During EMTECH INVEST Davos 2026, hosted at the iconic Grand Hotel Belvédère and Mountain Plaza Hotel, we brought together investors, policymakers, scientists, and technology leaders to address a fundamental question:
Is artificial intelligence merely optimizing the present — or is it capable of expanding the future?
On January 21, within our flagship program Technology, AI & Digital Compliance, this question crystallized during the keynote titled:
“Creating Abundance and Avoiding the Zero-Sum Trap: Prioritising Creation over Reduction for Responsible AI.”
The keynote was delivered by Yuval Dvir, Senior Advisor at SandboxAQ — a global technology strategist who previously helped architect large-scale product ecosystems at Google and now works at the intersection of AI and quantum technologies.
What made this moment particularly powerful was not simply the technological argument — it was the strategic reframing of AI’s role in the global economy. Yuval challenged the prevailing narrative that celebrates efficiency as the ultimate goal. Instead, he invited us to consider whether AI should be judged not by how many costs it cuts, but by how much new value it enables.
Below, we publish his article in full, as presented, without edits.
Creating Abundance and Avoiding the Zero-Sum Trap: Prioritizing Creation over Reduction for a Responsible AI
By Yuval Dvir
The current wave of Large Language Models (LLMs) has been an extraordinary success, progressing with a speed that has transformed how we communicate and process information. Driven by human curiosity and fueled by the capitalist engine that funded it, this era has provided the essential ecosystem – infrastructure, data, compute, and investment – required for the next leap in intelligence.
However, we must recognize that this wave has been overwhelmingly leveraged for efficiency and cost reduction. While this has increased shareholder value, it may create a zero-sum trap where AI’s gain comes at a cost by failing to grow the economic pie for the wider workforce. To build a truly responsible and sustainable future, we must pivot from reduction to creation.
Beyond the Artificial Silos of Human Cognition
For centuries, we have approached the universe through a lens of fragmentation, dividing the natural world into silos like Biology, Physics, and Chemistry. These are not inherent properties of the universe but a reductionary approach, due in part to a biological coping mechanism. Because the human brain cannot process the staggering complexity of reality in its entirety, we created these artificial blocks to reduce our cognitive load. Whether our biological brain can be upgraded to manage such complexity is a discussion for another time.
The universe, however, does not know where chemistry ends and biology begins. It operates in a singular, harmonious continuity governed not by human words, but by the mathematical laws that underpin it. This convergence is sometimes referred to as singularity by some or a root node by others. Our current focus with LLMs is the ultimate expression of this human limitation. LLMs are trained on words – human constructs designed to approximate a reality we can only partially perceive. As Meta’s Chief AI Scientist Yann LeCun argues, these models lack an understanding of the underlying physical reality.
The LQM: Exploring New Sciences and Interpreting “Unknown Sciences”
Responsible AI growth must transcend linguistics. The next frontier is the Large Quantitative
Creating Abundance and Avoiding the Zero-Sum Trap: Prioritizing Creation over Reduction for a Responsible AI
By Yuval Dvir
The current wave of Large Language Models (LLMs) has been an extraordinary success, progressing with a speed that has transformed how we communicate and process information. Driven by human curiosity and fueled by the capitalist engine that funded it, this era has provided the essential ecosystem – infrastructure, data, compute, and investment – required for the next leap in intelligence.
However, we must recognize that this wave has been overwhelmingly leveraged for efficiency and cost reduction. While this has increased shareholder value, it may create a zero-sum trap where AI’s gain comes at a cost by failing to grow the economic pie for the wider workforce. To build a truly responsible and sustainable future, we must pivot from reduction to creation.
Beyond the Artificial Silos of Human Cognition
For centuries, we have approached the universe through a lens of fragmentation, dividing the natural world into silos like Biology, Physics, and Chemistry. These are not inherent properties of the universe but a reductionary approach, due in part to a biological coping mechanism. Because the human brain cannot process the staggering complexity of reality in its entirety, we created these artificial blocks to reduce our cognitive load. Whether our biological brain can be upgraded to manage such complexity is a discussion for another time.
The universe, however, does not know where chemistry ends and biology begins. It operates in a singular, harmonious continuity governed not by human words, but by the mathematical laws that underpin it. This convergence is sometimes referred to as singularity by some or a root node by others. Our current focus with LLMs is the ultimate expression of this human limitation. LLMs are trained on words – human constructs designed to approximate a reality we can only partially perceive. As Meta’s Chief AI Scientist Yann LeCun argues, these models lack an understanding of the underlying physical reality.
The LQM: Exploring New Sciences and Interpreting “Unknown Sciences”
Responsible AI growth must transcend linguistics. The next frontier is the Large Quantitative
Model (LQM), rooted in the governing laws of science rather than patterns of human text.
Unlike an LLM, an LQM is not bound by the artificial blocks of human disciplines. By training models on fundamental mathematical constants and physical laws — the true “source code” of the universe — we enable AI to detect patterns we are biologically incapable of perceiving. We can create in-silico simulations that generate entirely new datasets not available on the internet. This may accelerate existing sciences — discovering new materials, speeding drug discovery, advancing battery chemistry — but more importantly, it opens the door to “Unknown Sciences,” the vast territories between and beyond our current fragmented silos.
LQMs can flip the zero-sum dynamic by moving from wealth extraction to wealth creation. Additional models can branch from these foundational systems, while LLMs continue to serve as the linguistic interface between humans and the world around them.
This shift transitions us from an economy of Reduction — doing the same more efficiently — to an economy of Creation, discovering what we did not yet know was possible. By exploring intersections between disciplines, we unlock what is known as the Medici effect: bursts of innovation born at the convergence of ideas. From theoretical mathematics to physical reality, new industries emerge — along with new jobs and professions. Just as the digital era created the data scientist, the era of New Sciences may require simulation architects and molecular designers.
To deliver AI that creates jobs rather than removes them, leaders must adopt a framework that prioritizes creation over reduction:
Invest in the Discovery Engine by allocating capital toward LQMs capable of navigating the mathematical substrate of the physical world and catalyzing exploration beyond current limits.
Resist short-termism. Capitalism is a powerful engine for scale, but it must fuel frontier exploration — not merely automation of the existing.
Commit to human-centered augmentation. As Dr. Fei-Fei Li emphasizes, the goal is not replacement but empowerment. By using AI to tackle complex problems, we equip humanity to perceive the universe as it truly is — interconnected and unified, from atoms to galaxies.
The current wave of AI gave us tools to communicate. The next wave may give us the power to create shared abundance. The era of reduction is over; the era of creation has begun.
EmTech Invest Perspective: From Dialogue to Deployment
Yuval Dvir’s argument highlights a broader shift we have been observing around World Economic Forum week in Davos. For more than seven years, EmTech Invest has convened investors and founders representing over $500B in AUM — not to predict trends, but to understand where capital, science, and governance begin to align.
The divide between reduction and creation is not theoretical. Capital allocated to efficiency concentrates value. Capital allocated to discovery expands it.
Across conversations this January, a consistent pattern emerged — a transition:
from automation to augmentation,
from optimization to exploration,
from extraction to expansion.
The next phase moves beyond discussion into implementation: targeted collaborations, structured initiatives, and cross-sector capital formation.
EmTech Invest will continue focusing on the intersection of creation-driven AI, capital allocation, and frontier technologies.
Monaco — June 2026 | Closed-Door Capital Gathering
For the first time, EmTech Invest brings its Davos network into a private, invitation-only setting in Monaco — shifting the conversation from macro dialogue to direct capital allocation.
The gathering will convene a curated group of global family offices, institutional investors, and frontier-technology founders to structure concrete investment pathways across AI, quantum, energy, and deep-tech infrastructure.
Unlike open forums, the Monaco convening is designed for decision-makers: fewer panels, more mandates — where relationships formed in Davos convert into transactions.
Davos — January 2027
Returning during World Economic Forum week, Emtech Invest will expand the discussion from responsible AI to creation-driven global systems — from energy and biotech to decentralized finance and quantum applications.
If you are building, investing in, or regulating the next era of intelligence — we invite you to join the conversation.
The question is no longer whether AI will reshape the economy, but which model of value creation will define it.
www.emtechinvest.com
For partnerships and collaboration inquiries:
media@emtechinvest.com