The IMD Tech Community kicked off 2026 by tackling the question that has preoccupied executives and investors for months now: Is the AI boom a bubble waiting to burst, or a genuine transformation? The panel discussion featured Karl Schmedders, IMD Professor of Finance, José Parra Moyano, Professor of Digital Strategy, and Christophe Meili, a Global Account Manager selling AI solutions, and was moderated by Juan Martin Ortiz Rocha, Head of Innovation at Cartier.
The discussion opened with an eye-catching statistic: $400bn in AI capex over the last two years needs to generate $2tn in revenue by 2030 to justify the investment. Schmedders drew parallels to previous technology cycles, noting that it took more than 30 years for electricity to fully transform manufacturing after Edison’s light bulb. The fundamental question is not whether AI represents real transformation (the panelists agreed it does) but whether returns will materialize fast enough to prevent a market correction.
The comparison to the dot-com bubble loomed large, with one critical difference: this time, real money is changing hands. Companies like Nvidia are generating billions in actual sales, not just promises. However, Schmedders raised concerns about companies like Oracle, heavily indebted and betting their future on AI infrastructure. The biggest uncertainty? Whether the massive capex investments in data centers and cutting-edge chips are truly necessary, or if breakthroughs like DeepSeek will demonstrate that similar results can be achieved with fewer resources.
“This capability opens entirely new domains, from refining individual-level predictions in insurance to transforming credit risk analysis.”
Parra Moyano offered a grounding definition: AI is “a system based on an algorithm that has learned from data and makes predictions.” But the conversation quickly moved beyond forecasting. Large language models excel at handling massive amounts of unstructured data, the messy reality most companies face. This capability opens entirely new domains, from refining individual-level predictions in insurance to transforming credit risk analysis.
Yet predictions alone don’t guarantee ROI. Schmedders cited Amazon’s zero-click ordering patent: they can predict what you’ll want, but haven’t implemented it because return costs would destroy the economics. The lesson: being world-class at one aspect of your business doesn’t guarantee competitive advantage. You need the entire value chain to work efficiently.
AI is a commodity... The same AI that I can buy, you can buy.- José Parra Moyano
Perhaps the most sobering insight came from Parra Moyano. “AI is a commodity… The same AI that I can buy, you can buy.” If everyone has access to the same technology, where’s the sustainable advantage? The answer lies in proprietary data, particularly data about how your organization operates, makes decisions, and solves problems. This internal operational data can’t be replicated by competitors.
The panelists emphasized that data quality matters more than many executives realize. As Parra Moyano warned, “No matter how good or bad you think your data is, it’s worse.” Companies that started their data housekeeping six to eight years ago now have a significant advantage in deploying AI effectively.
AI won’t eliminate jobs wholesale, but it will fundamentally change them. Journalists and coders no longer spend time generating text or basic code – that constraint has vanished. New constraints emerge around quality control and directing AI toward meaningful outcomes. As Parra Moyano noted, the skills of the future center on checking quality and providing will and intention.
The cultural challenge is immense. Companies pursuing AI with a substitution mindset will face employee sabotage – people will “forget” emails, provide wrong data, and undermine initiatives. The Klarna case study is instructive: after publicly celebrating AI-driven staff cuts, the CEO had to walk back the decision 11 months later and rehire. Trust, once depleted, is nearly impossible to rebuild, the panel warned.
Meili shared wisdom from the sales trenches: most AI pilots fail during scale-up for three reasons.
Meili shared wisdom from the sales trenches: most AI pilots fail during scale-up for three reasons. First, cost explosion: what costs pennies in a pilot becomes millions in production. Second, data reality: pilots run on clean or synthetic data, but production means confronting truly messy information. Third, risk and security: the sandbox environment of pilots gives way to enterprise systems vulnerable to attacks and cascading errors.
His advice? “You shouldn’t be doing an AI pilot. You should be doing a business pilot that happens to use AI.” Start with a clear business problem and measurable value, then apply AI as the solution.
The value often comes from process simplification, not the AI itself.
Start now, but focus on fundamentals. Don’t wait for the technology to mature, but don’t chase AI for its own sake. Begin with your data infrastructure. Document how your organization actually works, makes decisions, and solves problems. This proprietary operational data will be your competitive moat.
Embrace a growth mindset culture. Signal clearly that AI will enhance people, not replace them. Create incentives for employees to experiment and document both successes and failures. Organizations that learn to capture failure as data will have invaluable assets for training future AI systems.
Think process, not technology. Use AI as a catalyst to rethink unnecessarily complex processes. The “halo effect” of AI initiatives prompts teams to question legacy procedures that no longer serve their purpose. The value often comes from process simplification, not the AI itself.
Measure business value, not AI metrics. Define success in terms of business outcomes – reducing credit approval time to 60 seconds, increasing sales conversion rates, and lowering operational costs. Implementation of AI is simply the means to those ends.
Prepare for the long game. Transformation takes time. If your strategy depends on immediate ROI from AI, you’re setting yourself up for disappointment. The companies that survive and thrive will be those that combine strategic patience with tactical urgency – moving quickly on pilots and learning, while maintaining realistic expectations about when major returns will materialize.
The verdict on whether we’re in a bubble? The panelists couldn’t agree. Parra Moyano’s heart said crash, while his mind said sustained growth. Schmedders saw an inevitable correction given current valuations. Meili predicted a “cleansing,” where overextended players fail but the fundamental technology transformation continues unabated. Perhaps the most important insight was this: preparing your organization for the AI future doesn’t depend on predicting the bubble’s fate – it depends on building the right capabilities today.
Learn more about the IMD Tech Community and join upcoming events here.
As investment in AI accelerates, access to emerging technology skills is becoming a decisive driver of career progression and pay. New research shows that structural features of tech roles, not women’s choices...
Spotify CHRO Anna Lundström drives AI readiness, culture, and personalized well-being to empower employees and sustain Spotify’s innovative, human-centric growth....
Transform your AI adoption strategy from a high-stakes gamble into a portfolio of calculated moves using this plan, write Faisal Hoque and Paul Scade....
Organizations should let departments choose the right balance between humans and AI, and focus on outcomes, not rigid processes, to keep pace with rapid technological change. ...
Explore first person business intelligence from top minds curated for a global executive audience