The AGI Index: Measuring Hyperbolic Changes as a True Indicator of AGI Progress

Introduction

The AGI (Artificial General Intelligence) Index serves as a quantitative metric to evaluate the advancement toward AGI, focusing on hyperbolic changes in user interactions required to complete tasks. Unlike metrics that highlight incremental improvements, the AGI Index is designed to capture drastic reductions in interactions, which are the true indicators of AGI. Using the scenario of planning a vacation to Walt Disney World, this article demonstrates how the AGI Index can tangibly measure these hyperbolic changes in the context of Intelligent Assistants (IAs) like Lulu.

Defining the AGI Index

The AGI Index is calculated using the formula:

A near-zero AGI Index signifies a hyperbolic reduction in human interactions, indicating significant progress toward AGI.

Why Hyperbolic Changes Matter

In the realm of AGI, incremental improvements are inconsequential. Only hyperbolic changes of 100X or more in interaction counts serve as credible evidence of AGI progress. Therefore, the AGI Index becomes meaningful only when it reflects these massive reductions in required human interactions.

Case Study: Frank and Lulu Plan a Walt Disney World Vacation

Pre-IA Scenario

In a pre-IA world, Frank would engage in hundreds of interactions for tasks such as:

1. Hotel Reservations: Multiple steps to register, browse, and book.
2. Ticket Purchases: Navigating through various options and payment steps.
3. Airline Bookings: Comparing flights, selecting seats, and confirming.
4. Taxi Services: Setting pickup and drop-off locations, and choosing ride options.

Each of these steps contributes to an exceedingly high AGI Index.

IA-Enabled Scenario

With Lulu, Frank’s IA, the paradigm shifts drastically:

1. Data Reception: Entities like Walt Disney World, Delta Airlines, and Uber provide their data to Lulu.
2. Task Execution: Lulu processes this data and autonomously completes the tasks based on Frank’s pre-set preferences.

Frank’s interaction count drops to near zero, leading to a hyperbolic change in the AGI Index.

Quantifying Hyperbolic Changes

The AGI Index for Frank before and after adopting Lulu illustrates the hyperbolic change:

The difference is not incremental; it’s a hyperbolic shift of more than 100X, thus serving as a tangible indicator of AGI progress.

Conclusion

The AGI Index offers a quantifiable measure of hyperbolic changes in human-IA interactions, serving as a true indicator of AGI progress. Incremental improvements are irrelevant in this context; only drastic reductions in the AGI Index signify meaningful advancements toward AGI. The case of Frank and Lulu exemplifies this, highlighting the transformative power of IAs in reducing human interactions to near-zero levels.