Many people may come to believe that the AI systems they talk to are conscious. Knowing whether that can be true matters for deep moral and social questions, and for not being misled by systems that only seem conscious.
What we know
There is no recognized method to test consciousness in an artificial system.
What blocks us
The answer depends on the unsolved question of human consciousness.
Current leads
Indicator checklists drawn from scientific theories of consciousness.
Sub-questions
Is consciousness only a matter of performing the right computations, whatever the material the system is made of?
Why can't we just test an AI's behavior to find out if it is conscious, when it may be trained to imitate humans?
How can we avoid believing that AI systems are conscious when they only seem to be?