How to lead in volatile times: Key insights from 8th annual Berkeley Culture Connect Conference

Over 200 business leaders and academic researchers gathered at UC Berkeley Haas this month for the sold-out eighth annual Culture Connect Conference, tackling one of the most urgent questions facing organizations today: How do leaders guide their teams through relentless change—and the rise of AI—while maintaining their culture, mission, and humanity?

The conference, organized by the Berkeley Center for Workplace Culture and Innovation, brought together top executives from dozens of companies, including OpenAI, GM, Microsoft, Lattice, and Deloitte, along with leading academics from Stanford, Wharton, UC Berkeley, and more. Under the theme “Adaptive Leadership in Volatile Times,” the two-day event explored how organizations are navigating economic uncertainty, rapid AI advancement, and the evolving nature of work itself.

“This year’s theme could not be more timely,” said Dean Jenny Chatman, who co-founded the center and conference with Professor Sameer Srivastava. “What started as a conversation between colleagues—to create a community of researchers and business leaders who shared the view that organizational culture is highly consequential—has grown into this amazing community.”

The conference has become a vital forum for bridging academic research and business practice. The theme has also evolved with the times, from remote work and virtual organizations to post-pandemic organizational resilience to exploring generative AI’s impact on workplace culture. This year’s event turned to the broader challenge of leading through volatility itself.

“The combination of research-backed evidence from academics and practical advice from seasoned industry leaders yields a level of insight that could not be achieved by either perspective alone,” Srivastava said.

Academic research spotlight

The first day of the conference featured presentations from leading researchers from around the world examining organizational culture through the lenses of sociology, social psychology, economics, and computer science. The academic day concluded with the announcement of the Edgar Schein Best Student Paper Award, continuing the conference’s tradition of encouraging the next generation of organizational culture scholars. Di Tong, a PhD student from MIT, won the $2,000 award for her paper, “Manager Political Values and Gendered Work Organization.”

Mission as competitive advantage in the AI era

OpenAI Chief Economist Aaron “Ronnie” Chatterji, PhD 06, opened Day 2 with a fireside chat with Srivastava, offering a perspective on culture that comes from someone who has studied strategy as an academic and now lives it as a practitioner. Chatterji emphasized how OpenAI’s mission to “build AGI to benefit all humanity” serves as a powerful competitive advantage in the intense war for AI talent.

“You can’t win it with money alone,” Chatterji said. “The mission itself has made it so easy for me to thrive and enjoy the work. And it comes up in every single conversation, every single Slack, in a way that is pretty unique.”

Chatterji also highlighted OpenAI’s distinctive culture of questioning authority, including in open meetings with CEO Sam Altman. “It’s a very open culture where people question decisions in a way that’s very surprising to me,” he said, contrasting it with his previous experiences in government and academia, where people were more reticent about challenging a leader in public.

He warned that managers should avoid a top-down approach to implementing AI and instead focus on understanding the tasks people enjoy versus those that slow them down. “I really enjoy writing, and if someone came to me and said, ‘Hey, I’m going to do all your writing for you,’ that would not be a good pitch because you’re taking something away from me. If you say, ‘Hey, I can do your expense reports,’ I’d be more open,” Chatterji said. “You need to figure out the human piece before you can figure out the implementation piece.”

Shifting from “no, because” to “yes, if”

UC Berkeley Chancellor Rich Lyons, former dean of Berkeley Haas, joined Srivastava and Chatman to discuss adapting culture at an institutional scale. Fifteen years ago, Lyons led the Haas culture initiative—with the help of Chatman’s expertise—which led to the adoption of the four Defining Leadership Principles that remain the school’s bedrock.

Lyons said he’s been wrestling with the issue of how traditionally slow-moving universities can adapt at the lightning speed of today’s changes and emphasized the importance of “pace” as a cultural value. Lyons said he was recently introduced to a framework that can accelerate decision-making while maintaining thoughtfulness.

“We’re shifting Berkeley’s culture from ‘no, because’ to ‘yes, if,’” Lyons explained. Rather than defaulting to rejecting a new idea or request, this approach encourages people to identify conditions that need to be met, then bring creativity to addressing them. “‘Yes, if’ gets you to action,” he said, distinguishing it from “yes, and,” which facilitates conversation but doesn’t necessarily drive results.

Other key conference takeaways:

Leading through crisis with focus and calm: Arden Hoffman, BA 93 (rhetoric), Senior Vice President and Chief People Officer at General Motors, reflected on navigating multiple economic crises throughout her career at Goldman Sachs, Google, Dropbox, and GM’s self-driving vehicle unit Cruise before assuming her current role at GM in 2023.

She emphasized that the No. 1 skill she’s seen for effective leadership is being able to focus and “not to get too overloaded with the noise.” She emphasized that in volatile times, “leaders can either ramp up anxiety on a team, or they can calm it down. And I think leaders need to stay calm.”

Hoffman also commented that we’re in the early days of the AI revolution: “I think AI is overhyped in the short term and underhyped in the long term.”

Pay attention to marginalized workers: Lindsey Cameron, assistant professor at Wharton, invoked William Gibson’s observation that “the future is already here, it’s just unevenly distributed.” Cameron argued that marginalized workers serve as early indicators of broader trends. 

In service of her research, Cameron spent years as an Uber driver in the U.S. and six summers conducting ethnographic research on Uber drivers in developing countries. She noted that up to 25% of Americans now earn money through nonstandard work arrangements, while in countries like Nigeria, Brazil, and India, these workers account for up to 80% of GDP.

Organizations must adapt: Melissa Valentine, an associate professor at Stanford, presented research on “flash teams”—rapidly assembled groups of specialists convened for specific projects. “We, as a society, invented organizations to help us do complex work together,” Valentine said. Her message, outlined in her new book, is that organizations must become more fluid and modular, able to quickly assemble the right expertise for emerging challenges.

Rethink the office: Paul Cooper, Principal at TEF Design, along with One Workplace’s Chris Good, Chief Creative Officer, and Kate Rancourt, Director of Customer Insight + Growth, argued that COVID wasn’t actually a disruption but “an accelerant of a bunch of trends that already existed,” and that AI represents the true transformative force for workplace design. “We don’t want to come to the office just to simply satisfy an attendance checklist. We want to come to influence others and be influenced by others simply because we’re together, we’re sharing ideas,” Good said.

Focus on what’s in it for individuals: Lindsay-Rae McIntyre, Microsoft’s Chief Diversity Officer and Corporate Vice President of Talent and Learning, shared that the company’s AI coaching agent achieved 33% adoption with 90% repeat usage in just 90 days by providing context-specific guidance: “What our employees and our managers really want is an answer in the context of the expectations and the culture and the scaffolding of leadership for Microsoft.”

AI adoption is a change-management challenge: Deloitte Managing Director Cathryn van Namen said Deloitte’s research found that approximately 80% of AI projects fail, with lack of trust as a key factor. “It would behoove all of us to make sure that we have the right people in the room before we launch (AI),” she said. 

Transformation may take longer than expected, but go deeper. Stanford professor Chad Jones noted that transformative technologies have historically taken 30-50 years to fully materialize. “The good news from that is that we do have time, but I think we should use our time wisely.”

Specialized human labor becomes more valuable: Jones explained the “weak link” theory: “Jobs are bundles of tasks and the AI can automate three quarters of those tasks, but it just means the last 25% that we’re still doing becomes ever more valuable.” Several years ago, there were predictions that AI would take over the work of radiologists. “But today, there are more radiologists than there were then, and they are more highly paid,” he said. “In contrast, we can already see a world in which Uber drivers get fully substituted for by autonomous vehicles. To understand the effects of AI, we need to dig into the details of tasks.”

Human judgment remains critical: Sarah Franklin, CEO of the people management platform Lattice, used the analogy of calculators to argue that people must maintain domain knowledge to recognize incorrect AI outputs. “The skills most needed are definitely the human skills, the critical thinking, creativity, strategic planning, and really how we maintain our humanity when AI is there,” she said. “If we end up in a world where we just have robots talking to robots, it’s really not that interesting or compelling for anyone.”

AI has already made expertise harder to verify: Haas Professor Toby Stuart said AI can now generate content that “appears to be expert level, if not higher than expert level, so you can’t read it and distinguish it without yourself being an expert and without putting a lot of time into vetting it.” This creates acute challenges for hiring, performance evaluation, and determining true expertise across organizations. He predicted that the role of chief human resources officer may become even more important, and that AI “is going to radically and completely remake not only the organization of executive level work, but the organization of all work.”

Co-creating with AI: Léonard Boussioux, assistant professor at University of Washington, led attendees through a hands-on activity to demonstrate how AI can be used to foster creativity and bring groups together. After lunch, each table brainstormed words and ideas related to culture and chose a musical genre; Boussioux then used the information to prompt the Suno AI music platform to create “anthems” for each table. Through “vibe coding”—where developers verbally describe what they want an LLM to produce—he quickly assembled songs into an album to create a custom website. While none of the tunes would win a Grammy, they brought the group together in a fun, creative way.

Looking ahead: Culture as the foundation for adaptive leadership

The conference reinforced several critical insights: 

• Mission and culture become competitive advantages when compensation alone can’t win talent wars. 
• While AI will be profoundly transformative, the timeline may be longer than news reports suggest. 
• The most successful AI deployments focus on what’s in it for individual workers, with trust as a primary failure point. 
• And as AI handles more routine tasks, distinctly human skills become more valuable, not less.

The ninth annual Culture Connect Conference will take place in January 2027. For more information about the Berkeley Center for Workplace Culture and Innovation, visit haas.berkeley.edu/culture.