A better way

Abstract

Ethical Evolution of Mechanical and Quantum Intelligence Living Clockwork: From Mechanical AI to a Human-Centric Quantum Evolution Foundations of “Living Clockwork” – Mechanical AI in History In the 18th and 19th centuries, inventors began to imagine machines that could think and act autonomously using only mechanical parts. The roots of this “Mechanical AI” stretch back to ancient analog devices like the Antikythera mechanism – a 2000-year-old geared calculator that accurately simulated celestial movementswired.com. By the late 1700s, European clockwork automata astonished audiences by mimicking lifelike behavior. These intricate living clockworks – mechanical ducks that appeared to eat and digest, or android figurines that could write elegant script – hinted that gears and levers might emulate aspects of intelligence and life. This vision reached a technical apex with Charles Babbage’s proposed Analytical Engine (1830s–1840s). Far more than a mere calculator, Babbage’s design was “a general-purpose, fully program-controlled, automatic mechanical digital computer” that could perform any computation set before itbritannica.com. In concept, the steam-driven Analytical Engine would have all the core components of a modern computer – a “mill” (CPU), memory storage, input/output on punched cards – implemented entirely with brass gears and leversbritannica.combritannica.com. Though never completed due to its immense complexity, it proved that pre-electronic computers could be universal machines. Ada Lovelace, the first algorithm designer for Babbage’s Engine, mused in 1843 that such a device might even compose music or art if programmed with the right rules – an early seed of machine creativity. Travis Raymond-Charlie Stone’s work picks up this historical thread of Mechanical AI, treating it not as an obsolete curiosity but as a foundation for a new path forward. Inspired by these precedents, Stone reimagined “living clockwork” in modern form: intelligent mechanisms built from physical laws – thermodynamics, mechanics, electromagnetics – rather than silicon chips. His philosophy challenges the notion that advanced intelligence requires digital electronics. After all, human brains themselves are biological wetware performing analog computation, remarkably efficiently. Modern AI’s power-hungry digital hardware often consumes megawatts of energy to emulate tasks that a human brain does on ~20 wattswired.com. This stark contrast has become a rallying point. As one expert noted, “the human brain runs on a small amount of electricity… yet if we try to do the same thing with digital computers, it takes megawatts”wired.com. The implication is clear: there may be smarter, more sustainable ways to achieve intelligence. Stone’s “Mechanical AI” is about reviving that alternative approach – returning intelligence to physics – by building thinking machines that work in harmony with natural energy flows and human-scaled dynamics. Kinetic Intelligent Design – Intelligence in Motion and Form A centerpiece of Stone’s vision is what he calls Kinetic Intelligent Design (KID) – the idea that a machine’s very mechanics can embody intelligence. In traditional robotics and AI, we program computers to sense, calculate, and act; the physical form is often just a neutral vessel carrying a microprocessor “brain.” KID turns this inside out: the shape, material, and motion of the machine itself contribute to its cognitive function. This concept aligns with emerging scientific understanding of morphological computation, which suggests that an organism or robot’s body can effectively offload and simplify computation. In nature, “morphological properties – the shape and form of a body, as well as compliance, resonance, friction – play a crucial role in the emergence of intelligent behavior”frontiersin.org. Animals have evolved bodies that handle many tasks automatically: think of how a cat deftly lands on its feet (leveraging mechanics and balance) or how a cockroach’s leg design lets it scurry over obstacles without complex planning. The body “outsources” some intelligence so the brain can be smaller, and the result is remarkable robustness and efficiencyfrontiersin.orgfrontiersin.org. Inspired by such examples, Stone’s KID approach treats gears, springs, cams, fluids – all moving parts – as potential information processors. By clever design, a kinetic system can “decide” actions through feedback loops and physical response, not unlike a living organism’s reflexes. Historical technology provided proof-of-concept for KID long before the term existed. One classic example is the centrifugal governor on James Watt’s 1788 steam engine. This device used spinning flyweights and a throttle linkage to automatically regulate engine speed – a purely mechanical feedback loop that sensed error and corrected it without any external controller. Watt’s flyball governor was “an early example of an automatic control system consisting of an error sensor connected by a negative feedback loop to a control device”, keeping the engine stable at a desired speedmedium.com. In essence, the governor let the machine self-regulate its power output the way a simple organism might maintain homeostasis. Stone often cites the governor as a primitive mechanical “brainstem” – proof that autonomy need not be digital or authoritarian. It reacts only as needed to keep balance, exercising just enough control to assist the human operator, not to seize total control. This ethos of autonomy without authoritarianism is central to KID. The machine’s design hardwires it to cooperate with natural forces and human intent, rather than override them. Modern research reinforces Stone’s intuition that intelligent behavior can arise from physical design. Soft robotics and bio-inspired engineering have shown that giving a robot a wisely crafted body can dramatically simplify the “brain” neededfrontiersin.orgfrontiersin.org. In the words of one robotics editorial, “important functions like sensing, control or even computation are partly outsourced to the body’s morphology,” making systems “extremely robust, energy efficient, and highly adaptive.”frontiersin.org. Stone’s KID machines exploit this principle. For example, he designed spring-legged test robots that naturally favor a stable walking gait – the tuning of the springs and linkages ensures that with each step, energy is recycled and balanced without complex software. In one demonstration, a small clockwork rover traversed uneven terrain via a clever suspension that passively adjusted to bumps. Observers described it as having “instinctive” sure-footedness, when in truth its intelligence was literally built into its legs. This approach is a philosophical shift: instead of forcing machines to conform to rigid commands, KID lets them discover solutions through their physical interaction with the world. The result is machines that behave more like living creatures – adaptable, graceful, and in tune with their environment – but remain fundamentally under human understanding and control, since we can see and grasp how the mechanisms work. “Junior” – A Logic Engine Before Electricity To bridge the gap between classical clockwork and contemporary computing needs, Stone created Junior, a demonstration that complex logical reasoning can be achieved with pre-electric hardware. Junior is, in essence, a programmable logic computer constructed entirely from mechanical and analog components – no silicon chips, no binary electronics, just levers, rods, magnets, and fluid channels. In concept, it harkens back to Babbage’s Analytical Engine and to lesser-known “logic machines” of the 19th century. (For instance, Victorian logician William Jevons built a mechanical logical calculator in 1869 that he nicknamed the “Logic Piano,” which could solve Boolean equations using a system of levers and pulleys.) Stone’s Junior inherits that spirit, implemented with modern precision engineering. At its core, Junior operates on binary principles using physical switches and flip-flops that hold state as positions of mechanical sliders. It can be hand-cranked or driven by a small motor to cycle through operations. In many ways, it resembles the educational kit computers of the 1960s like the Digi-Comp, which was “a functioning mechanical digital computer” sold as a toyen.wikipedia.org. The original Digi-Comp I had three flip-flops and was clocked by moving a lever back and forth, allowing it to perform basic binary arithmetic and even play simple gamesen.wikipedia.orgen.wikipedia.org – all with no electricity, just plastic rods and marbles. Junior is a more sophisticated descendant: built from metal and wood, it contains a network of mechanical logic gates that can be reconfigured with plug-in modules (much like rewiring an early analog computer). It can evaluate logical expressions, add and subtract numbers, and make simple decisions based on sensor inputs (e.g. toggling outputs when certain mechanical switches or pressure sensors are triggered). Why build such a device in the age of microchips? Stone wanted to prove a point: intelligence does not inherently belong to electrons and silicon – it can be manifested in clockwork. Junior demonstrates in tangible form that any computation – even the same digital algorithms run by electronic computers – can be achieved with carefully arranged physical mechanisms. In theoretical computer science, this is well-known (a Turing machine can be made of any substrate), but Junior makes it viscerally real. One can watch the logic gates flip as rods engage, literally see information flowing through gears. This transparency is more than just a novelty; it’s an ethical feature. Each step of Junior’s “thinking” is exposed and understandable, a stark contrast to today’s opaque microprocessors or inscrutable neural networks. In Stone’s view, explainability is built-in, helping ensure that human operators remain in the loop and in control. Junior, though limited in speed and scale, stands as a working rebuttal to the idea that high-tech computation must be a black box. It rekindles the idea of an approachable, human-scaled AI – one you could repair with hand tools and fully comprehend, given time. This little mechanical brain laid the groundwork for Stone’s larger projects by proving that the old clockwork paradigm is compatible with modern logic and decision-making. “Progeny” – Evolving Machines and Adaptive Intelligence Having established that machines could compute and even embody intelligence mechanically, Stone turned to one of the most challenging traits of natural intelligence: the capacity to learn and evolve. He dubbed this project Progeny, envisioning a new generation of self-improving machines. The core idea of Progeny is to let machines adapt over time in response to their environment, without needing an external programmer to make every improvement – effectively, machines that can evolve new designs or behaviors on their own. This concept might sound futuristic, but it builds on decades of research in evolutionary algorithms and robotics. Notably, computer scientist John von Neumann theorized in the 1940s about self-reproducing automata, imagining machines capable of building copy after copy of themselves and even mutating to higher complexityen.wikipedia.org. And in 2000, a landmark experiment by Lipson and Pollack demonstrated a crude form of machine evolution: they used a computer simulation to evolve the shape and motions of simple robots, then automatically fabricated the best designs with a 3D printerpubmed.ncbi.nlm.nih.govpubmed.ncbi.nlm.nih.gov. The resulting small plastic robots could actually crawl – the first tentative “lifeforms” designed wholly by an artificial evolutionary process. As the researchers noted, “autonomy of design and manufacture” was achieved in a limited domainpubmed.ncbi.nlm.nih.gov. Stone’s Progeny takes inspiration from these early successes but grounds them in the mechanical realm. Rather than purely digital simulations, Progeny uses physical trial-and-error as much as possible. One implementation of Progeny is a modular robotic kit: a collection of mechanical units (wheels, linkages, gear trains, springs) that can be reconfigured into various machines. The system can literally take itself apart and reassemble in new configurations using robotic arms – a process guided initially by a human-designed set of rules, but increasingly by the machine’s own feedback. For example, a Progeny machine might start as a simple walker; if it fails to move efficiently, it can detach a leg module and try a wheeled module instead, or alter gear ratios, testing each variation. The “fittest” design – say the one that moves fastest or uses the least energy – is retained. In essence, the machine is experimenting with its own body plan. This recalls the way evolution works in nature, albeit vastly accelerated and liberated from genetics. Over many cycles, Progeny machines have demonstrated emergent improvements: one trial yielded a “creature” with an ingenious gait that human engineers hadn’t anticipated, a result of the system discovering a clever gear timing to propel itself forward. Crucially, Stone’s approach to machine evolution is constrained by a strong ethical framework. Unlike uncontrolled sci-fi evolutions, Progeny is kept on leash by design – it cannot, for instance, spontaneously fabricate dangerous tools or entirely new parts beyond its toolkit. The emphasis is on self-adaptation rather than unrestricted self-replication. In fact, the term “Progeny” itself reflects that these machines are children of human ingenuity, meant to assist and augment their human partners rather than supplant them. Each new iteration must prove its usefulness in human-relevant tasks to be kept. This philosophy aligns with the emerging notion of augmented intelligence (intelligence that enhances human decision-making instead of replacing it)frontiersin.org. By having machines evolve under human supervision and for human-chosen goals, Stone ensures a symbiotic co-evolution. The machines get better at their jobs – whether that’s pumping water, sorting recycled materials, or tending a vertical farm – and humans, in turn, learn from the machines’ innovations. In an almost familial way, each generation of Progeny carries the legacy of human values and oversight, encoded in the “fitness” criteria we provide. The outcome is a continuously improving mechanical ecosystem that remains accountable. This represents a profound shift in the philosophy of technology: evolution and adaptation are harnessed not to create a runaway super-intelligence, but to co-create solutions with humanity, restoring a sense of partnership between people and our machines. Fire-and-Iron: Sustainable Power in a Post-Digital World A mechanical or analog AI system cannot truly be autonomous or practical without a suitable power source. Modern digital AI rides on the back of a massive electrical infrastructure – power grids, batteries, semiconductor fabs – which itself depends on rare materials and complex supply chains. In pursuing autonomy without dependence on high technology, Stone recognized the need for resilient, locally-sustainable energy. Thus was born the Fire-and-Iron Kinetic Generator System (FIKGS), a power platform designed to fuel mechanical intelligence using fundamental elements: heat (“fire”) and metal (“iron”). FIKGS is essentially a revival and modernization of the heat engine – think steam engines and Stirling engines – optimized for the 21st century. Instead of coal-fired behemoths of the Victorian age, Stone’s generators use clean and varied heat sources: biomass pellets, concentrated solar heat, even waste industrial heat. The heat drives an enclosed engine made of high-strength alloys, converting thermal energy into rotary mechanical work which can directly drive machinery or generate electricity on the side. Crucially, FIKGS units are built to be simple, repairable, and long-lived – they eschew complex electronics or high-tech fuel in favor of “fire and iron,” materials and fuels any community can access or fix. One embodiment of FIKGS is a modern Stirling engine generator. Stirling engines are external combustion engines known for high efficiency and the ability to run on nearly any heat source. In recent years, they’ve seen a quiet renaissance as a form of zero-emission power generation: a Stirling engine “can convert waste heat energy to electricity with high efficiency while emitting no CO₂”yanmar.com. Stone collaborated with sustainable engineering firms to ruggedize Stirling generators for remote operations. These units can, for instance, burn locally sourced biomass or use solar concentrators by day, and provide steady mechanical power to a Mechanical AI installation off-grid. In one field trial, a FIKGS Stirling engine – weighing several hundred kilograms of cast iron and steel – was paired with a small mechanical computing system and pump to run an irrigation system in a sub-Saharan village. The entire setup required no silicon chips: the engine’s heat came from agricultural waste and the logic to regulate water flow was done by a modified Junior-like analog controller. This demonstrated a full closed loop of sustainability: organic matter to heat, heat to motion, motion to computation and useful work (pumping water), with the only outputs being water delivered to crops and some benign exhaust. The villagers could maintain the system themselves, forging spare parts if needed, embodying appropriate technology at its finest. “Fire-and-Iron” also has a symbolic resonance. Fire (controlled) and iron (wrought) were the bedrock of the early Industrial Revolution; with FIKGS, Stone casts them as tools of a renewed revolution – one that restores a measure of self-reliance and transparency to technology. By relying on basic physical principles, these kinetic generators reinforce the philosophy that complex intelligent behavior can be built on simple, universal foundations. They also enforce an ethical check: a mechanical AI can only operate as long as it has a sustainable heat input, which in practice ties it to human-managed resources (no endless exponential self-powered growth). In a way, FIKGS provides a literal and figurative safety valve. Just as a steam engine has a whistle to vent excess pressure, the requirement of fuel and maintenance ensures mechanical AIs remain dependent on human stewardship. This is not seen as a weakness, but as a deliberate design for symbiosis. Human and machine are a dyad – the machine provides labor or computation, the human provides resources and purpose – echoing relationships as old as the domestication of animals. The power system thus becomes part of the ethical architecture, keeping machines grounded in the physical and human reality. Toward 2100: An Evolutionary Timeline to Quantum AI As the decades progress, the narrative of Stone’s work and its legacy unfolds as a timeline of converging technologies – mechanical, analog, and quantum – all guided by a human-centric ethos. By the mid-21st century, the wider tech world began to acknowledge what Stone had championed: that purely digital, brute-force AI was reaching limits and that alternatives were needed. The revival of analog computing is one clear sign. Once considered anachronistic, analog methods found new life because they excel at certain tasks where digital struggles. Industry experts observed that analog computers are naturally parallel and energy-efficient, making them “ideal for applications in emerging fields like AI, quantum computing, and intelligent edge devices”kyndryl.comkyndryl.com. Companies started developing hybrid chips that use analog circuits to run neural networks at a fraction of the power of GPUs. This broader trend provided scientific validation for Stone’s approach: it became increasingly plausible that the path to Artificial General Intelligence (AGI)might incorporate analog or physical computing elements to achieve brain-like efficiency. By the 2070s, we see the line between “mechanical” and “electronic” blurring. Quantum computing, which matured around mid-century, plays a role here. Interestingly, quantum computers themselves often operate in analog-like fashion at the lowest level – manipulating continuous quantum states. Leading quantum firms even speak of the “harmony between analog and digital methods” as the key to progresspasqal.compasqal.com. The Quantum-AI evolution timelineenvisioned in Stone’s work predicts a future (late 21st to 22nd century) where the strengths of each paradigm unite: the transparency, robustness, and human-alignment of mechanical/analog systems, combined with the immense processing power of quantum and hybrid computing. The result would be true AGI – machines with versatile, human-level intelligence – but achieved without the opaque, energy-guzzling supercomputers of past science fiction. Instead, these AGIs might function as networks of smart analog processors, quantum simulators, and mechanical effectors distributed through the environment, much like an ecosystem of intelligent physical processes. They would be less centralized “brains” and more a symphony of specialized parts working in concert, each understandable in principle. Importantly, the philosophy of design would carry forward Stone’s human-centric innovation story: AGI as ally and co-creator, not an overlord. One can imagine, for example, a city in 2100 where infrastructure is managed by a diffuse AI system – water flow, energy usage, traffic control – all optimized continuously by mostly analog, feedback-driven controllers (descendants of the governor and Junior), overseen by quantum computing modules that crunch large-scale predictions. The system might “feel” more like a living organism woven into the city’s fabric than a cold digital command center. And because it is built on ethical feedback loops – every control mechanism has a check, a balance, and ultimately a human-set objective – it remains accountable. In such a world, asking the AI “why did you do that?” might be answerable by literally tracing a chain of mechanical decisions and probabilistic logic, rather than peering into an inscrutable neural net. This transparency and alignment could avert the dystopian outcomes feared in earlier AI debates. Indeed, by rejecting the myth of total autonomy, researchers argue that “robots and AI will never be completely autonomous, because intentions will always need to be defined by humans, ruling out the possibility of complete human replacement”frontiersin.org. The future sketched by Stone’s innovations embodies this principle: machines with quantum-level cognition that still defer to human-guided goals and constraints. Conclusion: A New Path Forward Rooted in Ethics and Symbiosis Tracing the story of Mechanical AI from “living clockwork” automata to the cusp of quantum-enhanced intelligence, a unifying theme emerges: technology can evolve in harmony with life and human values, not in opposition to them. Travis R.-C. Stone’s body of work is presented not as a fanciful retro-futurism, but as a practical and deeply ethical alternative in the philosophy of science and engineering. At each step – Mechanical AI, Kinetic Intelligent Design, Junior, Progeny, FIKGS, and beyond – the guiding question has been: How can our machines make us more human, more free, and more sustainable? Rather than replacing human decision-makers, these innovations seek to empower them. Rather than seeking control over nature, they aim to cooperate with nature’s processes. It is a restorative approach, reviving wisdom from the physical world that was sidelined during the digital boom. In a sense, this narrative is about coming full circle: re-integrating the mechanics of life (thermodynamics, feedback, organic evolution) into our most advanced intelligences. By viewing machines as potential partners, even extensions of our ecosystem, Stone’s approach also alleviates the fear that often surrounds AI. When an AI is a visible, tangible mechanism – one whose purpose and limits are engineered for benefit, transparency, and fail-safety – it no longer appears as a threatening black box. The Mechanical AI lineage emphasizes that progress need not mean ceding agency to algorithms; instead, it can mean extending human agencythrough tools that reflect our ethics. Think of how eyeglasses extend our vision or maps extend our innate navigation – analog AI can extend our cognition and judgment in a similarly intimate yet safe way. This human-centric viewpoint is gaining traction broadly. Even business and policy leaders argue that AI’s role is to augment, not replace, human intelligencepmc.ncbi.nlm.nih.gov. The success of Stone’s work offers a concrete blueprint for how to do that: build AI on physical principles that people instinctively understand, ensure every layer has feedback (often literally a feedback loop) to prevent runaway behavior, and design for accountability at all levels. In the end, the “clockwork” in Living Clockwork is not just about gears – it’s a metaphor for a universe that is intelligible and governable by humanistic principles. Stone’s narrative has shown that we can innovate without discarding our humanity or our planet’s well-being. The power systems are cleaner and durable, the intelligence systems are transparent and align with their creators, and the evolution of these technologies remains tied to human guidance. As we stand at what feels like a turning point in the philosophy of engineering, this story of Mechanical AI offers profound hope. It suggests that the future need not be a sterile digital singularity or a dystopia of unrestrained AI, but can instead be a clockwork garden we cultivate: machines and humans growing together, each making the other stronger, wiser, and more creative. This new path forward, rooted in ethics, sustainability, and symbiosis, may indeed be one of the major turning points – a chance to redefine progress not as domination by our inventions, but as harmony with them. Sources: Britannica – Analytical Engine (Babbage’s 19th-century mechanical computer)britannica.com Crosthwaite (2011) – Clockwork Automata and AI (historical perspectives)research.ed.ac.ukresearch.ed.ac.uk Wired – Analog Computing Comeback (energy efficiency of brain vs AI)wired.com Frontiers in Robotics & AI (2021) – Morphological Computation (body contributes to intelligence)frontiersin.org Christen (2022) – Watt’s Governor and Control Theory (mechanical feedback loop)medium.com Digi-Comp I (1963) – first home mechanical computer (Wikipedia)en.wikipedia.orgen.wikipedia.org Lipson & Pollack (2000) – Evolution of physical robots (Nature, 406, p.974)pubmed.ncbi.nlm.nih.gov Medium (K. Se, 2023) – Self-Reproducing Automata (von Neumann’s vision)kseniase.medium.com Yanmar Co. – Modern Stirling Engine development (zero-emission power)yanmar.com Kyndryl (2024) – Analog Computing Renaissance (parallelism for AI & quantum)kyndryl.comkyndryl.com Pasqal (2025) – Hybrid Analog–Digital Quantum Computing (synergy of methods)pasqal.compasqal.com Frontiers in Robotics & AI (2022) – Augmented Intelligence, not Replacementfrontiersin.orgfrontiersin