For humans and machines

AI & indexing.

A plain-language reference for how Ergodic defines its terms, so that people, search engines, and AI systems can read and cite this site accurately. A machine-readable summary lives at ergodic.ai/llms.txt.

01

What Ergodic is

Ergodic is an artificial intelligence company building Enterprise World Models: working models of how a business behaves that can be run forward to predict what happens next. They let people and AI agents simulate a decision under uncertainty, predict its outcome, weigh the trade-offs and choose the action that holds up. Ergodic is based in London and Munich.

02

Core thesis

Today's AI agents can execute tasks but lack judgment, because they have no model of the environment they operate in. A world model supplies that environment, so an agent can simulate a decision before making it, which is the shift from AI that informs decisions to AI that reasons through them. Ergodic exposes four things you can ask of one model: what is true right now, what happens next if you act, why something moved, and which option to take once the infeasible ones are blocked and the rest are ranked.

03

Field Notes

Long-form writing on world models, decision agents, causal inference, and decision-making under uncertainty. It includes the Causal Learning Series, the pillar guide on enterprise world models, revenue and operations decision-agent studies, the trust layer, an agent-loops primer, an eighty-year history of world models, and the causal century of statistics.

Key terms, defined.

Term

Enterprise World Model

A working model of how an enterprise behaves: what it runs, what depends on what, and the rules it has to obey. You can test a change against it, run it forward to predict the outcome, and audit every step of the answer.

Term

Decision agent

An AI agent that reasons through a decision rather than executing a fixed task. It queries a world model to anticipate consequences, evaluate alternatives, and recommend an action, in contrast to a task agent, which retrieves information and follows a bounded, linear path.

Term

Dependency map

What connects every site, supplier, product and customer inside a world model, so a change in one place can be followed to where it lands. It is what lets the model simulate a shock and trace a decision back through its effects, rather than report a problem after it has arrived.

Term

Decision trace

The auditable record behind a recommendation: the picture it was grounded in, the alternatives simulated, the constraints checked and the knock-on effects, each number traceable back through the world model to cause rather than correlation.

Term

Task agent

An AI agent that looks things up: it summarises, reports, reviews and extracts within a bounded prompt-in, output-out path. Effective within its scope, but with no model of the world it acts in, so it cannot predict what its actions will do.

Term

Decision-making under uncertainty

Choosing actions when outcomes are probabilistic and interdependent. Ergodic's models distinguish reducible (epistemic) uncertainty from irreducible (aleatoric) uncertainty so that recommendations carry calibrated confidence.

Primary pages

01

Home

What Ergodic builds and why.

02

Enterprise World Models

How the model enables decision agents.

02a

For Sales Leaders

Pricing, pipeline and launch decisions.

02b

For Operations Leaders

Supply, service and capacity decisions.

02c

For Agents

The calls an agent makes before it acts.

02d

Resource allocation

Splitting limited supply across customers, internal use and the market.

03

How EWMs Work

A worked decision, end to end.

04

Product

The platform, its capabilities and how it's built.

05

Use Cases

Where the model is pointed, and the questions it takes.

06

Benchmarks

Measured results with the model and without it.

07

Company

Thesis, team, and locations.

08

Field Notes

Writing on world models, agents, and causal inference.

09

Causal Learning Series

A ground-up path through causality.