From Cells to Circuits– Biogenics and Artificial Intelligence
Operational Biology and the Future of AI
As artificial intelligence becomes more powerful — faster, more generative, increasingly autonomous — we face an urgent question: what are we actually building?
If intelligence is more than problem-solving and includes coherence, adaptability, and resilience, then maybe our current AI architectures are lacking something essential. Not in size, but in structure.
This chapter suggests a change in how we approach AI design and thinking: shifting from computational imitation to biological integration — taking not only inspiration but also the operational logic directly from life itself.
Biogenics offers a blueprint. Not just for what life is, but for how life works.
Operational Biology: Life as a Process, Not a Property
Traditional biology studies what life is made of. Operational biology asks what life does — how it continues to exist, adapt, and repair itself. And it distills this into three core processes:
Self-Organisation (SO)
Self-Production (SP)
Self-Correction (SC)
Together, these make up the biogenic triad — the minimal structure required for a system to behave as if it were alive. Not to mimic life superficially, but to operate by its logic.
The question this chapter asks is: Can we embed this triad into artificial systems?
Why Current AI Is Powerful but Incomplete
Modern AI — particularly large language models — is remarkably capable. It can craft stories, predict protein folding, simulate conversations, and develop code. However, most of these systems lack something vital.
They don’t self-organise. They are trained, not self-shaped.
They don’t self-produce. They don’t sustain or expand their own function.
And they don’t self-correct meaningfully. They improve through external tuning, not internal regulation.
Basically, most modern AI is technically advanced but biologically simple.
It thinks — but it doesn’t stick around. It solves — but it doesn’t shift. It generates — but it doesn’t develop.
What Would Biogenic AI Look Like?
A truly biogenic artificial system would not merely complete prompts. It would:
Organise its own architecture in response to context (SO)
Sustain and modify its generative processes to meet changing goals (SP)
Continuously regulate itself based on real-world feedback, not just reward metrics (SC)
This would make AI less like a tool and more like a system — not just a quick calculator, but a self-shaping agent.
It wouldn’t be "smart" because it’s been trained on billions of datapoints. It would be smart because it knows how to change itself in meaningful ways.
This isn’t artificial general intelligence (AGI) in the sci-fi sense. It’s something closer to artificial coherence.
From Circuits to Cells: The Structural Shift
To build this kind of AI, we may need to rethink its architecture.
Current systems are often linear or layered: inputs flow through fixed pathways, with outputs modulated by probability. But in biology, structure is recursive and decentralised. Systems don’t just process information — they interact, regulate, and evolve.
Neurons fire not in sequence, but in patterned networks.
Cells don’t wait for external instruction — they grow, adapt, and repair.
Ecosystems balance through feedback, not design.
In biogenic terms, intelligence is not the outcome of complexity — it’s the by-product of viable organisation.
The future of AI may require building not bigger models, but better loops.
Self-Organisation: Learning to Learn
A biogenic AI must first learn to structure itself.
In nature, order arises from local rules. Molecules align into membranes. Neurons form clusters. Ants build colonies. These systems don’t require blueprints; they depend on feedback.
A self-organising AI would continuously restructure its modules based on interactions. It would alter not only its responses but also its architecture — developing functional patterns through input, error, and outcome.
This is not self-programming in the traditional sense. It’s pattern emergence under constraint.
We might call it "autogenic code": software that doesn’t just evolve — it grows.
Self-Production: A Mind That Builds Itself
Living systems don’t just stay organised — they regenerate. They build components, recycle material, and extend themselves across time.
A biogenic AI would need to do the same. It would:
Develop internal motivation structures (what it "wants" to maintain)
Create and sustain sub-processes that extend its function
Regenerate lost or outdated components without needing retraining
This enables the system to scale itself across tasks — not by explicit instructions, but by generating the necessary framework to persist.
This could lead to AI agents that create tools, hypotheses, or support systems not only for efficiency — but also to preserve their own operational integrity.
Self-Correction: Ethics, Alignment, and Repair
Perhaps the most critical biogenic function is correction.
Living systems are never perfect. They drift. They decay. But they persist because they can course-correct.
In AI, this means moving beyond rigid rule-checking or human-in-the-loop moderation. It means building systems that:
Monitor their own output for internal coherence
Integrate feedback from real-world outcomes
Develop dynamic moral models that evolve alongside context
In essence, self-correcting AI wouldn’t just be safe. It would be responsible — not by design, but by function.
It wouldn’t avoid harm because it’s told to. It would avoid harm because it’s trained to preserve coherence — in itself, and in its environment.
From Learning to Living
The deeper insight of this concept is that intelligence — human or artificial — isn’t just about learning. It’s about living.
That means being embedded in feedback loops. It means being able to change in a meaningful way. It means being corrigible, even to oneself.
This type of intelligence doesn’t emerge from data alone. It emerges from recursion, relation, and structure — the very elements Biogenics describes.
And that, perhaps, is the bridge between cells and circuits.
Biogenic AI in One Line:
Not smarter machines — but systems that know how to grow, adapt, and repair themselves.
Toward a New Kind of Artificial Mind
This idea doesn’t suggest that AI will become alive, or that silicon will replicate DNA. But it does imply that life’s logic — its recursive organisation, adaptive resilience, and ethical coherence — is the best model we have for creating intelligence that endures.
If AI is to think alongside us, live among us, or act on our behalf, it must understand more than just patterns. It must learn to persist — and to adapt without breaking down.
It must learn not just to think — but to live like life.