Not Nano-Bots, Nor Androids, Nor Cyborgs — but Golems
~ future structures, equipment, and robotics will be *infused* with Life ~
TL;DR — We eventually end-up organizing our industries rather efficiently, even when our initial attempts are zany or wasteful. Far-future robotics and construction, when done efficiently, will *not* consist of cyborgs, or androids, or nanobots, because those are not as effectual and reliable as Golems. Golems? Yes, dead materials like clay, wood, and stone, crystal and metal, when they are punctured all-through with microscopic capillaries and rigged with microscopic sensors, power supplies. They appear to be normal-sized ‘robots’, yet with details in each part of their body down to the cellular scale — and all that micro-machinery will be maintained and operated by custom-designed *bacteria* and fungi, because living cells are the best chemistry-set for their cost and availability. Crude mounds, knit by filaments, will see their world by using their whole, bare face itself to collect light, and move their limbs as carbon-fiber tendons twitch in each pumice’s pore. And, these Golems will be *assembled* to-specification for their unique tasks, without being grown or self-replicating. Each is a unique monstrosity.
“Batteries as Exoskeleton”
I waited years to see researchers pursue a natural optimum of robotic power-system design: while standard practice has been to give robots a skeleton or shell, with all its own weight that it must carry, AND add a battery-pack to power the entire system… instead, there have now been developments of Batteries-as-Skeleton! The battery-material is sturdy enough that you can make arms and legs out of it, which avoids your battery-weight by absorbing it into the skeleton-weight. Suddenly, the robot operates longer, has a higher power-to-weight ratio (making it more effectual in its environment — jumping higher, for instance), and there are fewer parts, which generally lowers cost of mass-production. I do expect future power-systems to far exceed current batteries, yet this way of looking at robotic design is the key insight for understanding “why Golems, instead of nanobots or cyborgs?”
Another facet of this perspective on design: self-replication. In nature, we see self-replication because… that was the only option. In contrast, *no* machine self-replicates. It requires tools that it cannot produce on its own. And, for that same reason, there are *no* companies which self-replicate! (Even Musk’s desire for “in-house” production is not complete — his own company can’t *make* the Giga-Press itself, for example!) Every business has suppliers. None self-replicate. And, the reason none of the businesses on Earth are trying to self-replicate is because: self-replication is slow, risky, and wasteful compared to mass-production of parts on specialized equipment, and further assembly.
Imagine: a factory that makes plastic bottles needs a few hundred different small machines, stamps, tool-tips, devices and sensors… and, once it has those particular components, then that factory can spit-out 100 soda-bottles per minute! Yet, the boss says “we’d be better, in some vague way I can’t measure, if our factory could self-replicate! Do that!” So… they now need to be able to *create* those different tool-tips on every machine, on their own… which requires other tools! They buy those other tools, but those additional tools are also components that they’ll need to self-replicate, so they have to get the tools that make *those* tools… Very soon, the boss has purchased *most* machines in existence — and his factory has NOT increased its bottle-production at ALL! What a waste!
No one bothers with “industrial self-replication” because such a system needs to keep every tool on-hand, even though it *rarely* uses each tool. That’s called a “low capital-utilization” and it’s a sign of doom for a business. You want all machines being useful all the time! A robot that can do every kind of task is actually *burdened* by all the features you had to include! Robots make more money and get more done when they are hyper-specialized, as all industrial robotics demonstrates.
That pattern will hold in the future, because it doesn’t depend upon *which* technology you have access to. “But, if we had nanobots, then they could self-replicate!” Yes, if we had nanobots, they could self-replicate… AND, those same nanobots could get MORE done if they specialize in giant vats, each vat processing their own specific chemical precursor or component part. If we had nanobots, we would keep them in vats specialized to their task, because that has higher throughput per dollar. Insisting that a nanobot self-replicate is insisting that it slow-down to do tasks that are better-done by a different specialist.
And, as chemists keep saying to futurists: your nanobot’s fingers won’t work at grabbing chemicals, without degrading that same finger-tip each time that you use it! The nanobots wear-down, chemically, by interacting. That same process happens in living cells — which is why so much energy is spent on cell-repair! So, chemists explain, if you want a nanobot that does chemistry and micro-mechanical fabrication and assembly, then you’ll want to give it an internal chemical repair-kit. The simplest kit, using abundant materials, for achieving diverse chemistry… is life itself. That’s the best nanobot.
But Machines are Better than Us
While bacteria and fungi are the supreme chemists and micro-fabricators, with comparatively ready, diverse mass-production, and safe programming of their functionality, the same is not true of macro-biology! When comparing robotics and machines to creatures our own size, then the machine’s capacity for *fast* and *precise*, *repeatable* motion and sensation makes them the clear winner. And, while our bodies are difficult to alter, robotics is plug-and-play. At human-body-scale, things will look like robots — and, given the progress in materials science, they will likely be made of ceramic or resin and fiber. (That is to say: made of clay, or pitch & flax.) At the micro-scale, they will be porous, super-structured, and teeming with bacteria!
The primary limitation to robotics for a few decades now, has been the *time* required for a stimulus-response. Whether a robot is autonomous, or teleoperated by a person, there is a *delay* induced when we try to do precise tasks in unstructured environments. The machine currently can’t think fast enough, or the bandwidth connecting to a human isn’t responsive enough! When wired-in to a teleoperator, roving robots at universities were able to clean-up coffee tables and fold laundry… twenty years ago! Once machine-thought hastens, all our old robotic forms will be plenty to outpace human performance. Our own bodies, tangled in the necessities of self-replication and growth-from-an-egg, can never compete.
A Floating Stomach
If microbes will rule the miniature-innards of Golems, while their person-scale bodies operate on familiar robotic principles, then what about the shape of factories and larger structures? Billowing, lobed, filled by twisting filaments that grasp at each component, held gently in a womb. A cloud made of jellyfish arms, chewing rocks and spitting metal into shape. Why? Structural engineering and logistics.
When we normally build structures, they take that shape because it’s easy for our tools and people to build. Those factories and skyscrapers are actually the opposite of good design, for doing what we intend —first, because they rely upon compressive strength, instead of tensile! Tensile strength is “how much force to pull it apart”, compared to compressive strength’s “how much you can stack on top of it before it gets crushed”. In general, the tensile strength of our best materials is dozens of times greater than our best compressive materials. Yup, that’s a HUGE increase in strength. What would that look like, as a structure, though?
Imagine an enormous ‘plastic tarp’ surrounding a dome that you’ve filled with air, pressurizing it compared to the air outside. The air inside is pushing against the tarp, trying to widen it and tear it apart — that is tensile strength, and that’s what we want. But, what about any weight held off the ground? Compressive strength MUST happen somewhere, to carry loads on higher floors. Easy: the column of air itself is a ‘pneumatic jack’, carrying the weight in its entire air-pressure down to the entire surface of the ground that it touches.
Yet, if you just sit on top of that inflated bubble, it will bow downwards and you sink quite a bit! When that happens, you’ll notice that the ‘imaginary lines’ drawn straight between points on the tarp start to get longer — the bubble is deformed. To absorb that force, we need MORE tensile strength, inside the bubble! They’re called ‘catenary cables’ when used in zeppelins and such, tugging on the walls from within in every direction. By pulling from all angles, within the dome, you resist bubble-deformation — you sink less when you sit on top of it!
Now, imagine those ‘catenary cables’ inside the pressurized dome are strung with zip-line bots hauling materials beside valves of fluids and gasses, with robotic arms dropping components into nets that drift beside them. Who needs a floor, when no humans walk there? Who needs separation by layers, then? The factory is a whole, thick web, extending in all directions without ceilings or walls, all the way to the pressurized skin surrounding it. Outside, it would look like an enormous marshmallow, quivering as the immense masses within were shuttled or hoisted-about. Inside the cocoon, sacks of microbes tailored to their task will be injected into waiting lumps of clay. Golem wakes.