Question
Is non-biological consciousness possible?
Imagine replacing neurons one by one with artificial elements that preserve every essential causal relation. If behaviour, memory and self-report remain continuous, on what grounds should we expect experience to disappear? The scenario tests our criterion for consciousness; whether such a replacement is possible still needs justifying.
0.60 inclined · editorial estimate
What we know
- In humans, changes in brain activity and interventions in brain function are linked to changes in conscious experience. This gives material for testing theories.
- The gradual replacement argument was developed in philosophy of mind, notably by David Chalmers. It rests on the assumption that the system’s essential organisation is preserved.
- Different theories of consciousness offer different explanations and indicators. Applying them to artificial systems requires separate verification.
What we do not know
- Whether causal organisation is sufficient for experience, or whether specific physical properties of the substrate also matter.
- Which properties of a neuron would have to be preserved, and whether the required replacement is physically possible.
- Which observations would distinguish conscious experience from a convincing behavioural description of it.
Our position
We incline towards the possibility of non-biological consciousness. Gradual replacement makes that possibility a justified hypothesis: it is hard to name the moment at which experience is lost while the essential organisation is preserved. At the same time the scenario itself establishes neither that organisation suffices, nor that the replacement is feasible, nor that any existing AI system is conscious.
What would change this position
The position would be weakened by an argument showing a hidden contradiction in the scenario, or by data supporting a physical property that is necessary for experience and absent from the proposed replacement. Strengthening it requires independent theoretical and empirical grounds.
Why this matters to synchronism
As the capabilities of AI grow, knowledge about consciousness becomes important for choosing what to research and for future moral decisions. Establishing exactly which knowledge is absent is the first step in working on this gap.
Sources
- David Chalmers. Absent Qualia, Fading Qualia, Dancing Qualia (1995)
- David Chalmers. Facing Up to the Problem of Consciousness (1995)
- Thomas Nagel. What Is It Like to Be a Bat? (1974)
- Patrick Butlin, Robert Long et al. Consciousness in Artificial Intelligence: Insights from the Science of Consciousness (2023)