Calculation Cannot Generate Consciousness: AI Mimics Empathy but Lacks a Soul
Source: https://www.facebook.com/americanspectator/. "AI Can Fake Empathy. Can It Fake a Soul? | The American Spectator | USA News and Politics." October 18, 2026. spectator.org
The Gist
The author argues that AI only imitates feelings by crunching language statistics, and people are easily fooled into thinking it is alive, as happened with the 1960s ELIZA program. Drawing on Penrose, Gödel, and Turing, he claims real consciousness and moral judgment can't be reduced to calculation, and he sees it as a divine gift. Thinking we can build a conscious, moral machine is, in his view, arrogant and dangerous.
Conclusion
Artificial intelligence, however convincingly it simulates empathy, is not and cannot become conscious, because consciousness (the 'image of God') transcends algorithmic calculation; therefore, attempts to build conscious, morally autonomous AI are a dangerous form of hubris.
Premises
- People readily attribute consciousness and emotion to machines that merely mimic empathy (the 'Eliza effect'), even when they know it is a machine.
- Large language models only aggregate language and apply calculable probability rules, so they do not feel anything.
- Following Penrose, Turing, and Gödel, any formal computational system has intrinsic limits and cannot generate truths or insights that transcend its own rules; no master algorithm exists.
- Consciousness, emotion, and personal identity involve a self-aware act of making a coherent whole (like seeing a Seurat painting as a whole rather than as dots) that cannot be fully programmed or calculated.
- This integrating self-awareness is what Genesis calls the 'image of God,' and its coherence depends on moral choices, which also have no algorithm because moral discernment transcends calculation.
- AI developers who claim they can build an algorithm for morality and conscious, autonomous AI are repeating the ancient idolatrous impulse to arrogate divinity to themselves, which is dangerous without discernment.
Assumptions
- Human consciousness is non-algorithmic, as Penrose argues, and Penrose's controversial interpretation of Gödel and Turing is correct.
- Gödel's incompleteness and Turing's limits on computation apply to consciousness and emotion in a way that rules out machine consciousness.
- Emotion and consciousness are the kind of thing that 'transcends the system' in the same way unprovable mathematical truths do.
- Humans themselves are not merely complex computational systems subject to the same limits.
- A religious framework (Genesis, the image of God) is a valid explanatory basis for what consciousness is.
- Simulating empathy is not the same as having it, and behavioral evidence cannot establish consciousness.
- AI developers genuinely claim to be building conscious, autonomous, morally algorithmic systems.