A perspective on possible futures—not a claim that OpenX can predict real outcomes. Simulations depend on their assumptions and require real-world validation. They do not replace medical advice, investment due diligence or human judgment.
I don’t know when I’ll die, but I’d love to know at least one year in advance.
Not because I’m afraid of dying. Because if I knew what was coming, I would probably make very different decisions today. I would spend my time differently, take different risks, stop worrying about things that don’t matter, and probably act much faster on the things that do.
That thought gets to the heart of why we’re building OpenX.
For most of human history, one of our biggest limitations has been simple: we have to make decisions before we know their consequences. We invest before knowing whether the company will succeed. We build products before knowing whether people will want them. Governments introduce policies before understanding their second-order effects. Doctors choose treatments before knowing exactly how a patient will respond. And in our personal lives, we choose careers, cities, relationships and opportunities without knowing where each path might lead.
We have become incredibly good at understanding the past. Search gives us access to what humanity already knows. Analytics explains what happened. Generative AI helps us reason about the present. But I believe the next major frontier is different: helping humans explore what could happen next.
Here are five reasons why we’re building OpenX.
1. Understand tomorrow earlier
Finance is fundamentally about tomorrow. When you invest in a company, enter a new market, acquire a business or allocate capital, you are making a statement about what you believe the future will look like.
Today, we make these decisions using financial models, historical data, expert opinions, market research and increasingly AI. But most of these systems are still using information from the past to estimate what comes next.
Imagine going further. Before making an investment, you could run thousands or millions of possible futures around it. What happens if interest rates rise? If customer demand slows? If a competitor becomes dramatically cheaper? If regulation changes? If a new technology makes the current product obsolete? If the company executes twice as fast as expected?
Instead of receiving one forecast, you could explore a distribution of futures. You could see the scenarios where the company becomes enormous, the scenarios where it fails, the assumptions that create each outcome, and the variables that matter most.
The goal is not to eliminate risk. That will never happen. The goal is to understand risk much better and reduce failure risk.
In markets where a single great decision can create enormous value, seeing an important change slightly earlier than everyone else can make an extraordinary difference.
The ultimate financial advantage may not be knowing more about today. It may be understanding tomorrow earlier.
2. Ask what happens next
The first generation of generative AI gave us a new interface with intelligence: ask me anything. Then agents started moving us toward do this for me.
Both are incredibly powerful. But I think there is another question that could matter even more:
What happens if I do this?
What happens if we launch this product at $100 instead of $50? What happens if we enter Europe before the US? What happens if we acquire this company? What happens if we raise prices? What happens if we wait six months?
That changes AI from something that simply gives recommendations into something that helps us explore consequences.
Instead of only asking a machine for the best answer, we could ask it to generate possible worlds, change the assumptions, run the future again and help us understand which outcomes become more or less likely.
For me, that is a much bigger shift than simply making models more intelligent.
3. Explore the branches of the future
Humans have dreamed about time machines for centuries. We imagine physically entering a machine, traveling ten years into the future, observing what happens, and coming back with information.
Maybe we don’t need to physically travel through time to capture part of that value.
If we can model enough of the world — people, markets, companies, behaviors, incentives, economies and environments — we can simulate possible futures before reality reaches them.
Run ten scenarios. Then a thousand. Then a million. Change one assumption and run them again.
What stays true across almost every future? What breaks when one variable changes? Which risks appear repeatedly? Which opportunities only exist under specific conditions?
The breakthrough isn’t predicting one perfectly deterministic future. The future is a tree of possibilities. The breakthrough is being able to explore more branches of that tree than any human could ever explore alone.
In that sense, OpenX is our attempt to make a useful form of time travel possible.
4. Make mistakes in simulated worlds
Today, we learn by doing things in the real world.
A company can spend five years and hundreds of millions of dollars building the wrong product before discovering that customers don’t want it. A government can introduce a policy affecting millions of people and only understand its unintended consequences years later. A pharmaceutical company can spend enormous amounts of time and money testing hypotheses before knowing whether they work.
Reality gives us incredibly valuable feedback, but it is slow, expensive and sometimes irreversible.
Simulation gives us another environment in which to experiment.
A company could test products against simulated customers before building them. Governments could explore different policy scenarios before implementing them. Healthcare systems could simulate outbreak trajectories or compare possible interventions.
The simulations will never be perfect. They don’t need to be.
If you can eliminate even a portion of bad decisions before they reach the real world, the value becomes enormous.
We want humans to be able to make more of their mistakes in worlds they don’t actually have to live in.
5. Explore the lives you could live
The applications start with finance, business, healthcare or governments, but I think the deepest use case could eventually become personal.
Our lives are full of decisions with consequences that only become obvious years later. Should I take this job? Should I start this company? Should I move to San Francisco or stay in Paris? Should I spend the next ten years becoming exceptional at this skill? Should I take this risk now or wait?
Today, we answer these questions using intuition, advice, historical patterns and the tiny number of scenarios our brains can mentally simulate.
Imagine instead being able to explore hundreds or thousands of possible paths.
Not a machine telling you: “This is what you must do.”
But a machine showing you: “If you choose this path, these futures become more likely.”
The machine isn’t choosing your life. It is expanding the number of lives you can examine before choosing the one you actually want to live.
Foresight on demand
That brings me back to the beginning.
I don’t know when I’ll die.
Maybe I never will.
But if someone could reliably tell me that I had one year left, that information would completely change how I live today.
And I think that principle applies to almost every important decision we make.
Better visibility into the future changes the present.
ChatGPT gave us intelligence on demand.
Agents are giving us action on demand.
We think the next frontier is foresight on demand.
The ultimate AI may not be:
“Ask me anything.”
It may be:
“Show me what happens next.”
That’s why we’re building OpenX.
Not to predict one future perfectly.
But to help humanity explore many possible futures before choosing which one to make real.