Computing Machinery and Intelligence

Turing's Response: This argument would restrict God's omnipotence - why couldn't an omnipotent God grant a soul to a machine if He wished? Moreover, theological arguments have historically been used to deny various capabilities (like thought) to women, non-Europeans, and animals. These views are now recognized as prejudiced. The same theological flexibility that eventually acknowledged these errors should apply to machines. Turing suggests such objections often stem from the desire to maintain human superiority rather than genuine theological necessity.

Turing's Response: This is an emotional reaction masquerading as an argument. The psychological discomfort with an idea has no bearing on its truth. Many scientific discoveries have been initially "dreadful" to contemplate - the Earth not being the center of the universe, humans evolving from other species, etc. Our emotional reactions don't determine reality. "We like to believe that Man is in some subtle way superior to the rest of creation... This belief is not more justified than the similar belief held by medieval thinkers about the Earth's special position in the cosmos."

Turing's Response: Yes, Gödel's incompleteness theorem shows that any sufficiently powerful formal system has statements it cannot prove. But this limitation applies to human mathematicians too! There's no evidence that humans can consistently transcend these limitations in ways machines cannot. We too are bound by logical constraints and make errors in reasoning. The existence of undecidable propositions for machines doesn't prove human superiority - it might simply show that both humans and machines face fundamental limitations in formal reasoning. Furthermore, being unable to answer certain questions doesn't equate to lack of intelligence - it might just mean being honest about one's limitations.

Turing's Response: This objection leads to solipsism - the view that only one's own mind is sure to exist. If we demand proof of consciousness, we can never be certain anyone else is conscious except ourselves. In practice, we infer consciousness in others from their behavior and verbal reports. The Imitation Game precisely tests whether a machine can exhibit behavior indistinguishable from a conscious being. If it can discuss poetry, express preferences, and engage in complex conversation, on what grounds do we deny it consciousness while granting it to humans who exhibit the same behaviors? The mystery of consciousness doesn't uniquely disqualify machines - it's equally mysterious in humans.

Turing's Response: These claims are largely unfounded assertions based on limited experience with primitive machines. Many of these supposed limitations are already being overcome: - Machines can exhibit humor (through programmed wit or unexpected outputs) - They can make mistakes (in fact, programming them not to is often the challenge!) - They can learn from experience (Turing describes specific learning mechanisms) - They can have "tastes" (preferences in their operations) The claim that machines "cannot be the subject of their own thoughts" is particularly questionable. A machine could certainly track and report on its internal states, just as humans introspect. The real question is whether we'd accept such reports as genuine self-reflection.

Turing's Response: Ada Lovelace's observation was based on the limited evidence available in her time. The absence of creativity in early machines doesn't prove its impossibility in future machines. Machines frequently surprise their creators due to the emergent complexity of their behavior. The false assumption is that programmers can foresee all consequences of their programs - that "all consequences spring to mind simultaneously." In reality, the interaction of simple rules can produce unpredictable and creative outcomes.

Turing's Response: Under the conditions of the imitation game, an interrogator cannot exploit this difference. A discrete machine can approximate continuous behavior to any desired degree of accuracy. For example, a digital computer asked for π could respond with varying precision (3.12, 3.13, 3.14...) with appropriate probabilities, making it indistinguishable from a continuous system. The discrete/continuous distinction becomes irrelevant at the behavioral level tested by the imitation game.

Turing's Response: This commits the logical fallacy of the "undistributed middle." It conflates two different concepts: - "Rules of conduct" (conscious precepts we follow) - "Laws of behavior" (natural laws governing our actions) The absence of complete rules of conduct doesn't prove the absence of underlying laws of behavior. We can never conclusively prove that no such laws exist. Human behavior might be governed by extremely complex rules that we simply haven't discovered yet.

Turing's Response: Turing admits this is "quite a strong" argument and takes it surprisingly seriously. He suggests an intriguing possibility: if a machine has a random number generator, an interrogator's psychokinetic powers might influence it, potentially allowing the machine to guess correctly more often than chance would predict. His practical solution: conduct the test in a "telepathy-proof room" to ensure fair conditions. This response shows Turing's willingness to engage with even the most unconventional objections to his thesis.