Introduction
Policy is prediction.
Every new regulation, budget allocation, or public intervention is a bet on the future—a forecast about how people, institutions, and markets will respond. But what if we're making these bets nearly blind?
In today's world, decisions are made in a landscape shaped by accelerating complexity: climate change, digital infrastructure, geopolitical instability, and growing inequality. Yet the tools available to policymakers haven't kept up. Most are designed for a world that no longer exists—linear, siloed, and slow.
We're building something different: Causal AI tools that help policymakers see, simulate, and shape the future with clarity. These tools won't eliminate uncertainty—but they will help us reason about it better.
What is the problem?
Imagine trying to navigate a maze blindfolded. That's kind of what it's like when policymakers try to anticipate how their policies will play out. Traditional methods often fall short, leading to:
- Reactive Governance: Dealing with problems after they happen instead of preventing them.
- Misaligned Incentives: Policies unintentionally causing the opposite of what they intended.
- Systemic Fragility: The overall system is becoming more unstable.
In short, we need better tools to understand the complex web of cause and effect in our socio-economic systems.
What Does This Mean in Practice?
We're working on two main components:
-
Policy Synthesis (Generation): Using AI to create policy ideas based on specific goals and limitations. For example, if the goal is to reduce carbon emissions, the AI could generate a list of potential policies.
-
Policy Impact Analysis: Evaluating those policies through simulations and data analysis. We can see how the policies might affect different groups of people, industries, and the economy as a whole.
Additional methodology explanation: Making Policy Impacts More Predictable
We define the above thesis in the following systems structure:
1. Complex Systems
Thesis: Any social system is inherently complex. The numerical modeling could not capture the interaction between actors, network effect, as well as the emergent properties within the system. Agent-based simulation goes from bottom up and able to simulate these. Human could only reason with such system limitedly but computation technology empower us.
2. Epistemic Systems
Anti-thesis: Even if we have all the right tools we need, scalable epistemic systems, human bounded rationality, we can only scale ~150 human relationships.
3. Causal Systems
Synthesis: It's difficult to trace cause and effect (this is both difficult for humans and AI systems). A policy decision may have indirect influences on markets, but we lack a market-based approach to understand those policy changes. Those decisions can also be "costly signals", and have adverse effects.
Causal AI involves a shift in perspective by asking (what-if and why) and finding answers that measure the effect of treatment variables, going beyond the classic machine learning prediction.
Conclusion
We don't just need smarter policies. We need smarter ways to make policy.
Causal AI won't replace human judgment—but it will augment it. It helps us map the maze before we step inside. By generating ideas, forecasting impact, and simulating unintended consequences, we can move from reactive governance to anticipatory stewardship.
In a world where every decision sends ripples through fragile, interconnected systems, we can no longer afford to guess. We must model before we act, and ask better questions before we answer them.
The future doesn't just need better outcomes. It needs better foresight.