Section 1
prof . dr . kathrin g . kind
interviews
Kathrin G. Kind: Interview
by Kamil Grund
It is a pleasure to be here. As someone who has spent over two deca- des navigating the labyrinth from Siemens and the Volkswagen Group to the top of QubitNexus. AI, I’ve seen technology move from the deterministic to the probabilistic, and now, to the truly “hyper-emer- gent.” Let’s dive into these complexities with the rigor of an academic and the pragmatism as a CEO. Kamil: From your experience bridging cutting-edge research and industrial implementation, what is the most critical factor that determines whether an AI initiative succeeds or fails in large organizations? Prof. Dr. Kathrin: In my experience, the pivot point between success and a catastrophic failure in large-scale AI is rarely the algorithm itself. It is Structural and Data Interoperability and the company culture and willingness to adapt and change from all levels of the hierarchy within management. Many organizations treat AI as a “plugin” rather than a fundamental rewiring of their nervous system. An initiative fails when there is a mismatch between the stochastic nature of AI and the deterministic expectations of legacy management, leadership and change mana- gement along its workforce and the ability to adapt and keep inno- vating whilst adapting a culture of growth, particularly making it open for all to continue learning from mistakes and feed this back into their knowledge management, just as the best AI models are trained . Success requires what I call Causal Literacy: understanding not just that a model predicts, but why it predicts, and having the data governance to feed it without poisoning the well. In such analo- gy, management needs to be technology literate at their foundational level, and as well as how it works and when it will not. Kamil: How do you see the role of quantum computing evolving alongside AI in the next decade — not as a competitor, but as a complementary te- chnology? What benefits for the world and future civilizations do you hope quantum computing will most enable? Prof. Dr. Kathrin: We shouldn’t view Quantum Computing as a rival; rather, think of it as the ultimate accelerator for AI’s heavy lifting. While classical AI struggles with the limits of energy capacity and performance complexity of certain optimization problems, Quan- tum—specifically via Quantum Approximate Optimization Algo- rithms (QAOA) —offers a pathway to solve “impossible” problems in real-time. It still will not replace yet all forms of computation, it will enhance it. We will be seeing from 2030 the rise of Quantum-von Neumann (conventional computers) hybrid architectures. It is time, the current architecture was invented in 1945! If we are to really have a singularity we need to advance, even beyong quantum, e.g. thermal computing. The Benefit: I hope for a “Computational Enlightenment” whe- re we use Quantum-driven AI to master Materials Science (for carbon capture) and Room-Temperature Superconductivity . At
QubitNexus. AI, we are already working on bringing these “cold” theories into “warm” industrial reality. Kamil: What are the most common ethical blind spots you observe in AI deployment today, and how can leaders proactively design against them? Prof. Dr. Kathrin: The most pervasive blind spot is what I call “Histo- rical Data Seduction”. Leaders often believe that because a dataset is “large,” it is “representative.” In reality, we are often training models on the shadows of our past prejudices. That is what the 5 v’s of data should always be observed ( Veracity, Variety, Volume, Velocity & Va- lue) and I even add a sixth in German (Vertrauen) for trust is crucial. Proactive Design: We must move toward Responsible AI by De- sign . This means implementing Diverse Redundancy —not just in hardware, but in human oversight—ensuring that the “human- on-the-loop” isn’t just a spectator, but a qualified auditor with the power to “pull the cord” when needed. We have seen already some AI Agents create extra costs by “looping along themsel- ves”, when a pair of humans, would have simply declared for that process step a new meeting and empathetic decision is needed. Kamil: Lately, the last major AI safety initiative by Anthropic announced they are pausing safety-focused development to follow competitors and compete in the market. Do you see this as a dangerous signal for the indus- try, potentially leading to a dystopian scenario where one wrong invention could have disproportionate global impact? How should companies and regulators respond? Prof. Dr. Kathrin: Pausing safety development to chase market parity is, frankly, a high-stakes gamble with a Global State of Exception . When we accelerate toward Artificial Superintelligence (ASI) wi- thout the corresponding “Brakes” of alignment, we risk a Black Swan event of unimaginable proportions! The Response: Regulators must move beyond “Voluntary Codes.” We need an International Agency for AI Safety (IAIS) , akin to the IAEA for nuclear energy. Compute-capping or “Safety-Taxing” those who bypass ethical guardrails isn’t just policy—it’s survival. Why? AI has only “Cognitive and Memorial intelligence”, it has a simulation of “Emotional Intelligence” which I question if it can be called that at all. One thing AI will never have is “Metacognition intelligence” that is the ability to: meditate, fantasize, imagine and explore beyond any trained or accessible data”. Only humans do, thus the limitations of usability end, where this unique human ability starts. Kamil: As a professor and educator, what core competencies do you believe future AI leaders must develop that are often overlooked in tra- ditional curricula?
8
DAVOS CLUB
Powered by FlippingBook