Why Context Has Become the Most Dangerous Asset in the World
By Vaughn Woods, CFP, MBA
Vaughn Woods Wealth Management | July 2026
“For I do not understand my own actions. For I do not do what I want, but I do the very thing I hate.”
— Romans 7:15
What My Thesis Didn’t Warn Me About
In 2007, I wrote a master’s thesis nobody read. It sat in a drawer, then a hard drive, then a cloud folder I forgot the password to. The title was clunky: Context: A neuroeconomic solution set for the integration of intelligent models (Woods, 2007). I was trying to answer a narrow question. If you have several smart models — statistical models, decision models, early machine learning systems — how do you get them to agree on what they’re even looking at? My answer was context. Not data. Not computing power. Context: the shared meaning that lets separate systems talk about the same thing in the same way.
I thought I was writing about finance. I was wrong. I was writing about power.
It took almost two decades, but the world caught up to that thesis. Today, the largest defense contractor in Silicon Valley, the biggest cloud data companies on earth, and entire national governments are racing to build exactly the thing I described in that drawer. They call it an ontology. They call it a semantic layer. They call it a universal catalog. I call it what I called it in 2007: the solution set that lets intelligent models work together. The only difference is the stakes. Back then, context was an academic curiosity. Now it is the operating system of civilizational power — economic, military, and ethical. This piece is my attempt to walk you through how that happened, why it matters to your portfolio, and why it should also matter to you as a citizen.
Page 1 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
Context Is the New Crude Oil
For most of the last decade, the popular metaphor was “data is the new oil.” I never liked it. Data, on its own, is inert. It just sits there, like oil in shale rock nobody has figured out how to extract. What turns data into value is refining it into something usable. That refining process is context: labeling what the data means, how it relates to other data, and what it’s for.
Here is the uncomfortable truth I’ve come to accept. Artificial intelligence models, no matter how large or sophisticated, are only as good as the context wrapped around them. A large language model with no context about your company’s supply chain, your government’s classified networks, or your firm’s compliance rules is a brilliant mind with amnesia. It can reason, but it cannot act responsibly, because it does not know where it is standing. Context is what tells the model where it is standing.
This is why the companies building context infrastructure, not just bigger models, are pulling ahead. Palantir Technologies reported first-quarter fiscal 2026 revenue of $1.6 billion, up 85% year over year, with U.S. commercial revenue reaching $595 million, up 133% year over year (Palantir Technologies, 2026). The company raised its full-year 2026 guidance to $7.65 billion. Those are not the numbers of a company selling software licenses. Those are the numbers of a company selling the plumbing that lets an organization’s entire universe of data and decisions become legible to a machine. That plumbing is context, industrialized.
Crude oil made nations rich because it powered engines. Context is making a smaller number of companies extraordinarily valuable because it powers something more consequential than engines: decisions. And whoever controls the layer that shapes a decision controls more than money. They control outcomes.
From Neuroeconomics to National Security: The Model Stack
Page 2 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
My original thesis lived in the world of neuroeconomics, a field that studies how the brain actually makes decisions under uncertainty, blending psychology, economics, and neuroscience. The core insight I borrowed from that field was simple. The human brain does not process the world as raw sensory data. It processes the world through context: memory, framing, prior experience, and relationships between concepts. Strip the context away and even a healthy brain cannot function. It cannot tell a threat from a friend, or a bargain from a trap.
I argued that intelligent computer models would face the identical problem. A model is a narrow reasoning engine. Stack several of them together, and you don’t automatically get a smarter system. You get several smart but isolated engines, each blind to what the others know. What binds them together, what I called the “solution set for integration,” is a shared contextual layer sitting above the individual models.
Fast forward to today, and this idea has become the literal architecture of national security infrastructure. On June 29, 2026, Palantir and NVIDIA announced a Sovereign AI Operating System Reference Architecture, integrating NVIDIA’s open Nemotron models with Palantir’s AIP, Ontology, Foundry, and Apollo products for classified and air-gapped government environments (Palantir Technologies, 2026). What makes this architecture significant is not the model itself. Nemotron is one of many capable open models. What makes it significant is that Palantir is model-agnostic: AIP lets government customers switch between accredited large language models, including those certified under FedRAMP, Department of Defense Impact Level 6, and Intelligence Community Directive 503 standards, depending on the mission at hand (Palantir Technologies, 2026). The model is replaceable. The context layer is not.
Speaking at the AWS Summit in Washington, D.C., Palantir executive Lucas Sammons described how sovereign governments are approaching AI not as a single tool purchase, but as an architecture decision, one that determines who can act on classified information and how fast (Sammons, 2026). That is neuroeconomics at the level of the nation-state. The brain has become the bureaucracy. The context layer has become the nervous system connecting classified data to real-time decisions.
Page 3 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
The Ontology Layer: Where Palantir Meets Power
Let me try to describe Palantir’s Ontology in plain terms, because the word itself scares people off. In philosophy, an ontology is simply a way of describing what exists and how those things relate to each other. Palantir borrowed the term for something surprisingly literal. Its Ontology sits above individual AI models as a semantic layer, mapping an organization’s data, systems, and processes into a machine-readable structure. In practice, this means a hospital’s patient records, a factory’s supply chain, or a military command’s logistics data are translated into a common language that any accredited AI model can read and act on responsibly.
This matters because it flips the usual sequence of automation. Historically, software followed rules that humans wrote in advance. Agentic AI, the current frontier where AI systems take multi-step actions rather than just answering questions, needs something different. It needs to understand the relationships between a decision and its consequences before it acts. The Ontology is what allows agentic AI to, in industry language, “pull decisions forward,” meaning the system can propose or even execute an action rather than merely summarizing information for a human to act on later.
There is also a defensive dimension that deserves more attention than it gets. The Ontology functions as a “semantic boundary,” preventing large language models from absorbing and permanently storing classified or proprietary information inside their own weights, the internal parameters that give a model its knowledge. In other words, the model can use the data to reason, but it cannot memorize and leak it later. This is not a minor technical footnote. It is the difference between an AI system a government can trust with classified material and one it cannot. Context, in this sense, is not just about capability. It is about containment.
Semantics, Metadata, and the Architecture of Control
Page 4 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
Palantir is not alone in this race, and that is exactly why I believe the thesis has graduated from theory to infrastructure. Snowflake, one of the largest cloud data platforms in the world, introduced what it calls the Universal AI Catalog inside its Horizon governance suite. The explicit goal, in the company’s own words, is to give data “business meaning” through a semantic layer, and it is now backing an industry-wide effort called the Open Semantic Interchange initiative to create open, vendor-neutral semantic standards (Snowflake, 2026). Read that carefully. A major cloud vendor is trying to standardize the very thing my 2007 thesis argued was the missing layer: a shared semantic structure that lets independent systems interpret data the same way.
Databricks, Snowflake’s chief rival, made the identical bet. Its Unity Catalog Business Semantics capability went generally available in 2026, allowing organizations to define governed metrics, dimensions, and business rules exactly once, then share that single definition across every dashboard, every query, and every AI agent that touches the data. Two competing trillion-dollar-adjacent companies, building the same missing layer, independently arriving at the same architecture. That is not coincidence. That is convergence on a real and previously underappreciated problem.
Chuck Brooks, writing in Forbes, frames the strategic stakes clearly: sovereign AI, meaning a nation’s ability to control its own AI destiny, requires owning the full stack, which he defines as control over the data layer, the model layer, and the interaction layer together (Brooks, 2026). Weakness in any single layer, he argues, undermines the whole structure. A country can build the most advanced model in the world, but if a foreign vendor controls the semantic layer interpreting that nation’s data, sovereignty is an illusion. This is the geopolitical translation of my old neuroeconomic argument. Whoever controls context controls interpretation. Whoever controls interpretation controls action. And whoever controls action, at scale, controls power.
Norm-Based Ethics in a Model-Driven World
Here is where the story turns from fascinating to genuinely unsettling, and I want to walk through it slowly because it deserves care.
Page 5 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
There is a long-standing observation, well documented by legal scholars and government ethics researchers, that the United States operates on a norm-based ethics system layered on top of incomplete law. In many areas of public life, the law sets a floor but relies on voluntary restraint, tradition, and the weight of institutional judgment to carry the load above it. What has historically restrained self-dealing behavior at senior levels of government — in the executive branch, in regulatory agencies, in corporate boardrooms — is not always statute. It is norms. The OECD’s 2026 integrity indicators rate the United States highly on conflict-of-interest regulation on paper, yet the researchers are careful to note that much of the system’s real integrity depends on these unenforced, voluntary norms rather than binding legal walls (OECD, 2026, as referenced in governance research).
Now overlay everything I’ve just described about context, ontology, and semantic layers onto that fragile arrangement. Palantir’s Gotham platform, built originally for intelligence and defense analysis, along with AIP and the Ontology, does not just store information. It maps relationships, patterns, and behaviors. It operationalizes norms, meaning it can turn an unwritten social expectation into a trackable, queryable, even automatable data point. A system that can map a norm can also, in principle, be used to identify when a norm is being tested, bent, or quietly abandoned, by anyone with access to the platform, for purposes both protective and predatory.
This is not a hypothetical concern confined to government. Research from the University of Chicago Booth School of Business in 2026 found that agentic AI inherits and amplifies human heterogeneity, meaning the outcomes these systems produce track the identity, incentives, and behavioral characteristics of whoever built or directed them (University of Chicago Booth, 2026). An economics analysis from MIT in 2026 went further, warning of a phenomenon researchers call “knowledge collapse,” in which agentic AI can provide individually helpful answers to a single decision-maker while simultaneously causing dynamic harm to the collective pool of human knowledge over time (MIT Economics, 2026). Put plainly: a tool can make you personally smarter today while making society collectively dumber tomorrow, because it quietly narrows the range of ideas and interpretations everyone relies on.
Page 6 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
When you combine a norm-based ethics system, one that depends on restraint rather than statute, with an AI infrastructure capable of mapping, predicting, and potentially exploiting exactly where those norms are weakest, you get something genuinely new. Not merely a technology risk. A governance risk. Context, my old academic obsession, has become the thing that decides whether norms hold or collapse, because context is what determines whether a system recognizes a norm being violated at all.
The Investor’s Dilemma: When the Infrastructure of Power Trades at 120x Revenue
Now, let’s talk money, because that is why you’re reading a wealth management blog and not a philosophy journal.
Palantir, the clearest public proxy for this entire context-and-ontology thesis, trades at a valuation multiple that would have been considered absurd for any software company a decade ago, frequently cited in the neighborhood of 100 to 120 times trailing revenue depending on the exact measurement window. Traditional valuation frameworks, the ones I was trained on and have used for most of my career, strain to justify multiples like that through discounted cash flow alone. So what is actually being priced?
I’d argue the market is pricing something closer to option value on civilizational infrastructure. Investors are betting that whoever owns the semantic layer, the Ontology, the Universal AI Catalog, the Unity Catalog, becomes structurally difficult to displace, the way owning the electrical grid or the interstate highway system would be. That is a real and defensible thesis. It is also a thesis that assumes governments and enterprises will keep paying premium prices for sovereignty, security, and semantic control, rather than commoditizing that layer over time as open standards like the Open Semantic Interchange initiative mature (Snowflake, 2026).
Page 7 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
This is the investor’s dilemma in its purest form. If context infrastructure becomes a genuine utility layer of the global economy, today’s valuations may look cheap in hindsight, the way early internet infrastructure investments eventually did for the survivors. If, instead, open standards commoditize the semantic layer the way TCP/IP commoditized internet plumbing, decoupling the value from any single vendor, today’s premium multiples become very hard to defend. I do not believe anyone, including the analysts covering these names professionally, can tell you with certainty which outcome wins. What I can tell you is that the underlying thesis, that context and semantic integration are foundational rather than incidental to AI’s value, is no longer speculative. It is being built, funded, and deployed in classified government environments as we speak.
What This Means for You
If you take one idea from this piece, let it be this. The next phase of the AI story is not really about which model is smartest. It is about which organizations, and which nations, control the context layer that tells those models what the data means and what to do about it. That was my thesis in 2007, written from a place of academic curiosity about how brains and machines integrate scattered information into coherent judgment. It has become, without my planning it, a lens for understanding some of the largest capital allocation decisions and geopolitical maneuvers happening right now.
For your own financial plan, this suggests a few honest, unglamorous conclusions rather than a hot stock tip. Exposure to this theme, if you choose to take it, should be sized like the high-uncertainty, high-consequence bet that it is, not like a core holding you’d stake a retirement on. Diversification matters more, not less, when a handful of companies are converging on the same architectural bet, because a shift in open standards could compress multiples across the group simultaneously rather than in isolation. And perhaps most importantly, pay attention to governance, both corporate and governmental, because the norm-based ethical scaffolding I described earlier was never built to withstand tools this powerful. As citizens and as investors, we should want the institutions using this technology, public and private alike, to be at least as thoughtful about context as the software now is.
Page 8 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
I never expected a drawer thesis about neuroeconomics to become relevant to how I think about your portfolio, or about the country. But context, it turns out, was never just an academic solution set. It was always about who gets to decide what the world means, and now, for the first time, a machine can help decide that too.
Context Changes Everything. So Does the Right Advisor.
In 2007, Vaughn Woods wrote a master’s thesis arguing that context — not data, not models — would become the decisive layer of intelligent systems. That framework now informs every conversation he has with clients about where the economy is heading and how to position a portfolio for it.
Vaughn Woods, CFP®, MBA is a wealth management advisor based in San Diego. He helps high-net-worth families, entrepreneurs, and successor trustees navigate portfolio construction, estate planning, and the kinds of structural market shifts described in this piece. His practice, Vaughn Woods Financial Group, Inc. is built on the conviction that the best financial advice is not reactive — it is anticipatory.
If you found this analysis useful and want to discuss what it means for your specific situation, reach out directly: vw@vaughnwoods.com or 1-800-374-4412 or follow.
Connect on LinkedIn: linkedin.com/in/vaughnwoods
References
Brooks, C. (2026, April 22). Sovereign AI: Why owning the full stack is the new strategic imperative. Forbes. https://www.forbes.com/sites/chuckbrooks/2026/04/22/sovereign-ai-why-owning-the-full-stack-is-the-new-strategic-imperative/
Palantir Technologies. (2026, June 29). Palantir launches engine for deploying NVIDIA Nemotron open models in sovereign environments [Press release]. Business Wire. https://finance.yahoo.com/technology/ai/articles/palantir-launches-engine-deploying-nvidia-105900604.html
Sammons, L. (2026, June 30). Palantir AIP and sovereign government AI [Video interview]. AWS Summit DC. https://www.youtube.com/watch?v=k-zJA0_tO6w
Snowflake. (2026, March 17). Intelligence and interoperability: Data catalog must-haves for AI. Snowflake Blog. https://www.snowflake.com/en/blog/universal-ai-catalog-data-governance/
Woods, V. (2007). Context: A neuroeconomic solution set for the integration of intelligent models [Unpublished MBA thesis].
Disclaimer: The views expressed in this blog are for informational and educational purposes only and do not constitute investment advice. Past performance is not indicative of future results. Please consult a qualified financial professional before making any investment decisions.
We are unable to accept orders via email. If you wish to place an order, please consult your registered representative or contact the home office trading desk at (800) 649-4554.
This email system is for business purposes only and any information, including attachments, transmitted in this email is not confidential. Any message may be reviewed by authorized compliance personnel and/or produced to regulatory agencies or others with a legal right to access such information.
Past investment performance is not indicative of future results. Securities offered through Bolton Global Capital, Inc., Bolton, MA. Member FINRA, SIPC. Advisory services offered through Bolton Global Asset Management, a registered investment advisor, 579 Main St., Bolton, MA 01740 (978) 779-5361.
Page 9 | For informational purposes only. Not investment advice.
Vaughn Woods Wealth Management | Investor Insights
Investors should be aware that there are risks inherent in all investments such as fluctuations in investment principal. Past performance is not a guarantee of future results. Asset allocation cannot assure a profit nor protect against loss. Although the information has been gathered from sources believed to be reliable, it cannot be guaranteed. Views expressed in this newsletter are those of Vaughn Woods and Vaughn Woods Financial Group and may not reflect the views of Bolton Global Capital or Bolton Global Asset Management. The information provided is for general informational purposes only and should not be considered individual recommendation or personalized investment advice. Representatives and Advisors of Vaughn Woods Financial Group are not tax or legal professionals, if you need tax or legal advice, please make sure to consult a tax professional/CPA and/or a lawyer. VW1/VWA0392