpenAI, Meta, Mistral, and DeepMind, including co-founder and CTO Barret Zoph and co-founder Luke Metz. By July 2025, the company had raised a record-breaking $2 billion seed round at a $12 billion valuation.
Soon after Murati fired Zoph for what was described as sharing confidential company information with rival firms. Within hours, Zoph and Metz had both rejoined OpenAI, along with another former colleague, in an announcement OpenAI made before Murati had even confirmed the full extent of the departures. It was the second co-founder exit in months, following Andrew Tulloch's move to Meta in October.
The lesson here is more of a warning that shared values and vision are not always enough for teams to align. The ability to work through disagreements and discomfort can truly make or break a company’s success.
3. Alexandr Wang - CEO, Scale AI
"Over time interviewing, I've found that I mainly screen for one key thing: giving a sh*t. To be more specific, there are two things to screen for: (1) they give a sh*t about Scale, and (2) they give a sh*t about their work in general."—Alexandr Wang
Image Source: Entrepreneur
In 2020, Alexandr Wang wrote a memo to the Scale AI team describing what he’d learned about hiring. He'd ask candidates what the hardest thing they'd ever worked on was, and listen for whether the answer came with real intensity or just sounded like the right thing to say.
This wasn't just a one-off memo. On Y Combinator's podcast, five years and hundreds of hires later, Wang said he still personally approved or rejected every single hire at the company. "I care a lot about every decision we make at the company," he said, adding that he could always tell when someone was "phoning it in" versus when the work meant something to them.
In June 2025, Wang left Scale to lead Meta's new Superintelligence Lab as part of a $14.3 billion deal that doubled Scale's valuation. The standard he'd applied personally for nine years, every hire, reviewed by him, now sits with Jason Droege as interim CEO. Cultural standards can and should outlive the leaders who set them.
4. Demis Hassabis - CEO, Google DeepMind
"Our mission is to solve intelligence, and then use that to solve everything else."—Demis Hassabis
Image Source: UCL
Demis’ mission hasn't changed even as the company has gone from 200 researchers playing Atari games to a Nobel Prize and AlphaFold.
In a 2016 talk at MIT, Hassabis described the specific culture he was trying to create: "the best from Silicon Valley startup culture with the best from academia", combining the open-ended, blue-sky thinking you'd find in a university research lab with the focus, energy, and pace of a top startup. Most companies pick one of these. Academic environments tend to reward depth and patience; startups reward speed and shipping. Hassabis built DeepMind explicitly to leverage both at once.
The lesson is that a long-term mission only means something if the day-to-day culture truly supports the kind of work that mission requires. "Solve intelligence" is a 50-year problem, and a culture optimised purely for quarterly output would have made that mission impossible from the start.
5. Ali Ghodsi, CEO, Databricks
"A company's culture is more or less the personality of its CEO…formalise that, so make culture principles out of that, then you're giving people the cheat sheet of how they can succeed in your company."
—Ali Ghodsi, CEO, Databricks
Image Source: Business Insider
Ali Ghodsi has made what he calls a controversial argument about culture. He claims that a company's culture is more or less the personality of its CEO, and his advice to founders follows suit. Don't write down some aspirational version of the culture you'd like to have. Write down an honest assessment of who the CEO actually is, formalise that into cultural principles, and then use those principles as the basis for who gets hired, promoted, and so on. He calls this the "cheat sheet", which provides a guidebook of how someone can succeed at the company.
It’s tempting to create a set of aspirational company values, but Ghodsi's version takes full accountability from day one.
6. Sarah Guo - Founder, Conviction
"Pace flows from the top. When leaders demonstrate comfort with imperfect information and rapid decision-making, it cascades through the organisation"—Sarah Guo
Image Source: Sarah Guo LinkedIn
"Conviction filters noise. Without it, speed makes for thrash. In environments flooded with information and leverage, beliefs need to be held steady long enough to be tested. Conviction is the temporary suspension of doubt that makes action possible. Sometimes that suspension of doubt needs to last years."
Sarah Guo spent nine years as a general partner at Greylock before leaving in 2022 to found Conviction, an early-stage venture firm built specifically around AI investing. She's been an early investor in Figma, Harvey, Mistral, and Cognition, among others.
Her leadership approach involves faith, holding a belief steady long enough for evidence to actually accumulate, rather than abandoning an approach after the first hurdle. Guo extends the same logic to people: "even traits we treat as fixed, like work ethic, ambition, and risk tolerance, respond strongly to incentives, expectations, and social context... agency is a product of system design. People act when they have ownership, consequences, and upside."
7. Melanie Perkins - CEO, Canva
"We wanted everyone at Canva to be the master of their own destiny and understand that this is the place where everyone can shape our company and culture."—Melanie Perkins
Canva runs on something called "close the loop," which means every complaint or feature request a user submits gets filtered into product development, and each issue must have a resolution.
Amy Schultz, Canva's Global Head of Talent Acquisition, talks about their purpose: "We have a simple two-step plan: to be one of the most valuable companies in the world but also to do the most good we can do."
The team at Canva also reinforce their mission through what they say no to. Perkins deliberately avoided taking on large amounts of early capital, which gave the team room to iterate slowly and stay close to customers, rather than taking on aggressive growth targets before the product was ready. At one point she paused expansion entirely to rebuild their product, choosing long-term scalability over short-term momentum.
8. Dylan Field - CEO, Figma
"Our internal culture at Figma encourages everyone in the company to help make Figma better."—Dylan Field
Image Source: Dylan Field LinkedIn
In a 2025 conversation on Figma's own blog, Dylan Field said: "Our internal culture at Figma encourages everyone in the company to help make Figma better, with initiatives like Maker Week. Many of our best features have come from these efforts."
An interesting example of Figma's innovative culture is the success of their multiplayer editing feature, which wasn't something users were asking for. Field and co-founder Evan Wallace built it based on a hunch from their earlier experiments with Google Wave, not customer feedback. Had they only built what users explicitly requested, multiplayer editing might never have happened.
The fine balance between listening to feedback and trusting internal conviction is what Maker Week formalises. It gives people across the company permission to build the thing they have a hunch about, the way Field and Wallace did with multiplayer.
9. Arthur Mensch - CEO, Mistral AI
"We've shown that with a small team focused on creating the best text generation models, we can develop world-class technology."— Arthur Mensch
Mistral has grown to a major AI lab with 350 employees since 2023, competing with companies many times its size and budget. Asked how, Mensch told Fortune: "We have created our own culture, which is low-ego and scrappy." He pointed to the best practices the founding team brought with them from DeepMind and Meta, shaping a mindset focused on shipping fast and maintaining rigorous scientific standards.
The founding team was purposely small, Mensch, Timothée Lacroix, and Guillaume Lample, kept that way intentionally to maintain speed, focus, and a high concentration of talent rather than growing headcount as a proxy for progress.
Mistral's commitment to open-weight models isn't separate from this. Mensch has said the goal was to bring openness and information-sharing back to a field that was losing it: "Since 2022, this has been starting to disappear, so we wanted to bring it back, and now several large companies have followed our path."
10. Reid Hoffman - Founder, LinkedIn & Partner, Greylock
"Pay attention to your culture and your hires from the very beginning."—Reid Hoffman
Hoffman had learned this lesson the hard way. At his first company, SocialNet, founded in 1997, Hoffman designed an org chart before he'd built anything else, specifying exactly what experience level each role needed. The company failed. Looking back, he said the structure he'd designed for the company's first version "would immediately fall apart as soon as something fundamentally changes or the company decides to pivot."
By PayPal, and later LinkedIn, his approach had inverted completely. Instead of hiring people who precisely fit a role that existed today, he hired generalists, people who could do the job in front of them but adapt when the company inevitably became something different. The "pay attention from the very beginning" advice is a direct correction of his own first mistake: the earliest hires matter most as they lay the groundwork for future success.
Cheat Sheet
What stands out across all ten isn't what these companies say about culture; it's how they enforce it.
Ghodsi's "cheat sheet" decides who gets hired, promoted, and let go. Canva's "close the loop" means every piece of user feedback has to reach a resolution, no exceptions. Mensch caps teams at five people. Field built an entire week into the calendar so people can act on a hunch the way he and Wallace did with multiplayer editing. Wang reviewed every hire himself, for nine years. Company beliefs are not lofty statements, but are fully embedded into day-to-day work.
