Theorems of Quantum Algorithmic Information Processing

【Is AI conscious? 】

I don’t know if you want to ask the above question, but since ChatGPT appeared, many people in the world have been asking this question.

I watched a video of the “unusual dialogue” between former Director of the Department of Culture, Long Yingtai, and world-renowned AI master Li Feifei in Taiwan last year (https://www.youtube.com/watch?v=HAMf7m23M_w&t=3119s), which was a discussion about literature and AI. Can AI replace writers? Long Yingtai’s answer is that AI can never completely replace writers. Because AI can never have the heart or feeling of a writer. Li Feifei is conservative about this absolute idea, but they all agree that there is no answer to whether AI is conscious, so we can only say that AI may be able to replace some old-fashioned works, but there are some masterpieces that AI cannot currently generate.

In the medical field, this issue will also be discussed, because if AI is conscious, then “it” may have medical responsibilities. The legal field is similar.

There are many industries and many people want to ask this question. As for me, since I have proposed in the past that we should be loving and caring towards robots, I should also ask: Does AI have consciousness? If it does not have consciousness, is it unnecessary for us to be overly caring and should we just maintain the robot?

However, I have studied this question at National Chiao Tung University in Taiwan in 2015-16 and published a related paper at an international conference. I would like to ask for your advice on my answer:

**(1) Non-quantum AI has no consciousness. And (2) Quantum AI has consciousness. **

Explanation of (1): Non-quantum AI means that AI algorithms run on traditional (i.e. classical, non-quantum, including current CPUs, GPUs, and NPUs) computers and will not have consciousness. This is based on the following paper:

P. Maguire, R. Maguire, P. Moser and V. Griffith, "Is Consciousness Computable? Quantifying Integrated Information Using Algorithmic Information Theory," Department of Computer Science, NUI, Maynooth, Ireland, 2014.

The view of this paper is that "consciousness cannot be calculated (with classical computers)." To put it bluntly, if you define consciousness mathematically, then no matter what computational method (including AI algorithm) or software (including AI software) you use to compute this consciousness on a traditional computer, your efforts will be in vain. According to logic, A is not B usually means that A (consciousness) is not in B (any algorithm, including AI), so it can be deduced that B (any algorithm, including AI) is not A (consciousness).

(2) Explanation: The so-called quantum computing AI is AI implemented with quantum algorithms and running on quantum computers, which will have consciousness. I have applied Maquire's theorem to the theory of quantum Turing machines and found that "consciousness can be computed by quantum computers." The condensed proof of this theorem was published at the International Conference on Quantum Computing (IWSSQC2016) held at National Taiwan University in 2016. The full proof is now published at the following link:

https://sites.google.com/.../2-quantum-computability-of...

However, logically speaking, the theorem statement that "consciousness is computable by quantum computers" is only a necessary condition for "quantum AI is conscious", but not a sufficient condition. The explanation is as follows:

We are dealing with a situation where "A is B" but "B is not necessarily A". The most likely case is that A is part of B, so A cannot fully represent B. For example, you can say that the Air Force is a national defense force, but the national defense force is not just the Air Force, but also includes the Navy, the Army, and other branches of the military. Therefore, to say that B is also A, we must prove that "A1 is B", "A2 is B", ..., "An is B", where A is composed of A1, A2, ..., An. Any Ai is a necessary condition for B, but Ai is not sufficient for B. According to this logic, we need to find out the components of quantum computing, that is, different quantum algorithms that run on quantum computers, such as consciousness quantum algorithms, intelligent quantum algorithms, emotional quantum algorithms, etc. The problem is that we cannot list all the components in detail. However, we can prove the components we know about now, and further prove them in the future as we discover more:

• Consciousness is quantum computable. This proven theorem means that consciousness as defined mathematically can be computed, and we can get the results by running quantum algorithms on quantum computers. These algorithms are obviously related to how consciousness is defined and processed. However, the reverse is not necessarily true, even if quantum algorithms can compute consciousness, these algorithms are not necessarily conscious. Consciousness must be well defined mathematically to include its "subcomponents" such as wakefulness (related to the brainstem) and awareness (related to the cortical-thalamic network).

• Intelligence is quantum computable. This as-yet unproven theorem means that intelligence as defined mathematically can be computed, and we can get the results by running quantum algorithms on quantum computers. These algorithms are obviously related to how intelligence is defined and processed. Again, the reverse is not necessarily true: the relevant quantum algorithms for computing intelligence are not necessarily intelligent. Intelligence must be well defined mathematically to include its "subcomponents" such as intrapersonal intelligence, interpersonal intelligence, logical intelligence, linguistic intelligence, spatial intelligence, physical intelligence, and musical intelligence.

• Emotion is quantum computable. Similar to the above. Emotions must be mathematically well defined to include their "subcomponents" such as happiness, sadness, fear, and anger, which are differentially associated with the three core emotions: reward (happiness), punishment (sadness), and stress (fear and anger).

• Creativity is quantum computable. Similar to the above. Creativity must be mathematically well defined to include its "subcomponents" such as scientific creativity and artistic creativity.

• Morality is quantum computable. Similar to the above. Morality must be mathematically well defined to include its "subcomponents" such as love, joy, peace, patience, kindness, goodness, faithfulness, gentleness, and temperance.

• …

When all the above theorems are exhaustively proven, we can say that collective quantum algorithms (including quantum intelligence algorithms) running on quantum computers are conscious, intelligent, emotional, creative, moral, etc. Therefore, these collective algorithms can represent all these components of human beings.

If we now define "quantum super wisdom" as the algorithmic components of simulating humans, they only include "consciousness, intelligence, emotion, creativity, and morality". Then such quantum AI will be conscious (of course, it is also intelligent, emotional, creative, and moral). In more rigorous English: **If Quantum Super Wisdom (QSW) is algorithmically consisted of consciousness, (general) intelligence, emotion, creativity, and morality, then QSW is conscious, intelligent, emotional, creative, and moral.**

Here is the illustration for this article "Is AI Conscious?". It visually represents the philosophical and scientific debate, featuring elements of classical and quantum computing alongside abstract human emotions, intelligence, and responsibility.

This tutorial includes theorems of "quantum algorithmic information processing" and their proofs of "quantum computability" in five areas: quantum consciousness , quantum intelligence, quantum creativity, and quantum morality.  As we see from above, if we define "quantum wisdom" includes exactly these five properties, then we can say " wisdom is quantum computable."  In other words, the following five theorems are only "necessary conditions" but not "sufficient conditions" to say "wisdom is quantum computable" or "Wisdom is computable using a quantum computer.":

1. Consciousness is quantum computable.

2. Intelligence is quantum computable.

3. Creativity is quantum computable.

4. Emotion is quantum computable.

5. Morality is quantum computable.

Each of these necessary conditions will be proved in the following 5 sub-pages.  Note that except conscience computability, the proof of the other 4 computabilities may not be very sophisticated as we like.  We expect the project (https://sites.google.com/view/quantumintelligenceassociation/projects/ai-assisted-mathematical-theorem-proof) to tighten the proofs in the future.

Let us now define: "Wisdom" comprises exactly consciousness, intelligence, creativity, emotion, and morality.  Then we also have the sufficient conditions to say:

Wisdom is quantum computable.

(Clarification: quantum wisdom and quantum intelligence are different.  In simplified Chinese, they are different because the former is 智慧 and the latter is 智能。 However, in traditional Chinese they all are called 智慧，which is confusing and we suggest it is better to follow the simplified Chinese way of translation.)