You know what? I've been thinking pretty heavily on this one. If a robot could actively consume energy from its habitat(like BEAM), actively respond to its surroundings through sensory interaction(like BEAM), adapt its behavior in response to changing circumstances(like some BEAM), self-replicate in part(unlike any robot), grow and develop independently after initial construction(again, unlike any robot), and contain internal coding that helps control how the machine does all the preceding(like digital robots), wouldn't it technically be alive?
In a way, BEAM is actually less life-like than a digital robot, not in terms of behavior, but in that not all life forms have a nervous system, but all life forms, and even some non-life forms, have some form of internal coding, which BEAM lacks. Which is why a digital-BEAM hybrid might be as life-like as it gets short of reproducing itself.
I've wondered why a replicator wouldn't be considered alive. And it's because a traditional replicator can do everything except grow and develop. Once it is completed, it stays that way for the rest of its working life. Which is why I specify that a truly living machine would only replicate itself in part, but enough that the partial replica could function enough to finish itself. This way each individual continues to develop itself to completion independently of its parent, thus being alive.
I'm not saying it would be self-aware or anything like that. It would be more analogous to a single-celled organism than to multicellular life. I don't understand why people think a machine could never be alive. They think that since a robot is not made of living cells, it can't be alive. Well, neither is a living cell, yet it's alive. Just like a cell, the living machine is not made up of living components, but rather, it itself as a whole represents the single, indivisible unit of life. Because a cell is construed of non-living components as well. The nucleus isn't alive on its own, nor the mitochondria, nor the ribosomes, nor the membrane, nor any organelle. But they work together as a unit of life. Just as much can be said about the living machine. Its processor isn't alive, nor its motors, nor its sensors, nor its power system, nor the reproductive equipment, nor any other single system or component that make up the robot. But they still work together to form that indivisible living unit. I think the confusion arises from the fact that people are used to indivisible units of life being microscopic, whereas a robotic unit of life likely wouldn't, since that would be much cheaper than nanotechnology. And a living machine could also be instructed to manufacture something aside from itself, such that different social tiers could be composed. The highest level being granted to reproduce more bots, while a lower tier works to manufacture and sort any commercial or industrial products they might be instructed to, as well as industrial product, and the lowest tier gathering and distributing resources, and commuting products to a shipment point.
And now for another one of my famous unrelated side notes.
I might try to determine the formula that drives the behaviors observed in HexBug Nanos. I've seen everything to obstacle avoidance to social communication to threat response. Heck, I've even observed mine display a phobia to heat and light. And they do all of this without any sensory or control apparatus. I have a feeling that they are, in fact, responding to their own vibration that their environment distorts and resonates back through their bodies. When placed in an open area, they seem move about radially in a roughly cosine pattern, until they hit something. But as the vibration is transmitted through the ground, they seem to be able to respond to the vibration patterns emitted from other individuals, and this must result in the "playful" and "assisting" behaviors they appear to emulate. This probably also explains why they follow each other around occasionally. As for the thermophobia I've seen them exhibit, it likely has something to do with the exposed area having expanded slightly and thus conducting vibrations differently than darker areas, evidently in a way the bugs don't quite like. It likely may also have to do with the expansion and contraction of their silicone outer bodies. I tested this once by heating the stove and placing the bugs on the counter right next to it, facing into it. Every single time I tried this they always pulled away vigorously, and went about in every other direction except toward the heat. In another instance, I went out onto the back patio at night with the floodlight on, and placed several down. They generally went about in all directions, but they seemed to prefer to stay in the shadows, because whenever any one went into a lit area it immediately retreated back into the dark. I'm not entirely sure how they could sense the light, but it is obviously different than being repelled by intense heat. There's no way light from an incandescent flood lamp could have a strong enough effect on any material it shines on to alter the way it conducts kinetic energy.
They also seem to have some measure of group decision making power. Also, as their legs wear in, they seem to develop a memory of repetitive patterns, thus exhibiting a measure of learning(which is probably why my old, beaten down ones seem to be wiser than my newer ones). And when I put them in a lego house, they always went straight for the door(still have no idea how they managed pathfinding like that). So somehow they have achieved some sort of "cockroach intelligence" without any control apparatus whatsoever. Or at the very least, something more like "slime-mold intelligence".
Okay, so perhaps this actually does have something to do with the original topic. If I can extract an equation that defines and predicts accurately how the nanos behave, that also gradually modifies itself according to how the a real nano would adapt, it could be extremely promising for BEAM, because all an analog circuit really does is emulate a single equation. So it can be thought of as a non-universal computer. Depending on how intricate that equation is, an analog circuit may very well outperform a microcontroller toward a dedicated purpose. A single circuit could serve as the entire control system for the machine and the machine could exceed a programmed one in performance. If I get the formula I'll likely expand on it greatly, in order to accomplish higher order behaviors. This sort of circuit would be a great control system for a living robot, such that the digital memory would only serve as a template for development, replication, and repair. Perhaps instead even another analog circuit could be used to guide the replication process, and a closely associated one for development and repair. Of coarse, I'm no longer talking about using a digital system to guide the machine's operations. Now we're talking about a true brain. A very simple one, but intricate enough to guide the machine through complicated instinctual processes, like replication and feeding, and measurably cognitive processes like learning and pathfinding. The only problem is whether it could still be considered alive at this point, because the digital internal coding has been replaced with a brain that emulates the same functions. I suppose the connections in the brain could be considered its coding, but at this point the machine would be less comparable to a cell, since it uses a sub-living brain in place of an equivalent nucleus. But why does it have to be comparable to a cell?