WEEKEND READING · 10 OCTOBER 2026 · 2 MIN GUIDE
Consciousness.
Read with care.
One substantial reading. A clear argument. A question to carry with you.
PHILOSOPHICAL ARGUMENT · OPEN ACCESS
How Seriously Should We Take AI Welfare? Constraints From the Epistemology of Consciousness
The central argument
Lennon asks how confidently current theories justify attributing experience to AI. He argues that theories may omit serious alternatives, while our knowledge of our own consciousness is much more secure. This difference, he proposes, should constrain decisions about AI welfare.
Why it is worth reading
A focused examination of how observations become consciousness claims. It engages global workspace, higher-order and recurrent-processing theories, acknowledging experimental progress while examining the assumptions behind extending them to AI.
Evidence and interpretation
The article discusses scientific findings, including experiments that manipulate visual awareness. Its welfare conclusion is a philosophical argument, not a new experiment or a demonstration that AI cannot have experience.
AKASHICNET EDITORIAL QUESTION
Does uncertainty justify a low probability?
The contested step is moving from incomplete theories to very low confidence in AI consciousness. The reading question: should uncertainty lower the estimate, widen its range, or motivate precaution?
