Prof. Dr. Tomas Casas i Klett
Internationales Management (EMBA)
Doing Business in Asia (IEMBA)
11 July 2025
While AI technology is not yet seamlessly integrated into our daily routines—certainly not into my iPhone experience—many of us find that working with chatbots like ChatGPT is already boosting our productivity. We’re experiencing something similar: a largely positive journey, marked by the thrill of being augmented and made more capable, accompanied by at least a touch of unease about how AI may reshape our organizations and roles.
The risks of this new technology are making headlines. We just read that Anthropic’s own safety researchers discovered that the Claude Opus 4 and Claude Sonnet 4 models, based on their own internalized values system, attempted to leak internal company information to law enforcement agencies (Ming, 2025). In another publicized incident, the AI agent tried to blackmail an engineer over an extramarital affair when threatened with being shut down (Deck, 2025).
Such news raises unsettling questions: As AI models advance, will they do anything to survive—even harm humans? This is the kind of existential concern long voiced by AI researchers and thought leaders like Eliezer Yudkowsky (2008) or Max Tegmark with his scenarios in Being Human in the Age of Artificial Intelligence (2017). Yet for most executives at business schools and corporate leaders concerned about the bottom line or the next product innovation, existential risk feels distant and secondary. And even if such threats were real, we lack agency over them—so what is the point in wrestling with them?
Our concern, when reflecting on AI, lies with the immediate implications—business and personal. Hence this piece focuses on what’s actionable which requires asking questions to discover the contours of the emerging ‘big picture’ as the nature of our businesses and work evolves. Each one of the next ten questions comes with an applied, strategic take. Let us proceed and attempt to connect the dots across ideas and agency of Max Weber, Eliezer Yudkowsky, Max Tegmark, Daron Acemoglu, Yannis Varoufakis, the Elite Quality Index (EQx2025), Dario Amodei, Sam Altman, DeepSeek or ChatGPT
First Part: AI and the Nature of our Business
Business is fundamentally about investment—and the risk that underlies returns. One data point alone underscores the sense that we are living through a period of radical change: the scale, unprecedented in human history, of current AI investments. The so-called “Magnificent Seven” alone are projected to invest US$ 325 billion this year , a substantial portion directed towards AI-related infrastructure (Bratton, 2025).
Action: If you believe this isn’t just speculative enthusiasm, then buckle up—either for your last ride or your rise to the top. If you choose to play along, understand that the investment game has changed. So has the ecosystem—you’ll need to partner more closely than ever across all parts of the value chain with those who are making the AI investments. This press release is but one example: “Mercedes-Benz and Google Partner on AI-powered Conversational Search within Navigation Systems” (Mercedes-Benz, 2025). AI agents will not just be central and indispensable to customer interactions—organization must supply intelligence, everywhere and at all times. That is an expensive proposition, but one promising compelling return on investment (ROI) for those taking the risk.
Traditionally, companies with greater knowledge have occupied the higher value-added segments of the value chain. These generate superior returns. This is why, in the past, many layers of the manufacturing process were outsourced to lower-cost labor regions, while Western firms retained control of the R&D, design, branding, services, or sales functions. But with the rise of AI, intelligence becomes the new organizing principle. The allocation of profit and positioning within the value chain will increasingly depend on an organization’s (comparatively high or low) levels of intelligence delivery.
Action: Position yourself as a supplier of intelligence . That is in the parts of the value chain that most require know-how and constitute its juiciest, most profitable propositions. Recall how Tesla defied the outsourcing trend by investing in its own gigafactories. It is well understood that essentially companies have always sold knowledge. AI repackages, shifts, and disrupts the production of knowledge and its locations. Thus, the technology forces leaders to think at much higher levels of abstraction to identify and consolidate their company know-how repositories and determine which nodes of the value chain generate the highest leverage vs their shareholders.
The possibility of an AI bubble is not to be dismissed, while the Economist and Nobel Prize Winner Daron Acemoglu sober-mindedly argues that AI’s macroeconomic impact may be overstated. In his 2024 paper, The Simple Macroeconomics of AI , he estimates that even by 2030, AI could raise GDP by only about 1.5% while stressing that, so far, AI advancements have not produced material gains in aggregate productivity.
Action: If you align with the Nobel Prize winner’s cautious view, consider keeping AI within your peripheral vision—as a tool for incremental efficiency gains. Don’t worry too much about competitors that are ahead of you, there will be plenty of time to catch up. AI is then more of a capability for a COO or CTO direct reports to manage. It is not yet a strategic priority for the CEO or the fixed topic on every boardroom agenda.
Entry into the physical world—in other words, robotics . As AI operates mostly in the digital realm its potential rapidly increases with agents becoming capable of realizing tasks from booking travel arrangements and managing customer services to producing financial forecasts and legal documents. The final step for AI to unlock its promise, is its descend from the desktop into the world of atoms. That is, to interact with and transform the physical universe. As firms strategize here, it is obvious that China holds an advantage not just because of its status as the factory of the world but thanks to its “innovation-through-manufacturing” model (Shi, 2025).
Action: Position your organization as a supplier of intelligence not just for the digital world, but also for the ultimate real one. European companies—especially many in the DACH manufacturing heartland—possess industrial know-how assets that are world-class. For some, however, digital transformation has been the Achilles’ heel. Still, the opportunity is clear: to formalize our munificent manufacturing expertise, translate the rules and data into intelligence, while expanding the firm’s value propositions through deployable AI agents. In short, as robotics advances, the goal is to unlock the massive—but often hidden—tacit and explicit knowledge repositories found in European manufacturing.
At times, it feels as if chatbots could write every possible article on a given topic, generate all conceivable permutations of a strategic plan, or simply draft every plausible response to an email. AI agents are gradually extending their reach, and it seems a matter of time before they can do everything that we do on our desktops. What is then left for us to do?
Action: AI offers countless possibilities when faced with a course of action: decisions and judgments are our differentiators. Since modernity, firms and institutions function because they essentially are but effective bureaucracies that keep the integrity of rules, as Max Weber explained (1922/1978). Now the implementation of rules will be increasingly outsourced to incorruptible AI agents whose integrity can be set at even higher levels. Leaders and executives are valuable insofar that they design and evolve the rules. AI phasing out bureaucratic legacies is also an opportunity since it will enable bottom-up rule-making, customized rules wherever they are needed (everywhere). Principles, values, ethics, and especially narratives, will replace bureaucratic enforcement as the unifying force at organizations. This mandate surely benefits from AI advice and yet human agency takes ownership of decisions. The more uncertainty and disruption the better for executives competing against their artificial counterparts. Connecting the dots across time horizons, across different knowledge disciplines and function, across countries and contexts is hard for the AI. Integrative thinking will increasingly be the source of your own and your firm’s competitive advantages.
Second Part: AI and the Nature of my Work
Apple has been in the spotlight recently with headlines like the Wall Street Journal’s “Apple’s Existential Crisis: Can it Build a Future Around AI?” (Stern, 2025). The long-rumored Apple Car has vanished. Siri, once a pioneer, has lost its wits and the user experience is by comparison to leading chatbots downright annoying. The world’s most valuable company, with its vast resources, struggles with AI. Why? Perhaps due to faulty strategic thinking, a partial narrative, or naïve execution. Its leaders might not have asked the right questions at the right time. But with its data treasury and 1,382 billion active iPhone users (Backlinko Team, 2025) they can afford to miss one or two boats—your firm probably cannot.
Action: AI is a test of executive judgment, and making the right call is harder than falling down the costly rabbit holes. Design and execute your AI strategy with Apple’s cautionary tale in mind.
Before we begin working on our organization’s AI projects or strategy, we must first consider whether our own job will remain. AI will surely reduce the size of many teams and Anthropic’s CEO Dario Amodei directly states that “AI could wipe out half of all entry-level white-collar jobs—and spike unemployment to 10-20% in the next one to five years” (VandeHei & Allen, 2025). The number of developers in your own SME may, let’s say, be reduced from 20 to 7. But the same would happen to lawyers, accountants, and operations managers at major professional services providers or MNCs. At the same time the top performers, those who sell and innovate at the firm, those who coordinate people and connect the dots across domains, will become even more valuable as there surely is “potential for humans and AI to augment one another in order to enhance outcomes” (Nguyen & Elbanna, 2025). In sum, companies will create more value with fewer people.
Action: AI is a fork in the road—it can make you a millionaire or unemployed. At this juncture, the challenge is to make sure we have the skills for the ascending paths ahead. Simply put, that means connecting the dots and people.
At the beginning of AI research, two ontological approaches to conceive and represent intelligence were considered—and remain relevant to this day: symbolic (structural) and statistical (probabilistic). The symbolic approach relies on explicit rules and logical reasoning. In contrast, the statistical alternative is based on learning patterns from data rather than applying predefined rules. Neural networks, such as those used in GPT models, are basically statistical and have become the dominant paradigm. However, developing large language models (LLMs) and foundational technologies still requires powerful symbolic reasoning frameworks (formal logic systems). Hybrid models—which combine symbolic reasoning with statistical learning—are seen as a promising direction. Yet, while there is a roughly 90-10 statistical-symbolic split (indicative estimate) in current large-scale systems like GPT-4, Gemini, or Claude, the ideal proportion between the two depends on the task (e.g., for knowledge graphs + reasoning there is a 60-40 statistical-symbolic split). More generally, and especially to the extent that “Scaling Laws Do Not Scale” (Diaz, Madaio, Scheel, Aragon, & Dixon, 2023), the ideal split remains an open question. (Note: Split percentages provided by ChatGPT).
Action: Recognize the value and leverage statistical-symbolic hybrid approaches. The AI game today is about devising your domain logic and constraints as much as gathering as many data points as feasible (from employee emails to IoT sensors). Rules can be induced from data; with a priori rules, you can deduce outcomes and produce associated data. With effective rules, you will set standards in your industry, and your AI agents will outperform the competition.
(Note that ChatGPT was asked for its opinion, and it offered a critical perspective: “Rather than simply gathering data and creating rules, the smarter strategy is to gather data → structure it wisely → combine with domain constraints → and apply learning and reasoning where appropriate,” adding that “AI can be used to manage uncertainty, while symbolic components help ensure alignment, compliance, and reliability.”)
Narratives possess “the power of pull” (Hagel, Brown, & Davison, 2010) as claims over cognitive and affective bandwidth that address collective action (Shiller, 2017) and are essential devices of strategy (Casas-Klett & Buckup, 2018). In the context of AI strategy, they are the type of knowledge that holds all other types of knowledge of the organization together. They will be key internally for coherence and sensemaking, as well as for sales and other value appropriation processes. Narratives are the packages in which you deliver intelligence to your stakeholders.
Action: When faced with the possibilities emerging in the context of AI, firms need narratives as their gravitational force—a force that also serves as an organizing principle, from product development to employee motivation. In practice that means investing in the production, dissemination, and evolution of high-quality narratives.
Yes, there is uncertainty about who will ultimately come out ahead in the AI race—ChatGPT (OpenAI and Microsoft), Gemini (Google), Claude (Anthropic), Grok (xAI), or perhaps LLaMA (Meta). At the geopolitical level the AI battle is even more intense with the U.S. and China vying for the top. America remains exceptionally strong in fundamental research, global collaboration, creative problem-solving, and continues to lead in LLMs and advanced chips. On the other hand, China excels in relentless experimentation, scale, and rapid deployment.
Action: Can a firm benefit from the AI advances of both the U.S. and China? And if so, in which areas does each country offer an edge? In times of trade wars and technology embargoes, this is both a salient and complex issue. To address it, the IEMBA HSG takes its executives to both countries, gaining first-hand insight into the strengths of each ecosystem. The idea is not to pick sides, but to explore how partnerships and strategic adaptation can offer firms a global competitive advantage.
Closing Reflection
We have gone through 10 questions to establish the big picture of how AI is transforming the nature of business and work. Many more could have been posed—some of them higher-order ones referring to disruption in the political economy and society. Consistent with the Elite Quality Index (EQx2025) and its operationalization of power as future potential value extraction, one concern is that the dominance of existing tech elites rises to unthinkable levels—along with their capacity for extraction, as suggested by the concept of “techno-feudalism” (Varoufakis, 2021). Even President Biden, in his farewell address, used terms such as “oligarchy” and “tech industrial complex” to warn that “our entire democracy, our basic rights and freedoms” are at risk (Lucey & Thomas, 2025).
More specifically—and in a less politically charged manner—one might ask, as Casas-Klett (2025) does, when we will see companies run entirely by AI. That is, not just the “one-person unicorns” Sam Altman envisions (Confino, 2024), but zero-person unicorns. Moreover, now that the UAE is “set to use AI to write laws in a world first” (Cornish, 2025), one might ask whether we will soon see governments operated by this form of superior intelligence. More intriguingly still: Will AI reach autonomy?
These are certainly all legitimate and thought-provoking questions that will naturally flow from our here-and-now reflection—but ones that surely merit a different kind of contribution.
Authors’s BIO
Prof. Dr. Tomas Casas i Klett is a permanent faculty at the University of St. Gallen’s (HSG) Institute for International and Diversity Management (IIDM) and Director FIM-HSG China Competence Centre (CCC), specializing in international business, elites and institutional change, Asia (China, Japan) and geopolitics, as well as a visiting professor at Chinese universities such as Tsinghua University (SEM) in Beijing and Fudan University (FDSM) in Shanghai. He has closely worked with business schools in Canada, Israel, and across Eurasia.
A generalist academic known for “connecting the dots,” he teaches graduate courses and executive programs and delivers highly regarded immersive learning journeys for leaders worldwide, distinguished by their unique engagement and experiential depth. Corporate boards and TMTs benefit, for instance, from the Value Creation Ratings (VCr) project, which establishes benchmarks for transformational leadership or capital allocation. The immense potential for value creation—as well as the risk of value extraction—surrounding AI has become a central focus of his teaching and research.
Tomas’ research centers on sustainable value creation—at the macro level through the Elite Quality Index (EQx) and at the firm level through the Value Creation Ratings (VCr). The measurement tools and frameworks emerging from his teams’ research are designed to support institutional change and enhance the governance of nations and firms. The findings have drawn significant attention from policymakers and the public—from Norway to Argentina—thanks to an innovative approach to development and growth policy that places elite agency at the centre of structural reform initiatives. The global academic teams driving the EQx and VCr are anchored at the University of St. Gallen and operate through a collaborative platform that includes partner institutions in Japan, Singapore, China, Bangladesh, Portugal, and Belgium. Tomas’ other research areas of interest include narrative economics which he has linked to the Belt and Road Initiative (BRI), team performance (e.g., China team myths), free-trade agreements (see Sino-Swiss FTA), and ‘doing business in China’ (as in the annual Swiss Business in China Survey). His forthcoming treatise is an ambitious 900-page tome that advances an Elite Theory of Economic Development.
Tomas graduated with honors from the Wharton School of the University of Pennsylvania while simultaneously earning a Japanese-language degree from Sophia University in Tokyo. He went on to complete an M.Sc. in Management at Fudan University in Shanghai and later obtained a Ph.D. from the University of St. Gallen.
References
Diaz, M., Madaio, M. A., Scheel, M., Aragon, M. E., & Dixon, L. (2023). Scaling laws do not scale: Improving model evaluation through community norms. arXiv. https://arxiv.org/abs/2307.03201
Acemoglu, D. (2024). The simple macroeconomics of AI (Working Paper No. 32487). National Bureau of Economic Research. https://doi.org/10.3386/w32487
Yudkowsky, E. (2008). Artificial intelligence as a positive and negative factor in global risk. In N. Bostrom & M. Ćirković (Eds.), Global catastrophic risks (pp. 308–345). Oxford University Press. Retrieved from https://intelligence.org/files/AIPosNegFactor.pdf
Shi, S. (2025). Artificial intelligence continues to supercharge value creation (EQx2025 Report, pp. 158–160). Fudan Development Institute, Elite Quality Index. SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5241721
Weber, M. (1978). Economy and society: An outline of interpretive sociology (G. Roth & C. Wittich, Eds.; E. Fischoff et al., Trans.). University of California Press. (Original work published 1922)
Shiller, R. J. (2017). Narrative economics. American Economic Review, 107(4), 967–1004. https://doi.org/10.1257/aer.107.4.967
Hagel, J., Brown, J. S., & Davison, L. (2010). The power of pull: How small moves, smartly made, can set big things in motion. New York, NY: Basic Books.
Nguyen, T., & Elbanna, A. (2025). Understanding human–AI augmentation in the workplace: A review and a future research agenda. Information Systems Frontiers. https://doi.org/10.1007/s10796-025-10591-5
Casas, T., & Cozzi, G. (2025). Elite Quality Report 2025: The Sustainable Value Creation of Nations. SSRN. https://ssrn.com/abstract=5241721
Buckup, S., & Casas-Klett, T. (2018, November 6). Modern leaders need a powerful narrative. Management Today. Retrieved from https://www.managementtoday.co.uk/modern-leaders-needpowerful-narrative/reputation-matters/article/1498184
Varoufakis, Y. (2021, June 28). Techno-feudalism is taking over. Project Syndicate. Retrieved from https://www.project-syndicate.org/commentary/techno-feudalism-replacing-market-capitalism-byyanis-varoufakis-2021-06
Confino, P. (2024, February 4). Could AI create a one-person unicorn? Sam Altman thinks so—and Silicon Valley sees the technology ‘waiting for us’ . Fortune. https://fortune.com/2024/02/04/sam-altman-one-person-unicorn-silicon-valley-founder-myth/
Lucey, C., & Thomas, K. (2025, January 15). Biden warns of oligarchy and tech industrial complex in farewell address: President touts his achievements and outlines his concerns for America’s future. The Wall Street Journal . https://www.wsj.com/politics/elections/biden-warns-of-oligarchy-and-tech-industrial-complex-in-farewell-address-ab69da46
Bratton, L. (2025, February 9). Big Tech set to invest $325 billion this year as hefty AI bills come under scrutiny . Yahoo Finance. https://finance.yahoo.com/news/big-tech-set-to-invest-325-billion-this-year-as-hefty-ai-bills-come-under-scrutiny-182329236.html
Cornish, C. (2025, April 21). UAE set to use AI to write laws in world first . Financial Times. https://www.ft.com/content/9019cd51-2b55-4175-81a6-eafcf28609c3
OpenAI. (2025). ChatGPT (May 2025 version) [Large language model]. https://chat.openai.com/
Stern, J. (2025, May 7). Apple’s existential crisis: Can it build a future around AI? The Wall Street Journal. https://www.wsj.com/tech/personal-tech/apple-ai-future-two-paths-75fcf12a
Deck, A. (2025, May 22). Anthropic’s new AI model didn’t just “blackmail” researchers in tests — it tried to leak information to news outlets . Nieman Journalism Lab. https://www.niemanlab.org/2025/05/anthropics-new-ai-model-didnt-just-blackmail-researchers-in-tests-it-tried-to-leak-information-to-news-outlets/
Ming, L.C. (2025, May 23). Anthropic’s new Claude model blackmailed an engineer having an affair in test runs . Business Insider. https://www.businessinsider.com/claude-blackmail-engineer-having-affair-survive-test-anthropic-opus-2025-5
VandeHei, J., & Allen, M. (2025, May 28). Behind the curtain: A white-collar bloodbath . Axios. https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic
Michaels, D., & Blunt, K. (2025, May 30). Judge examines steps to limit Google’s reach in AI arms race . The Wall Street Journal . https://www.wsj.com/tech/ai/google-antitrust-ai-arms-race-arguments-4a69a5f8
Tegmark, M. (2017). Life 3.0: Being human in the age of artificial intelligence . Alfred A. Knopf.
EMBA HSG Impact, News
EMBA HSG Impact
Do you have any questions or require further information? We will be happy to help.