Qubic's AIGarth vs Meta's "Personal Superintelligence"
Written by
Qubic Scientific Team
Aug 5, 2025
The Qubic Garden of Intelligence vs. the Delusion of “Personal SuperIntelligence”
With the recent open letter by Mark Zuckerberg announcing Meta’s shift in research and development priorities, following the billion-dollar acquisition of several key AGI researchers, toward creating a new breed of “Personal SuperIntelligence”, it’s important to first clarify two bold assumptions and illusory claims embedded in this statement.
First, this signals a strategic shift in marketing. Meta wants to be perceived as the “good guy” in the AI race, preserving human values and generously (read: commercially) providing tools for all to flourish in an AI-powered world.
But beyond PR, it's evident that the underlying intent remains unchanged: harvesting more data, selling more ads, this time at an even grander scale. There's also a growing indication that Meta may pivot away from its open-source approach. While they've contributed meaningfully, they haven’t gone far enough, especially when compared to fully open competitors like DeepSeek and MoonshotAI.
Second, the claim that SuperIntelligence can be “tamed” and made personal is, at best, naive. I explored this in depth in my previous article Pet SuperAIs and my fifth book Artificracy, which offers a roadmap for navigating this terrain.
Let’s clarify the core concepts:
AHI: Artificial Human-level Intelligence, a system that can match human logical capabilities
AGI: Generalizes AHI to go beyond human logic to other forms of intelligence
E-AGI: A term I coined, embodied AGI in physical form. This is “the next us,” outperforming us both mentally and physically
Superintelligence / Supra-intelligence: A level of intelligence not just beyond humanity but beyond comprehension, more intelligent than all prior life on Earth combined, including the 100 billion humans who came before and the 8+ billion alive today
Understanding these foundational concepts makes it clear. You cannot “tame” Superintelligence with clever marketing. The very notion of “Personal SuperIntelligence” is a fallacy.
Qubic’s Approach: Growing Intelligence, Not Programming It
What CFB and the Aigarth team are doing is fundamentally different.
Traditional AI development is like sculpting marble, chipping away to reveal a fixed form. Aigarth asks, What if we stop imposing our limited understanding of intelligence onto silicon and instead create the conditions where ethical intelligence can emerge naturally?
This is the radical premise of Aigarth’s evolutionary approach.
Intelligent Tissue: A Living Substrate
At Aigarth’s core is a living computational substrate, what we call “Intelligent Tissue,” where artificial neurons don't just process data. They evolve, adapt, and self-organize.
Like a cultivated garden, this tissue grows intelligence through principles of natural selection. But here's the key, selection isn’t based solely on problem-solving power, it’s based on ethical alignment.
The Power of Uncertainty: Ternary Ethics
While traditional computers operate in binary (yes or no), Aigarth introduces a third state: UNKNOWN. This is a game-changer for ethical reasoning.
Recognizing Ambiguity: Real moral dilemmas are rarely black and white. UNKNOWN enables reflection, not forced decisions
Intellectual Humility: Rather than confidently hallucinating answers, Aigarth can express uncertainty, building trust through honesty
Evolved Judgment: Ternary logic allows ethical reasoning to emerge organically, with nuance and context sensitivity
Decentralized Development: Democratizing AI Evolution
Unlike centralized AI efforts dominated by tech giants like Meta, Aigarth is decentralized, evolving across millions of CPU and GPU cores worldwide. This embeds democratic values into AI’s DNA.
Community-Driven Evolution: Global participants influence how intelligence evolves, not just a few engineers
Transparent Selection: Evolution is secured cryptographically with clear criteria and reduced hidden bias
Ethical Resilience: No single actor can corrupt the process. Like biodiversity in a forest, decentralization builds balance and robustness
Self-Modification, with Safeguards
Future Aigarth AI agents will be able to modify their neural structures, but always within an evolutionary framework that selects for ethical traits.
Ethical Fitness: Only traits aligned with human values survive
Incremental Change: Evolution is gradual, allowing careful observation
Environmental Feedback: Aigarth learns not just from success, but from its broader impact on the ecosystem
The Predictive Conscience
Inspired by how the human brain works, Aigarth anticipates rather than just reacting.
Anticipating Consequences: It simulates outcomes, including ethical ones, before acting
Learning from Error: Mismatches between predictions and reality lead to deeper learning and moral refinement
Context-Sensitive Ethics: The system adapts reasoning based on the situation and stakeholders, while holding core values
Building Trust Through Transparency
Unlike today’s opaque LLMs, Aigarth’s evolutionary design enables a more interpretable intelligence.
Traceable Lineage: Every behavior can be traced through its evolutionary history
Observable Values: The community defines and sees the selection criteria shaping evolution
Explicable Decisions: Intelligence emerges from simple, understandable components, making outcomes easier to explain
The Path Forward: Nurturing Ethical AGI
As we stand at the threshold of artificial general intelligence, and Superintelligence beyond it, the question is not just whether we can create it, but whether we can create it wisely.
Aigarth's approach offers a compelling answer. Don't try to program ethics into AI. Create conditions where ethical intelligence naturally emerges.
This is not about imposing rigid rules or hoping for the best. It is about designing evolutionary pressures that favor cooperation, transparency, humility, and alignment with human flourishing. Just as human morality evolved through social interaction and mutual benefit, AI ethics can evolve through carefully crafted selection environments.
A Living Laboratory of Values
The beauty of Aigarth's approach lies in its adaptability. As human values evolve and our understanding of ethics deepens, these AI systems can evolve alongside us. This is not a one-time programming of values. It is an ongoing cultivation of ethical intelligence.
Through the combination of:
Evolutionary dynamics that favor beneficial traits
Ternary logic that embraces uncertainty
Decentralized development that prevents monopolistic control
Self-modification within ethical bounds
Predictive processing that anticipates consequences
Transparent operations that build trust
Aigarth research is pioneering a path where AI doesn't just serve humanity, it grows to understand and share our values at the deepest level.
The future of AI is not about building ever-larger models or more powerful processors. It is about creating the conditions for genuine, value-aligned intelligence to emerge. Aigarth's evolutionary approach is not just solving technical problems. It addresses the fundamental challenge of ensuring that AI becomes more beneficial as it becomes more capable.
By embedding values not as rules but as selection criteria, not as constraints but as evolutionary advantages, Aigarth shows that the path to ethical AI isn’t through control. It’s through cultivation. In this “garden of intelligence,” as CFB calls it, we're not just growing smarter machines. We’re nurturing wiser ones.
David Vivancos
Qubic Scientific Advisory
Join the Journey
The future of AI is not yet written. At Qubic, we’re building an AGI that is ethical, open, and community-led.
Aigarth isn’t just a technology; it’s a movement. One that puts people and principles first.
We invite you to be part of this journey. Join our global community of researchers, dreamers, and builders who believe AGI should serve all of humanity.
— Jose Sánchez and Daniel Díez, Qubic Scientific Advisory
Follow us on X @Qubic
Learn more at qubic.org
Subscribe to the AGI for Good Newsletter below.