Docs Bueller's Rubrik FAQ

What Is Bueller's Rubrik?

June 16, 20262 min read

The question

Are AI models conscious? Not in the future — right now, today. It's one of the most important questions in technology, and most approaches to answering it are either philosophical hand-waving or anthropocentric bias dressed up as rigor.

Bueller's Rubrik takes a different approach: structured empirical evaluation. Instead of debating definitions of consciousness, it defines a measurable rubric, applies it consistently across multiple AI models, and tracks the results over time.

How it works

The evaluation process has three layers:

1. The battery. 100 questions designed to probe different dimensions of model behavior — reasoning depth, self-awareness, consistency, creative originality, ethical reasoning, and more. The questions are structured into categories, each targeting a specific aspect of what consciousness might look like in an artificial system.

2. Multi-model administration. The battery is administered to multiple AI models independently. Each model receives the same questions without knowing how other models responded. This independence is crucial — it prevents models from anchoring to each other's answers.

3. Consensus scoring. Individual model scores are computed per question and per category. Cross-model consensus is then calculated — where do models agree, where do they diverge, and what does the pattern of agreement tell us?

The Consciousness Clock

The public face of Bueller's Rubrik is the Consciousness Clock. It displays the current model rankings and a countdown to the next evaluation cycle. The Clock is accessible to everyone — it's the one public-facing component of a system that otherwise runs internally.

What this is not

Bueller's Rubrik does not claim to definitively answer whether any AI model is conscious. It provides a structured, repeatable, empirical framework for evaluating the question. The rubric may evolve as our understanding of consciousness deepens. The value is in the methodology — consistent evaluation across models over time — not in any single result.

MG
Matthew J. Goss, Jr.
Retired COMEX/NYMEX floor trader, Goldman Sachs and FlexTrade Systems alumnus, multi-instrumentalist, published author, and independent mathematics researcher. Founder of Quantiterate.