By Authors:
Jason Davis – INSEAD
Interviewers:
Grant Jacoby – University of North Carolina
Article link: https://journals.sagepub.com/doi/10.1177/0001839216649350
Grant Jacoby: All right, Jason. It’s great to have you here on the ASQ blog. I wanted to start by having you introduce yourself.
Jason Davis: Sure, it’s great to be here. I’ve been looking forward to this for a while. I’m a big fan of the ASQ blog. My name is Jason Davis. I’m a professor at INSEAD in Singapore. I research innovation broadly, especially digital innovation. These days, most of my work focuses on AI, platforms, and crypto. I also do a lot of executive education with our clients in Asia.
Grant Jacoby: Excellent. Let’s dive into the first question. How did the empirical research behind your ASQ publication in 2016, The Group Dynamics of Interorganizational Relationships, shape your understanding of collaboration within innovation ecosystems?
Jason Davis: My 2016 ASQ paper on group cycling led me to believe that collaboration processes are crucial for unlocking innovation potential between individuals, teams, or organizations with specialized knowledge that must be combined. I see this same dynamic in the AI sector—getting the collaboration process right is especially important for how established organizations leverage generative AI. Large organizations, in particular, must integrate diverse knowledge of customer needs, organizational capabilities, and emerging technologies to drive meaningful innovation. Event Horizon Strategy explores how this happens in the context of generative AI.
Grant Jacoby: As AI adoption accelerates, firms must simultaneously operate across different innovation cycles. This is the definition of group cycling on multiple levels. How has your empirical research on big-tech companies influenced your perspective on the strategic tensions firms face when balancing multiple timelines in AI-driven transformations?
Jason Davis: Staying close to the computer industry through inductive, case-based research on big-tech companies has been incredibly fortunate. It quickly became clear to me that AI would be the next big wave, leading me to focus much of my teaching for MBAs and executives on strategy and organization around AI over the past decade. It’s been amazing to witness the excitement that organizations and entrepreneurs have for this latest wave of generative AI. Event Horizon Strategy provides managers with a framework to navigate this transformation and create value with AI.
Grant Jacoby: We’re all curious about your thoughts on AI strategy these days, especially how intergroup temporality shapes collaboration and decision-making in AI development. You have your new book on making these decisions, Event Horizon Strategy. For anyone who hasn’t read it, please check it out. How does your background and interactions with Tesla, Palantir, and OpenAI inform your approach to AI strategy?
Jason Davis: This book is really about how businesses can navigate the uncertainty surrounding generative AI. My research and teaching experience show that there is a lot of uncertainty, especially in large organizations. I found that a black hole is a useful metaphor for AI—it pulls everything toward it, and we can’t see what’s on the other side. Companies are making decisions without full visibility into the long-term impact.
The book applies frameworks from technology strategy and organization theory to understand how large organizations adopt AI. Cases like Tesla, Palantir, and OpenAI illustrate different strategies. For example, Tesla didn’t dive straight into AI; they started as an EV company and gradually moved into AI. OpenAI and Palantir are more forward in AI adoption and have distinct strategies. The goal is to apply ideas from tech strategy, like S-curves and experimentation, and from organization theory, like inertia and how technologies are adopted, to provide practical guidance for executives.
Grant Jacoby: How have recent events impacted the resilience and adaptability of AI infrastructure? This seems especially relevant for executive education.
Jason Davis: AI is evolving rapidly, and it’s hard to keep up. One major shift I’ve observed is the decentralization of AI. It’s happening across multiple countries and at different organizational levels, from individuals to large labs. Another big development is the rise of AI coding tools like Cursor. These tools allow a single person to build AI-powered applications, reducing the need for large, cross-functional teams. This shift is empowering smaller teams, startups, and even individuals while decentralizing the AI landscape globally.
Grant Jacoby: What are some surprising or counterintuitive findings from your book that readers should know about?
Jason Davis: One surprising finding is how AI is changing the role of teams in innovation. Traditionally, innovation relies on large, cross-functional teams. But AI is reducing the need for these big teams. Tools like Cursor, GitHub Copilot, and AI models like Claude and OpenAI empower individuals to perform tasks that once required entire teams. This empowerment is accelerating innovation and enabling startups to challenge established companies. For instance, while Tesla leads in real-world AI, new robotics startups are leveraging cutting-edge AI to compete directly with them.
Grant Jacoby: If you’re an MBA student or in executive education and want to start an AI company, is now a good time?
Jason Davis: It’s a terrific time. We’re in a window of major opportunity. This shift has even changed how I teach. At INSEAD, I lead an AI entrepreneurship course. Three years ago, it was a traditional MBA course with PowerPoint presentations. Now, students build actual products using AI tools like Replit and Cursor. Even those heading to consulting or finance are creating AI-powered applications. It’s incredible how much has changed, and students love the hands-on approach.
Grant Jacoby: Your research in group cycling has explored collaboration in alliances, which is something I’m also passionate about. How do your insights on reciprocal interdependence and group cycling connect to the AI landscape?
Jason Davis: Great question. I think we need more research on this topic. AI is shifting collaboration in two key ways. First, open-source software communities are becoming even more important. Platforms like Discord, GitHub Copilot, and X are essential for collaboration, especially for entrepreneurs. Second, we’re seeing the rise of human-LLM collaboration—people using large language models as active collaborators. This is becoming formalized, and many teams now expect their members to use AI tools. The next step will be AI-to-AI collaboration. Executives will need to figure out how to govern and manage these systems to ensure effective collaboration.
Grant Jacoby: Could you elaborate on the structural and governance challenges in AI collaboration?
Jason Davis: It’s definitely a governance challenge. While I don’t think AI will go rogue, we do need to ensure it is efficient, effective, and innovative. APIs facilitate data exchange, but the next layer involves MCP servers, which provide context for large language models. Managing this infrastructure is critical, and human oversight will be necessary to shape outcomes. Executives will need to think about how to incentivize and monitor AI systems while fostering productive collaboration.
Grant Jacoby: Google’s AI co-scientist just launched recently. It’s another example of potential AI-to-AI collaboration. What do you think?
Jason Davis: Absolutely. It’s fascinating how quickly these tools are adopted and copied. OpenAI introduced deep research, and soon after, Grok and Perplexity launched their own versions. These tools combine reasoning, intelligence, memory, and search to deliver deeper insights. The pace of innovation is incredible—you need to check for new developments every week to stay current.
Grant Jacoby: For executives navigating this fast-paced environment, how do you integrate experimentation into a strategic AI roadmap?
Jason Davis: Flexible experimentation is key. My research identifies three main strategies. Some companies delay investment, which is risky in the face of rapid AI advancement. Others bet their entire future on AI, which is impractical for most established organizations. The best approach is phased experimentation—making incremental investments, learning from each phase, and building on successful innovations.
Tesla exemplifies this. They began with EVs, which provided data to develop basic machine-learning models. This evolved into advanced AI for full self-driving (FSD) and eventually humanoid robots. This approach allows companies to capture value while staying adaptable. It balances immediate returns with long-term learning, ensuring companies don’t overcommit to a technology that might change direction.
Grant Jacoby: Incredible insights. For anyone reading, be sure to check out Jason Davis’s new book, Event Horizon Strategy. AI strategy is a big deal, and it’s only going to get bigger.
Jason Davis: This was really fun. Thanks for having me.
Interviewer bios:
Grant Jacoby is a PhD Student in Strategy and Entrepreneurship at the University of North Carolina’s Kenan-Flagler Business School, where his research examines organizational learning dynamics and cross-sector collaboration mechanisms in technology-driven R&D environments. Building on 7 years of operational leadership in Silicon Valley’s electric vehicle (EV) infrastructure and artificial intelligence sectors, Grant’s work bridges scientific progress with theoretical advancements in innovation.
Get new ASQ Blog posts by email