The world keeps changing. Your analysis should too.

Hinsley combines expert judgment with continuous AI monitoring to keep scenarios, signals, and assumptions current - so major decisions can be weighed against multiple plausible futures, not a single aging forecast.

See how it works Talk to us

Used by practitioners across leading intelligence organizations, think tanks, and geopolitical teams - and continuously refined with their expertise.

The Problem

A new operating model for strategic foresight.

Everything else in your business is measured continuously. The decisions that matter most are still made on a phone call and a hunch - and nobody notices when the ground shifts underneath them.

Hinsley makes foresight continuous. Forecasts of the critical indicators update as new evidence arrives, refreshing both the likelihood of alternative outcomes and the analysis that explains how events are unfolding.

What teams reach for today

Expert calls and offsites

Valuable judgment in the moment, but no durable record, score, or update cycle.

Briefings and syndicated research

Timely and expert, but produced for someone else's market and already aging when it arrives.

Dashboards and predictive analytics

Excellent where data is clean and the future resembles the past. Strategic uncertainty rarely does.

AI chatbots

Fast and flexible, but ad hoc: no persistent model, shared memory, or accountable history.

Trusted in consequential environments

Trusted by leading organizations in government, defense, intelligence, and capital markets

I wish I had had access to a tool like Hinsley when I was serving as a State Department diplomat and policymaker.

Adoption of Hinsley by national security and diplomatic agencies would significantly enhance analysis and decisionmaking. Hinsley can enforce rigorous methodologies and provide forecasting capabilities beyond current training. This would improve the clarity of U.S. Government analysis and the efficacy of policy decisions.

Geoff Odlum, Former U.S. State Department Foreign Service Officer and Acting Deputy Assistant Secretary of State

Enterprise Foundation

Two decades of operating where security and trust are non-negotiable.

Cultivate Labs builds and operates forecasting applications for institutions with exacting security, deployment, and accountability requirements.

Explore enterprise security and deployment

Shared to air-gapped

Deployment options

Choose our shared cloud, an isolated environment, or a fully air-gapped deployment for classified environments.

SOC 2 Type II

Independently audited

Cultivate Labs and its applications undergo a SOC 2 Type II audit every year.

U.S. + Europe

Top-secret deployments

Our applications have operated in top-secret environments on both continents. Our work spans intelligence agencies, think tanks, NGOs, and some of the world's largest private-sector companies.

Since 2006

20 years of forecasting programs

The work began with Inkling Markets at Y Combinator in 2006 and evolved through two decades of forecasting programs into the applications we operate today.

The Methodology

From a rough question to a living assessment.

Cultivate's Continuous Probabilistic Foresight is the method behind Hinsley, shaped through two decades of work with senior leaders in intelligence organizations, think tanks, and other high-stakes institutions. It turns uncertainty into a rigorous assessment that can be challenged, updated, and used to inform decisions.

Frame the question

Turn a rough topic into a question tied to a real decision. Map the evidence, actors, disagreements, and gaps before drawing a conclusion.

Structure the problem

Break the question into drivers and signals. Build genuinely different scenarios, then turn the most useful signals into forecast questions with clear outcomes and deadlines.

Forecast, assess, and challenge

Bring human and AI forecasts together while preserving each estimate and its reasoning. Draft the assessment, make uncertainty visible, and challenge it before release.

Keep the assessment live

Publish a versioned assessment, monitor for new evidence, re-forecast, and revise the view when the facts warrant it.

New evidence reopens the right part of the analysis, so the assessment changes when the facts do.

Our Design Principles

AI drafts. Humans decide.

AI can research, forecast, and draft. People review, revise, override, and decide what is published or acted on.

Every output shows its work.

Sources, forecasts, rationales, alternatives, revisions, and outcomes remain labeled and traceable, creating a measurable record of what changed and what proved reliable.

Uncertainty stays visible.

Hinsley does not manufacture certainty. Assumptions, alternatives, confidence, and weak evidence remain explicit.

Intellectual foundations. Hinsley's foundations include the work of Richards Heuer and Philip Tetlock; IARPA's ACE program and the Good Judgment Project; RAND scenario analysis; and the analytic standards in ICD 203 and UK PHIA.

Human + AI Forecasting

The engine that keeps your foresight continuous.

Hinsley keeps critical indicators active by combining machine-scale coverage with structured human judgment. Run AI continuously, return to a standing forecaster network on a cadence, or launch a focused campaign when a decision needs targeted input.

Always on

AI Forecasting Agent

Research agents scan open-source and proprietary information on your schedule. Multiple frontier models independently weigh base rates, current evidence, and time remaining, then combine into one probability estimate with a written rationale. As the evidence changes, the forecast runs again.

Research current evidence

Forecast independently

Aggregate and revise

Explore AI Forecasting

Persistent participation

Forecaster Networks

Publish questions to a dedicated site for your existing network of experts and forecasters. Request one-time or recurring forecasts, route them by topic or assignment, capture rationales, and track accuracy over time.

Recurring requests
Written rationales
Accuracy tracking

Targeted collection

Forecasting Campaigns

Create a focused request for human input on one analysis. Invite selected participants, open it to account members, or share a link to collect forecasts on specific questions and likelihood assessments across a scenario set.

Question forecasts
Scenario likelihoods
Flexible access

A Live Analysis, End to End

From one strategic question to a living outlook.

EXAMPLE: What will define the global energy demand and policy environment by 2035?

A real Hinsley workspace connects the research brief, alternative futures, observable indicators, and recurring forecasts in one auditable analysis.

Scenario Trends

Four futures, tracked continuously

The scenarios stay visible as their likelihoods are revisited each week.

For illustration purposes only.

Open Electric Surge

17%

Electricity-led · Open

Fortress Electrification

36%

Electricity-led · Fragmented

Hydrocarbon Buffer World

15%

Mixed-carrier · Open

Fragmented Fossil Retrenchment

32%

Mixed-carrier · Fragmented

Continuous Forecasting

Critical signals, updated weekly

Will global data-center electricity demand in 2030 be at least double its 2025 level?

48%

estimated probability of Yes

+1%

past 7 days

Estimated weekly history based on public projections and observed developments. The latest value is our current forecast.

What changed this week

Strong electricity-demand growth and continued data-center expansion nudged the estimate upward. The IEA's updated outlook places 2030 demand near 950 TWh, just below the 970 TWh threshold, while grid connections, equipment shortages, and rapid efficiency gains constrain the upside.

Our 48% estimate treats a doubling as genuinely close, but not the base case. The central outlook misses by about 20 TWh. Faster adoption of energy-intensive AI workloads could clear the threshold, while infrastructure bottlenecks or faster efficiency gains would keep demand below it.

2025 baseline: 485 TWh
2030 threshold: 970 TWh
IEA central outlook: 950 TWh

Additional indicators being continuously forecasted

How fast demand electrifies
Whether new load growth is exceptional or manageable
Whether grids can keep up
Whether flexibility scales in time
Whether policymakers back firm clean power
Whether gas remains the dominant insurance technology
Whether coal really exits or politically persists
Whether consumers tolerate the transition
Whether supply chains stay globalized or fragment

The forecast is not a separate dashboard. It is part of the assessment.

Research, scenarios, indicators, forecasts, and revisions stay connected, so leaders can see both the current view and why it changed.

Where It Fits

Three ways teams are putting Hinsley to work.

Geopolitical analysis

Maintain a living assessment of conflicts, alliances, policy shifts, and escalation risks as new evidence changes the outlook.

Escalation pathways and actor behavior
Policy shifts and leading indicators

Corporate strategic foresight

Track alternative market, regulatory, technology, and competitive futures so strategy stays current between planning cycles.

Market and regulatory futures
Technology and competitive shifts

Portfolio & program lifecycle

Revisit long-term capital, program, and portfolio assumptions against changing scenarios before commitments drift out of date.

Capital exposure and program assumptions
Milestones that trigger reassessment

Have a priority question in mind? We'll build your first scenario and indicator set, on us.

Start with a complimentary trial centered on the question that matters most. We'll help structure the workspace so your team can experience Hinsley with a real use case before making a commitment.

Start a free trial