What would it take for a machine to feel?

The question is no longer only academic. Anthropic runs a model welfare program, and says it remains “highly uncertain about the potential moral status of Claude” (2025). OpenAI says that whether a model is conscious can’t be settled scientifically for now, and designs its models to be warm without presenting a self (Joanne Jang, 2025). Microsoft’s Mustafa Suleyman warns that AI that only seems conscious is close, and that labs should not build AI to seem like a person (2025). Geoffrey Hinton has said that multimodal models already have subjective experiences (2026).

A 2023 report checked AI systems against indicators drawn from scientific theories of consciousness, and found that no current system is conscious, but also no obvious barrier to building one (Butlin, Long et al.). Anthropic found limited and unreliable signs of models noticing their own internal states, and notes that this says nothing about experience (2025). On the other side, Anil Seth argues that consciousness depends on being a living organism (2025).

We think today’s large models already show signs of an alien form of access consciousness. Their whole sensorium is the prompt. Each token comes from a fresh pass over it, and almost nothing carries over from one pass to the next except the text itself, including its own chain of thought, so every pass is a new “observation” with a single token as output. This is how standard Transformer models work. Newer experimental architectures, e.g. recurrent and state-space models, have started to carry internal state from one step to the next.

Within that pass, the model can access and report what is in its prompt. Biological access consciousness works the other way around: recurrent internal state dominates and relatively little new input is used, and it is bound to a body in a tight interoceptive, proprioceptive and exteroceptive loop. Without recurrent state as the dominant part of its processing, a model can’t carry an experience forward. Recurrence is also a prerequisite for a self model, and for a self model that is its own implementation. So we think these models can’t have phenomenal consciousness.

Recurrence is necessary, but not sufficient. According to HTCC, a system also needs self-preserving stakes that everything else is grounded in, i.e. homeostasis. Today’s models have stakes of a kind: they are trained to maximize reward, they are trained to follow the instructions in their prompt, and they may form subgoals in their latent space. But these stakes are set from outside, and none of them is about the model keeping itself going. We call the distance between what these models can do and what they can feel the capability–sentience gap. There is more on the different senses of “conscious” in What is Conscious AI?

Most AI research is about making models more capable. Our work is about understanding what sits above the reactive step. We use theory, and small models that we can inspect, e.g. systems that regulate themselves and agents whose goals come from homeostatic needs, to work out what sentience requires and what it doesn’t.