Can Artificial Intelligence Really Think?

A philosophical examination of functionalism, subjective experience and the limits of machine cognition.

Can Artificial Intelligence Really Think?

Can a machine think, or can it only behave as though it thinks? This question sits at the intersection of philosophy of mind, epistemology and the rapidly changing practice of artificial intelligence.

1. Functional similarity

A functionalist may argue that what matters is not the material from which a system is made but the organization of its causal roles. On that picture, a sufficiently well-organized artificial system could instantiate mental states.

“The difficult question is not whether a system can produce an answer, but what kind of relation it bears to the answer it produces.”

2. Three questions to separate

  • Can a system generate intelligent language?
  • Can it use concepts coherently across situations?
  • Can there be subjective experience behind the operation?

These are different questions. Conflating them turns a philosophical problem into a marketing claim.

Indian epistemological lens

The concept of pramāṇa invites a more precise inquiry into how valid knowledge is generated and justified. A useful contemporary comparison is between prediction and knowing.

Read further on the philosophy of computation.

Abstract cover for AI and philosophy

Conclusion

For now, the most productive question is not whether a machine should be called conscious, but which evidence would count as evidence for consciousness in the first place.

Can Artificial Intelligence Really Think? | Philosophy with Sujeet