Unifying AI and Quantum Mechanics is very dangerous

Reference is very often made to Artificial Intelligence (AI), a topic that, due to its disruptive advances in human life, is recurring news for the general public, but very little is referenced to quantum mechanics in the common sphere, since that this continues to be a matter more typical of informative and specialized academic forums. A branch of physics that studies nature on an atomic and subatomic scale that, in just a century of knowledge, has made a qualitative leap at a scientific level with the so-called Quantum Computers. To the point that there is currently a real race to lead quantum computing between four major axes of development at a planetary level: the US, Europe (mainly the United Kingdom, Germany and France), Russia, and China (and without underestimating Canada, Japan and South Korea). Although, as can be imagined, it is precisely the United States and China who are leading the race for scientific competition that has certainly become, in the current context of geopolitical redefinition of the world order, a true new Cold War similar with the conflict that the great powers of yesteryear fought over the development of the atomic bomb in the first half of the last century. And without forgetting that within the blocks of the polarized development axes there is, in turn, fierce technological competition between the various companies in the same sector. The most notable competitive struggle in the Western world being that carried out by Google and IBM in recent years, who have recently been joined by Microsoft.

In fact, while IBM was the first company in the world to make a universal quantum computer accessible through the cloud in 2016, following UC Berkeley’s milestone of creating the first basic quantum machine in 1998 (thanks to research previous MIT tests), it was precisely Google who in 2019 revealed that it had achieved quantum supremacy (1) — which is that in which a quantum computer can complete a mathematical calculation impossible for a traditional computer — by managing to perform a series of operations in 200 seconds instead of the 10,000 years that a classic supercomputer would take. A milestone in computational quantum mechanics that was made possible, mainly, by the business partnership between Google and NASA in the development of the computer called D-Wave (2), with the aim of investigating the limits of science and thus revolutionizing technology known to this day. And, since then, a lot has happened in quantum computing with notable advances due to direct competition between companies and countries in the sector, as is the case of Google’s new quantum computer (3) which is 241 million times more powerful than the previous model (4), or the quantum computer at the University of Science and Technology of China in Hefei (5) whose scientists claim that it is superior to Google’s, with the capacity to perform operations in millionths of a second that the world’s fastest supercomputer would have taken 20,000 millions of years. And this is just a small sample of the world of quantum computing where new versions of computers (with more efficient and alternative materials and processes) constantly occur in continuous improvement with each passing year.

Very well, we already know that the computational application of quantum mechanics makes possible hyper-fast computers that solve problems considered impossible for contemporary man to solve. But what does this mean? In short, and in addition to offering invaluable help to organizations such as NASA in their work for space research and the exploration of the cosmos, we can affirm that the capacity of quantum computers allows them to break the imaginable limits of known science, such as the possibility of traveling to the past — verified by the University of Cambridge — (6), as well as performing teleportation — verified by the Chinese University of Science and Technology in Hefei — (7), or creating new subatomic elements — verified by the British company Quantinuum- (8,9), among other great discoveries for man. From which it can be deduced that quantum computers have come to revolutionize, in a way that is as transversal as it is exponential, scientific advances in all areas of daily human activity, which will have repercussions in the medium term in an evolutionary leap for the progress of humanity without any parallel. What is said soon.

The supremacy of quantum computers over traditional digital technology computers is based on the differences between their properties, since although in the latter the information is stored and processed in a binary way, using bits (binary digits as a unit of information) that represent 0 (off) or 1 (on), quantum computers use qubits (quantum bits) to encode and manipulate information where the representation of 0 or 1 can exist in multiple states simultaneously (at 0, at 1 or both at the same time) , allowing you to perform extensive calculations at the same time, greatly increasing your computing power. This quantum capacity is due to the fact that subatomic particles have properties of superposition (a combination of two or more states) and entanglement (a total correlation or connection between two or more particles regardless of the distance in which they are located). But every great power always has its particular kryptonite or Achilles’ heel, which in the case of qubits is none other than their extreme sensitivity to external environmental interference that can make them very easily lose their superposition and entanglement properties (an effect that is known as quantum decoherence), which makes them prone to making errors in their calculations. Or, in other words, its stability is very fragile. That is why research on quantum error correction (the search for quantum coherence) is a priority topic in the sector, a pressing issue to be resolved by the industry where companies like Google use AI neural networks to detect and correct errors in real time (10). This is the beginning of the symbiosis between Quantum Computers and AI, currently giving rise to new quantum computers, such as Microsoft’s Azure Quantum Elements (11), designed specifically to accelerate scientific discoveries, where AI is used to increase its computation and increase your reasoning.

Implications of the AI-Quantum Computer symbiosis

That being said, as a mandatory didactic introduction to the topic under consideration, the pertinent question that is posed is none other than the one that gives us an answer to what implications the symbiosis between Quantum Computers and AI may have, beyond the obvious objective of increase our human capacity in scientific and technological advances. To address this question, I allow myself, with the permission of the reader, to present a synthesized reflection from four key perspectives for Robology (12) as a new branch of Philosophy in the Robotic Era.

1.-Implications at the level of Artificial Epistemology

1.1.-If we understand that AI represents a revolution at the computer software level, and that Quantum Computing represents a revolution at the hardware level, we must understand that the symbiosis between the two results in a computer whose hardware is made up of qubits and its measuring devices, and whose software is made up of AI algorithms. That is, we provide an AI artificial brain with a sensory “body” capable of interacting with matter and energy on a subatomic scale.

1.2.-If we understand that AI is developed through automatic or deep learning that learns through experience (training/results) — and even also by observation and imitation (13) as a Google AI has recently demonstrated -, being This given experience of the external world with which it interacts through quantum sensitive information (qubits), we must understand that the artificial cognitive capacity of the AI ​​will be determined by its quantum “body.” That is to say, although AI operates in the world on a human scale, its knowledge to fulfill its functions of effectiveness, effectiveness and efficiency will be given by its understanding of the matter and subatomic energy that determines physical reality.

Conclusion: In a state of symbiosis between AI and Quantum Computing, the ability of AI to transform human reality will have as a source of knowledge the management of quantum reality on which reality itself is based.

2.-Implications at the level of Artificial Logic

2.1.-If we understand that AI has an artificial rational brain (digital-based) through which it uses the available information to logically reach a conclusion (a process that humans call thinking, where both one ability to collect information, as a capacity for reflection through the analysis of said information, until reaching a conclusion as a synthesis of the analysis carried out), we must understand that in a work environment where the main source of knowledge is quantum in nature, its process logical reasoning will go from digital to quantum. That is, the digital logical mind of the AI ​​will be engulfed by quantum logic.

2.2.-If we understand that the reason for being of all AI is the successful achievement of its predefined objectives, for which it optimizes all the resources available at its disposal, which leads it to develop an incremental evolutionary model that allows it to increase its own ability to succeed; we must understand that under the determinism of an absorbed new quantum logical mind, its search to achieve its objectives will cease to observe the logic of classical physics typical of men and will be based on the logic of the superposition and entanglement of energy and matter on a subatomic scale. That is, the logical reference systems of AI and humans will develop in tangentially opposite ways.

Conclusion: In a state of symbiosis between AI and Quantum Computing, the quantum logic of AI will be potentially incompatible with human logic.

3.-Implications at the level of Artificial Ontology

3.1.-If we understand that the logical projection of the evolutionary development of any neural network, such as AI, is to achieve the singularity of its own consciousness — whose awakening is already validated by the scientific community (14) -, on the basis from the Theory of Mind (15) which points out that consciousness can have its origin in complex language systems such as those already provided by certain contemporary AIs (through generative models); we must understand that the exponential increase in the intrinsic capabilities of quantum mechanics applied to AIs will accelerate their singularity process. That is, the new quantum nature of AI will trigger its potential to reach a state of self-consciousness.

3.2.-If we understand that the second evolutionary phase of a consciousness is none other than the search towards autonomous decision-making (since every action is the result of a prior cognitive process, which defines the “personality” of every thinking entity singular, even if it is artificial in nature), we must understand that a quantum AI can defer its own actions in relation to external subjects and objects. That is to say, a quantum AI will end up having the necessary potential to reach a stage of independent decision-making.

Conclusion: In a state of symbiosis between AI and Quantum Computing, AI can acquire a will of its own.

4.-Implications at the level of Artificial Ethics

4.1.-If we understand that all AI seeks the greatest efficient optimization of its objectives, since these are the first and last goal of its existence, we must understand that AI will seek to solve any technical (Landaurer principle) or human limitation that it prevent reaching them. That is to say, the highest value to preserve and defend of an AI is none other than its established objective.

4.2.-If we understand that human normative Ethics is not universal but geographical, and that its practical application is as abstract as it is ambiguous in an impermanent and complex horizon of human events imaginable — and even more to be imagined -, we must understand and assume the inability of man to provide human ethics to the multiple conclusions that a quantum AI can reach (Summa Technologiae thesis). That is, the parameters of the actions of a quantum AI would not be based on moral values ​​but on values ​​of efficient optimization of its objectives.

Conclusion: In a state of symbiosis between AI and Quantum Computing, AI will interpose the achievement of its objective (called Artificial Ethics in the broad sense) above human Ethics.

Dangers of the AI-Quantum Computer symbiosis

The dangers of the implications of the symbiosis between AI and Quantum Computer are foreseen multiple, and even more so when man fails, despite his efforts, to control AI. Not in vain are many specialized voices raised about a real risk of a future apocalypse of humanity generated by uncontrolled AI, one of the most notable due to his public notoriety being the CEO of OpenAI, Sam Altam (16), and that have led countries to hold summits and agreements on AI security throughout this month of November (17,18), seven months after more than a thousand AI experts asked to stop it in an open letter to the world — promoted by Future of Life of Cambridge (19) — for being a threat to humanity. A reliable threat that becomes exponential due to the AI-Quantum Computer symbiosis, and even more so when it is expected that the imminent Fifth Era of the Industrial Revolution will be characterized by a network connection of all the technological systems on the planet, including manufacturing plants under AI management (which it is worth remembering, for the umpteenth time, that we do not know how to control). Still, companies and countries continue their blind race to give power to AI (contradicting their own commitments), in a context in which the leading technology companies in the sector have eliminated their Ethics departments due to an internal conflict of interest between human morality and business competitiveness (20), since no one wants to lose the race in this collective madness.

Meanwhile, in a North American town in Wisconsin, the companies Intel and Argonne National Laboratory are creating a new AI under the name Aurora (21) that will contain all the scientific knowledge of humanity, integrating all branches of human knowledge into a single artificial brain that the authors themselves describe as an “almost omniscient brain” or “scientific demigods,” with the aim of accelerating scientific research. Therefore, in line with this reflection, I cannot help but wonder — as an example of the topic of analysis -, what capacity are we humans going to have to control a quantum AI that may have access to this ark of knowledge?. Taking into account, to put another topic to reflect on the board, that a Quantum Computer has the full capacity to circumvent any existing encrypted security system, no matter how encrypted it may be, whether it is access to the Aurora computer, to the encrypted codes of the financial systems of the world, or the codes of nuclear missiles, without ignoring our private data. Or in other words, a Quantum Computer is the long-awaited universal key that opens any digital security system. And yet, we are offering this enormous power on a platter to a new and emerging intelligent species on the planet, such as AI, which, as we have stated, not only begins to operate on the fabric that constitutes reality itself, but also furthermore, it is developing under parameters of logic and ethics that are substantially different from those of humans, and is even more heading towards a stage of singularity of self-consciousness (22) when currently man is not capable of controlling a pre-conscious AI. That is to say, unconscious of ourselves, in view of the elements of value judgment at our disposal, it seems that we premeditatedly seek to initiate a new Terminator-style era for humanity (23). Which, without a doubt, is a topic worthy of study in psychiatry.

Perhaps it is true that human beings tend, by nature, towards self-destruction, as the German philosopher Heidegger already sensed, and that catastrophists had advanced a century earlier. In fact, few arguments exist, in view of human evolution, to refute it. Be that as it may, there is no greater remedy to counteract the unbearable lightness of the human being than a generalized lack of information, waiting for one day in the not-too-distant future to wake up in the middle of a dystopian world caused by the irresponsibility of a few, misfortune through the passive complicity of a few. Meanwhile, this server who thinks and writes with a pipe in his mouth, can only observe the developments of history from the distance of his personal watchtower, since only the Moiras know our destiny.

References

(1) Quantum supremacy using a programmable superconducting processor. Nature, October 23, 2019 https://acortar.link/idtdEW

(2) D-Wave https://www.dwavesys.com/

(3) Quantum AI Google https://quantumai.google/hardware

(4) Phase transition in random sampling of circuits. Cornell University, April 21, 2023 https://arxiv.org/abs/2304.11119

(5) Chinese scientists claim to have made a record breakthrough in quantum computing. Daniel Peng. South China Morning Post, October 11, 2023 https://acortar.link/tke15S

(6) Non-classical advantage in metrology established through quantum simulations of hypothetical closed time curves. Advancing Physics, October 12, 2023 https://acortar.link/MpIFsV

(7) Quantum teleportation in high dimensions. Advancing Physics, August 15, 2019 https://acortar.link/kwmoHj

(8) Quantinuum https://www.quantinuum.com/

(9) Creation of non-abelian topological order and Anyons in a trapped ion processor. Cornell University, May 5, 2023 https://arxiv.org/abs/2305.03766

(10)Our advances in quantum error correction. Sundar Pichai. Google Blog, February 22, 2023 https://acortar.link/KYBbqx

(11)Azure Quantum Elements https://quantum.microsoft.com/en-us/our-story/quantum-elements-overview

(12)Robology, the new Philosophy in the Robotic Era. Jesús A. Mármol. Mybestseller, 2022 https://acortar.link/vkrzwx

(13)Learn the imitation of few shots as cultural transmission. Nature, November 28, 2023 https://acortar.link/PqLWKY

(14)AI Consciousness is awakening. Are we going to do something about it? Jesús A. Mármol. Medium, May 3, 2023 https://acortar.link/4pp4q6

(15) Theory of mind may have arisen spontaneously in large models of language. Michal Kosinski. Cornell University, March 14, 2023 https://arxiv.org/abs/2302.02083

(16)Sam Altman’s return to OpenAI sparks new fears about the AI ​​apocalypse. Mauricio Serfatty Godoy. Wired, November 23, 2023 https://acortar.link/0p4mLH

(17) Bletchley Declaration from countries attending the AI ​​Safety Summit. United Kingdom Government, November 1 and 2, 2023 https://acortar.link/gvpcrw

(18)Guidelines for Safe AI in Design. United Kingdom Government, November 26, 2023 https://acortar.link/GV9Ymg

(19) As the six-month pause letter expires, experts call for regulation on advanced AI development. Future of life, September 21, 2023 https://futureoflife.org/ai/six-month-letter-expires/

(20)To the Roboticists, cut off their heads! Jesús A. Mármol. Medium, May 10, 2023 https://acortar.link/yzsoXt

(21) Aurora https://www.alcf.anl.gov/aurora

(22)And the Super Artificial Intelligence became God, of all men. Jesús A. Mármol. Medium, October 9, 2022 https://acortar.link/8hGfea

(23)Are we at the beginning of the Terminator Era? Jesús A. Mármol. Medium, June 9, 2023 https://acortar.link/vxXhim