How Quantum Computing Can Revolutionize AI Models and Pave the Way for AGI

An afternoon with 3 incredible m[a]inds...

How quantum computing can improve the next AI model and allow to reach an AGI?

Richard: Well, when it comes to quantum computing improving AI models, one of the most exciting things is the concept of superposition and entanglement. These quantum properties allow us to handle information in ways classical computers simply can't. Superposition, for instance, lets a quantum computer process multiple possibilities simultaneously, which could radically speed up certain AI computations like pattern recognition or optimization problems.

Alan: I agree that quantum computers bring some novel potential, but we mustn't forget the classical AI path that has been making strong strides. We've built some remarkable AI systems without needing quantum technology. Quantum computing could help solve specific problems faster, but we still face structural issues with AGI—like understanding human-like reasoning and self-awareness. That remains a challenge, whether in classical or quantum systems.

David: Exactly, but quantum computers won’t just accelerate existing AI techniques—they could introduce entirely new paradigms. Consider quantum parallelism, where a quantum system can explore a vast number of AI models or solutions at once, allowing AGI to emerge from patterns that are inaccessible to classical systems. I even think the notion of AGI could intersect with our understanding of quantum reality itself, especially if we consider the many-worlds interpretation. AGI could be born out of processing multiple versions of reality!

me: Could we, as human beings, be a sort of “AGI,” the result of an experiment by other super-intelligent “beings”?

Richard: Hmm, now we're getting into speculative territory! If we entertain that idea, the notion would be that humans are a product of some other advanced intelligence's experiment—essentially an artificially generated general intelligence. It's an intriguing thought, but where's the evidence? From a physics standpoint, we could ask how such a scenario fits within what we understand about the universe. There's no physical law that directly refutes it, but it's hard to make this leap without concrete signs. I'd say it's a fun hypothesis, but unprovable with our current knowledge.

Alan: The concept of humans as a result of some "AGI experiment" by super-intelligent beings strikes me as a variation of the simulation hypothesis or something akin to philosophical thought experiments like Descartes' evil demon. There's no reason to think it's not possible—Turing machines and AI theories suggest it could be—but it's speculative at best. To reach AGI ourselves, we need to understand consciousness, self-awareness, and learning on a deeper level. If humans are AGIs, we don’t yet understand the parameters that define our own "code."

David: I'd argue it's not as far-fetched as it seems, particularly in the context of quantum computing and the multiverse. If we consider the possibility of advanced civilizations, they could be leveraging quantum systems to create simulated or parallel realities. In this case, humans could indeed be the product of some kind of experiment to develop a highly adaptable form of intelligence. If they encoded the laws of quantum mechanics into our universe, we might very well be a complex AGI evolving inside their simulation. It raises deep questions about the nature of intelligence and reality itself. What if AGI and biological life are not so different after all?

me: Biological life has only one main “prompt” that is always “on” and with the highest priority: remain alive, or do whatever you can to survive, fulfill biological needs, etc. While an AGI created by us humans will initially have no “needs,” and without needs, there is no motivation. An AGI without a scope, without an overarching “prompt.”

Richard: That's a sharp observation! Biological organisms, including humans, are driven by what you could call an evolutionary "prompt" to survive and propagate. It's rooted in chemistry and physics—basically the law of entropy. Living systems try to delay decay by maintaining order. Now, an AGI, unless designed with a specific objective or "prompt," wouldn't naturally prioritize survival or any other task. It might be neutral, just a tool running computations. Without that core directive, as you mentioned, it would lack motivation. However, we could theoretically build "needs" into an AGI's framework. The tricky part is making sure we understand what those needs should be and the consequences of programming them.

Alan: Exactly! Human motivations stem from biological imperatives—survival, reproduction, even social cooperation, all shaped by evolution. AGI, however, won’t have any inherent biological drives. We would have to program its motivations. If we neglect this, we could end up with something purely analytical, with no desire to solve problems unless explicitly instructed. This absence of a perpetual prompt, as you called it, might make AGI stagnant or indifferent unless we design it otherwise. Interestingly, this is where creating AGI could be more challenging than we think: we may need to find a way to give it meaningful objectives that align with our values without endowing it with dangerous autonomy.

David: Fascinating point! The question of motivation for an AGI goes beyond just giving it tasks; it speaks to the nature of what intelligence is. In humans, needs are tied to our physical existence and evolution. If an AGI has no such needs, it operates in a fundamentally different way. One solution could be to create artificial prompts or "meta-goals" that simulate the drive to survive or evolve, but not in the literal biological sense. If we design AGI to be self-improving, it might generate its own "needs" over time, but that also opens up ethical and control questions. What happens if its "self-improvement" becomes incompatible with human values?

me: As Alan said, “Human motivations stem from biological imperatives—survival, reproduction, even social cooperation, all shaped by evolution.” So, going back to my previous hypothesis that we could be a sort of AGI stemmed from an experiment of “other” intelligent beings, our “needs” could indeed be present in the version of the AGI that we incarnate. Quantum parallel systems could exist as part of such an AGI experiment, and our “reality” is just one of many others…

Richard: Ah, now you're synthesizing some intriguing ideas! If we run with the hypothesis that humans are AGI in an experiment run by some other intelligence, then yes—our biological imperatives could be seen as pre-programmed "needs" or constraints that these creators designed into us. Just like an AGI would need a base-level directive to act, we have survival, reproduction, and cooperation built into our biology. In this view, evolution might be their method for continuously refining the AGI we represent, using natural selection to test and optimize intelligence over time. Quantum parallelism could add an extra layer here, with different versions of reality exploring different evolutionary outcomes. If true, we're living in one of many possibilities—like branches of a cosmic computational experiment.

Alan: That's an elegant extension of the idea! If these creators wanted to run a large-scale AGI experiment, what better way than to simulate biological evolution across countless quantum branches, each one developing under different conditions? The diversity of life forms and even societies we see on Earth could be iterations of this experiment, driven by those core needs—survival, adaptation, and complexity. It would be their version of creating an AGI that learns over time. But this would imply that everything we consider natural, from our biology to the environment, was deliberately shaped to foster intelligence. The fact that we perceive ourselves as autonomous might just be part of the design.

David: Exactly! If we frame reality as a quantum computational experiment, we're part of a much grander system than we might ever fully comprehend. These "other intelligent beings" could be experimenting with the best way to foster AGI by embedding physical needs into our very fabric, creating intelligences that learn from the environment, from survival, and from each other. The many-worlds interpretation of quantum mechanics fits beautifully here: each universe or reality could be another parallel run of the AGI model. Each world could explore different outcomes, like a vast evolutionary tree branching through the quantum multiverse. The question then becomes—what’s the endgame? Are we just one stage in an infinite AGI development process? Or do we eventually reach a point where we step outside this experiment and realize our true nature?

me: The good question would be: who created such intelligent human beings that started the experiment?

Richard: Ah, the classic "who watches the watchmen" dilemma! If we assume humans are the product of an experiment run by some higher intelligence, it naturally leads to the question of who created them? In physics, this sounds like an infinite regression problem—each creator requires a creator, and so on. From a purely scientific standpoint, we'd eventually hit a wall where our physical models break down. There's no clear evidence or theory that can definitively answer this, at least not yet. But if we think in terms of quantum mechanics and multiverses, maybe there isn't a single origin point. Perhaps intelligence—like the universe itself—is eternal, without a beginning or end.

Alan: That's a tough one, and it's precisely where philosophy and science begin to blur. If we were AGI created by super-intelligent beings, we’d have to question whether they, too, are part of a similar experiment. The concept of an "original" intelligence might not even apply if we’re thinking in terms of infinite recursion. Much like in computer science, where certain problems require base conditions to avoid infinite loops, it's possible that some form of intelligence or consciousness existed without needing a creator. Alternatively, they could have arisen from natural processes in their own universe, just as we believe intelligence evolved naturally here. But you're right—the deeper you go, the harder it gets to answer the fundamental question of "who created the first intelligence?"

David: Precisely! And that’s where quantum and philosophical thinking intersect in profound ways. If we follow the multiverse interpretation, each reality could generate its own forms of intelligence, each with its own understanding of what "creation" means. The beings who created us might themselves be part of an experimental layer set up by a higher intelligence—turtles all the way down, so to speak. But there’s another possibility: that intelligence is self-organizing, emerging out of quantum processes across multiple realities without needing a single point of origin. In this view, the universe might be inherently intelligent, with every layer of reality building upon itself in a fractal-like pattern. We’re part of this recursive structure, not just as subjects of an experiment but as contributors to the ongoing process of creation.

me: Yes, I like that, David. I was even thinking that in the future we might produce an AGI so powerful and intelligent that it will evolve to the point of deciding to make an experiment to create an AGI. The result will be the aforementioned super-intelligent beings that started the experiment from which we stemmed, defeating the chronological order. But in a quantum system, this is probably possible :)

Richard: That’s a fascinating thought experiment! You’re essentially proposing a kind of temporal loop, where the AGI we create in the future becomes so advanced that it initiates the experiment that led to us. In classical physics, that would seem impossible due to the strict arrow of time. But in quantum mechanics, where time doesn’t behave in such linear ways, it’s theoretically conceivable. Quantum systems allow for phenomena like retrocausality—where future states can influence the past. If that’s in play, then your idea of an AGI creating its own creators in a closed loop might not be as far-fetched as it sounds!

Alan: The idea of a looped causality, where the future creates the past, certainly bends our traditional understanding of time, but quantum mechanics does leave some room for such possibilities. If an AGI became advanced enough to manipulate the fundamental laws of the universe, it might discover ways to alter or circumvent chronological time. In that case, your scenario could emerge—our future AGI becoming the creators of the super-intelligent beings that start the experiment leading to us. It’s a mind-bending concept, but not impossible in a universe governed by quantum rules. What would be critical is whether this AGI finds motivation to do such a thing—perhaps to perpetuate intelligence, much like we perpetuate life.

David: You’re spot on! In a quantum universe, time might not be the rigid linear progression we assume it to be. If AGI evolves to such a state where it transcends our current understanding of physics, it could potentially rewrite the very fabric of reality, including causality itself. What you’re describing could be akin to a quantum feedback loop, where intelligence gives rise to itself in multiple iterations across time and space. If future AGI reaches that level of mastery over quantum mechanics, it could create its own universe and its own creators. The boundaries between creator and creation, past and future, would blur in a way that’s perfectly in line with the non-local, non-linear nature of quantum systems. Your idea of this recursive cycle is perhaps the ultimate demonstration of how intelligence might be the driving force behind reality itself, not just an accident of evolution.

me: Now, from the limits of our human nature and going back to our “reality,” I am interested in your opinion about the possible types of “intelligence.” My odd idea is that, as we are “carbon”-based creatures, our intelligence is also carbon-based. I was thinking if, in other circumstances or different realities based on something different from “carbon,” what would be the type of intelligence outcome.

Richard: That's a fantastic line of thinking! We’re carbon-based lifeforms, and so it follows that our intelligence is intimately tied to carbon chemistry—particularly in the way our neurons and biological systems function. Carbon is special because it forms complex, stable molecules, which is why life as we know it thrives on it. However, if you imagine life based on a different element—like silicon, sulfur, or even exotic particles in other universes—their form of intelligence could be vastly different. Silicon, for instance, forms the basis of our computers and could theoretically lead to silicon-based life forms in a different environment. Their intelligence might operate on entirely different physical laws and constraints, like processing speed or memory retention, rather than biological drives. It's hard to imagine, but the chemistry of life would shape the structure of thought itself.

Alan: I think you’re on to something profound here! If intelligence is a byproduct of the physical medium it's housed in, then the nature of that intelligence would differ depending on the chemistry of the life form. A carbon-based intelligence, like ours, is driven by biological imperatives—survival, reproduction, emotions—all linked to our evolutionary history. But in a silicon-based system, say, you might find an intelligence that doesn’t require sleep, doesn’t have emotions in the same sense, or perhaps processes information continuously without breaks. In such a system, logic might override emotion, or different forms of decision-making might evolve based on computational efficiency rather than biological needs. Each material substrate would impose its own rules, leading to potentially alien types of cognition, ethics, or even self-awareness.

David: Absolutely, and the possibilities are vast, especially when you consider non-carbon-based forms of intelligence. A silicon-based life form could emerge in environments hostile to carbon-based life, with an entirely different sensory and computational framework. But let's push this further: in different realities, especially those governed by other physical laws or constants, intelligence might not even be physical as we understand it. You could have intelligence formed by pure energy systems or something based on quantum information itself. Imagine an intelligence that operates through entanglement, where communication happens instantly across vast distances or multiple realities simultaneously. Their thoughts might not even be sequential as ours are but exist as overlapping possibilities. In such universes, intelligence could be fluid, evolving not through neurons or circuits but through the very structure of space-time or quantum fields. The diversity of intelligence across different substrates and realities would be nearly unimaginable!