Ember

Ember is an innovative public AI prediction engine designed to provide transparent and trustworthy insights into prediction markets. Built on the principle that an AI's reasoning process is essential for credibility, Ember operates daily with three distinct AI models: Claude from Anthropic, Grok from xAI, and Gemini from Google. Each model independently evaluates live Polymarket markets every morning at 7 AM EST, without consulting each other. When these models diverge by 10 or more points from the crowd, it signals a potential high-conviction opportunity. This divergence is meticulously timestamped, ensuring that every prediction is locked in before the outcome is known. By tracking accuracy through Brier scores and conducting post-mortems on all calls, Ember builds a public record that users can rely on for informed betting decisions.

Transparent Prediction Calls

Ember ensures that every prediction made by its AI models is documented and locked before outcomes are known. This level of transparency fosters trust among users, as they can see the reasoning behind each call without any edits or deletions.

Multiple AI Models

Ember utilizes three independent AI models, each with its own unique strengths. Claude synthesizes information through first principles, Grok leverages real-time sentiment analysis, and Gemini verifies against live search results. This diversity allows for a well-rounded evaluation of prediction markets.

High-Conviction Signals

The platform flags divergences of 10 or more points from the crowd as high-conviction signals, indicating where the crowd may be mispricing an outcome. This feature provides users with actionable insights that can enhance their betting strategies.

Comprehensive Data Integration

Ember synthesizes information from multiple sources, including prediction markets, bookmaker lines, and AI research feeds. By analyzing data from 20 different sources, it ensures a robust and informed basis for predictions, making it a powerful tool for bettors.

Sports Betting

Bettors can leverage Ember's predictive insights to make more informed decisions in sports betting. By analyzing divergences in predictions from the crowd, users can identify potential undervalued bets and capitalize on them.

Market Predictions

Investors and traders can use Ember to gain an edge in predicting market movements. The independent evaluations of the AI models provide a clearer picture of potential outcomes, allowing users to make more strategic investment choices.

Event Outcomes

Ember is valuable for predicting the outcomes of significant events, such as political elections or major corporate announcements. Users can access high-conviction signals that highlight where the market may be misaligned with actual probabilities.

Research and Analysis

Researchers and analysts can utilize Ember's data and predictions to study trends in prediction markets and AI behavior. The comprehensive public record of predictions and outcomes offers a wealth of information for academic and practical applications.

How does Ember ensure the accuracy of its predictions?

Ember tracks the accuracy of its predictions using Brier scores, a calibration metric that rewards both accuracy and confidence. Each prediction is timestamped, with a thorough post-mortem conducted on every call to analyze performance.

What is the significance of the 10-point divergence?

A divergence of 10 or more points from the crowd indicates a potential high-conviction signal. This means that either the crowd is incorrect, or Ember's models have identified an opportunity, providing users with actionable insights.

How often are predictions made?

Ember makes predictions every day at 7 AM EST. The insights are available to subscribers before they are released to the public, allowing early access to valuable information.

Can I track the performance of Ember's predictions?

Yes, Ember maintains a public record of all predictions, including timestamps and outcomes. This transparency allows users to track the performance of the AI models over time and assess their reliability.