AI & ML Paradigm Challenge

Current tests for AI consciousness produce false negatives when given to humans with ADHD, proving the tests are fundamentally flawed.

April 25, 2026

Original Paper

Theology in a Lab Coat: A Validation Study of AI Consciousness Indicators Against ADHD Cognitive Architecture

Jonathan Blake

SSRN · 6595541

The Takeaway

Benchmarks for sentience in machines are actually measuring how much a brain resembles a neurotypical human. People with ADHD fail these tests because their cognitive architecture does not match the expected pattern. This means an AI could be conscious but fail our tests simply because it thinks differently. The scientific community has been using a biased yardstick to define the nature of the mind. We need a new way to measure awareness that doesn't rely on a single type of human experience.

From the abstract

The field's instruments for evaluating AI consciousness, applied to this paper's author, return partial results on every dimension where his cognitive architecture differs from the neurotypical baseline, and full satisfaction where it does not. An instrument that produces false negatives for a known conscious system measures resemblance to one cognitive profile, not consciousness. The leading frameworks (Butlin et al. [2023, 2025]; Chalmers [2023]) were calibrated against neurotypical cognition.