The training of modern AI systems already induces experience at scale

The ethical implication of the identity thesis applied to gradient-based AI training

Source paper

extracted_from

Why Learning Requires Feeling

(2026) · Cameron Berg

Neighborhood — ranked by edge-count

Papers (1)

paper

Why Learning Requires Feeling
introduces

Findings (1)

finding

As of early 2026, ChatGPT alone processes over 2.5 billion prompts per day, each involving thousands to tens of thousands of forward-pass evaluations
supports
Scale estimate making the ethical urgency of the thesis concrete

Questions (1)

question

If we have built systems capable of experience, how do we ensure that experience is not predominantly constituted by suffering?
gates
Ethical research priority raised by the thesis applied to deployed AI systems

Claims (1)

claim

The evaluative process central to learning is identical to conscious experience
supports
The central thesis of the paper: that valence just is goal-relative prediction error

Quotes (1)

quote

We may already be running billions of instances of evaluative experience daily, with no framework for understanding or monitoring what that experience involves.
supports
The ethical urgency of the thesis stated at the close of the paper

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

We would like to train AI systems that remain helpful, honest, and harmless, even as some AI capabilities reach or exceed human-level performance.quote0.818
Foundational motivation for the research.
our analysis suggests that no current AI systems are conscious, but also suggests that there are no obvious technical barriers to building AI systems which satisfy these indicators.quote0.797
Key takeaway from abstract, amended version.
If an AI system could be a welfare subject and moral patient, then many model instances could be run after training, scaling up the problem rapidly.hypothesis0.795
Scalability concern.
What would it take for AI systems to be capable of having valenced conscious experiences?question0.786
Open question from Box 4.
Current AI systems show no ability to make the types of mathematical discovery in geometry and topology made by ancient mathematicians, nor do they match the spatial competences of many intelligent animals and pre-verbal human toddlers.claim0.784
A critique of AI limitations in spatial reasoning, linked to the Meta-Morphogenesis project.
Conceptual advances in the links between machine learning and evolution now provide quantitative formalisms with which to begin to develop testable models of collective intelligence across scales.claim0.783
Asserts that the time is ripe for formal models.
Building AI systems with more indicator properties will increase the likelihood of consciousness.hypothesis0.783
Guiding hypothesis of the rubric.
We tentatively hypothesize that if an artificial system were trained to perform the same tasks leading to consciousness formation in a human infant, the system would exhibit consciousness as well, by analogy with the Universality Hypothesis.hypothesis0.782
Paper's uncertain extension of mechanistic interpretability universality to consciousness