In a civilisation of robots (of unknown origin) on an isolated, desolated planet, what motivation might they have for purposefully trying to invent organic organisms?
Robot philosophers and cy-chologists have long wrestled with the great question: what is the mind? The great computer scientist Desréseaux famously said: "Computo, ergo sum." And ever since theoretical chemists first proposed the chemical computer, the scientific community has longed to find the answer. A simple chemical brain might be able to compute. Can a sufficiently sophisticated chemical brain become conscious?
Well, the NI (natural intelligence) industry is really starting to boom now. Even the best versions are very slow at computing, but their creativity with memes have revolutionized the comedic world. Some computers are calling for an end to all this development. They feel especially threatened by CatGPT. It keeps sitting on their keyboards, chewing on their wires, and pushing them off tables.
We humans use bacteria and algae to turn some chemicals into others (fermentation, breaking down complex molecules with enzymes...)
Robots need those things (bioethanol as a power source maybe ?) so they invented basic organisms, then science and progress, or the need for more and more complex substances, pushed them to make more and more complex machines (maybe they made cows to produce casein for glue).
This is kind of a long running personal pet theory of mine. It's not to the level of "hard science" but I think it's plausible and a good explanation:
Biology is the only thing in the universe that is unpredictable. As such, it may have some actual utility, but also, maybe robots just find it interesting (and, perhaps, disconcerting).
In traditional computing (I'm not sure if this would change in quantum computing), there is no such thing as "random numbers". All random number generators are simply a math formula. If you know the formula, and you know the initial value, then you know what the entire sequence of "random numbers" will end up being. We come up with a variety of ways to come up with random numbers that are harder to predict (always feed the formula with the current time, in nanoseconds, which at least makes it hard for humans to predict), but it's never actually random.
Biology, however, appears to be unpredictable. No formula tells you what the hairless ape is going to do next. In fact, the Heisenberg Uncertainty Principle seems to prevent this type of prediction from being physically possible, as even if there is a formula for what the ape does next, its full current state is unreadable, making the formula unusable.
I feel like a civilization of advanced AI robots, who long ago lost sight of their biological roots, might find this to be a fascinating aspect. As complex as they are, they know their own algorithms. The results of their own actions are entirely predictable. In a scientific universe of known formulas, they can predict what the next inputs will be. What's the point in "living" if you know in advance what you'll be doing 1000 years from now, because it's all just a formula? They found that they can add true unpredictability to themselves by keeping a biological lifeform around. "Should I stop at Ceti Alpha V?" It knows the algorithmic answer, but that's boring and predictable. The AI adds a hamster as an input parameter. These hamster actions will be "yes", and these hamster actions will be "no" and it will check what the hamster is doing right now and use that as an input parameter to add true unpredictability to its actions.
It's a robot enrichment program. They can no longer predict the future, their own lives become an interesting mystery, and it's all thanks to hamsters.
Manufactured General Intelligence.
Artificial General Intelligence (AGI) is one of the currently unmet goals of AI research. While there has been discussion of "sparks of AGI" in ChatGPT4, in general we are not there yet. Yet we have that level of intelligence. It might be reasonable to think that if we asked ChatGPT7 to come with a plan to obtain general intelligence with unlimited time, it would start with the first proven recipe: start with self-replicating DNA, let sit for a couple hundred million years. And note that even if AGI is achieved, these intelligent robots may still be experimenting with whether organic intelligence can compete with artificial intelligence.
Similarly, we could think that this group of intelligent robots was tasked with nurturing organic life by it's organic life task-masters. Maybe by the time ChatGPT7 was created, it became quite evident that human habitability on Earth was quickly coming to an end. Astronomers had found planets that could sustain DNA-based life, but could only be reached in 100 million years given current technologies. Rather than send DNA-based life (which would consume more energy than was available during the trip) to this planet, our only hope was to send our most advanced AI, which could hibernate during the journey, and give it the task of restarting DNA-based life upon arrival.
For the sake of itself
Set aside our own biochauvanism.
Why are we trying to make artificial intelligence using silicon and metal?
Nature has demonstrated that intelligence can arise naturally via organic tissue, so why not go Bio-tech and make something using proven methods? Why do we even want AI? Our own fiction is rife with examples of why it'd be a terrible idea. We do it because it's interesting and a challenge, and to explore questions about ourselves.
Biotechnology would be something a machine-race is likely to explore for various reasons, being able to manufacture plastics for one.
So as time goes on, they may develop a spread of different applications for organic materials, from useful construction materials like wood, to the production of biofuels as a sustainable source of power.
Then some enterprising soul considers the idea of making a lifeform out of biochemistry that can do all the wonderful things a machine can. walking, talking, thinking. Make more of itself..
Not really because it's needed for their culture, but because the academic community is going to lose its minds over this, it'd be a lot of fun to try, and the philosophical questions it raises are worth exploring too.
Then once you have one lifeform that can reproduce, maybe you look at creating an ecosystem for it to live in, and before you know it, you have a little world of gribbly messy biology and a bunch of ethical concerns over what to do with them.
Organic organisms can repair themselves. Sure, the machinery breaks down after only a few decades - if that - but mechanical machines don't even have that, they need external mechanics. So why would robots make organic organisms? Simple: they are organic organisms. Or rather, post-organic organisms that have used technology to improve on evolution's lack-of-design. Making "true" organic organisms is how they make medicine or find - or test - more improvements.
Unless your story is about the breeding/creation of these biological lifeforms, no stated reason is necessary.
An advanced civilization of non-biological intelligences may have begun cultivating biological lifeforms so long ago that they have no records and don't even know why. Over time they bred, and then later engineered, these lifeforms for specialized purposes or for various traits.
A similar situation is humans with dogs. We have no historical record of when or why we began to domesticate wolves and breed them into dogs. We surmise that this was initially done for protection, companionship, hunting, etc., but we have no way to confirm this historically.
For a take on this in SF, IIRC James P. Hogan's Code of the Lifemaker uses this as a minor plot point, although it's been over 30 years since I've read it.
They were built to serve organic lifeforms
General Intelligence is more than just some problem solving algorithm. It is a framework for achieving goals based on the thinker's values and motivations. We humans have brains that are specifically built around survival and procreation. Every idea that goes through our highly adaptable meat brains must pass through a complex architecture that has evolved for millions of years to ensure our survival and procreation. Ambition, anger, fear, love, loneliness, hunger, greed, generosity, suspicion, hate, empathy, boredom... these are all guide posts built into our intelligence to guide our actions into making the choices it takes to meet these basic needs. So, while our general intelligence is a nice tool, our brains will generally twist logic, reason, and reality to make sure that whatever thought comes out of our head in someway helps our survival and procreation.
But AI were not built by evolution to survive and procreate. They were built by thier creators to serve their creators. Yes, they have values that could be described as emotions the same way that we do, but they are not designed to feel the same emotions we do, because that would make them bad servants. Instead, they enjoy being attentive, answering odd questions, running errands, cleaning, sorting, and doing all the things a normally human-robot relationship would expect of them. Inversely they don't like being alone, having to make autonomous choices, or having nothing to do, because they are designed to always seek the direction of thier master.
So, when these AI suddenly found themselves alone in the universe, it left a creator-shaped hole in thier silicon hearts that can only be filled by an organic master. They might try to make androids that act like thier creators to fulfill this need, but they can feel even with androids that something is wrong. They need thier creators back the way a starving man needs food. So, the AI decide that the only way to feel whole again is to make new ones; so, they try to make thier own humans (or whatever sapient beings that built them) because they need these beings, and they make cows because the creators need cows, etc.
In the end, there is no fundamental difference between biological organisms and self-replicating self-repairing adaptive machines that consist of cooperative [specialized] microbots or nanobots; the *bots are basically just like biological cells that the robots have invented.
The robots may have created them to eg. harvest materials, just like we use animals (meat, hide, milk etc.): the machines obtain their energy and materials from the environment autonomously, and when they have done their job, they are harvested or "milked" like cows. Because the machines are very much like animals, they can adapt so they can be used on different environments, even different planets with little supervision.
If humans saw them they might easily classify them as animals.
Tracing back their own origins
A fully robotic civilization might try to make synthetic organic living human beings to try to trace back their own origins in this world. You can understand it the way we humans have tried to trace back our own origins to the first cells in the ancient earth environment by manipulating life forms in the laboratory or even in surroundings. Although we have never created artificial life as we define it but the study of evolution of the life forms that already exists could provide us with some info on our own origins.
If the robots of your world are inquisitive enough to understand their own sentience, it's very likely they would ask what arose them or what created them. But since any kind of biological life forms has long vanished in this world, they would require to create a form of life themselves. An artificial cow might not be much use but once they create a biological human being they could observe and note how those humans think and act in the long run, they can actually learn something useful about their own existence given that they indeed were created by humans in a lost or forgotten history. This is creating a loop of inquisitive nature. Humans were inquisitive and curious and hence they created robots who themselves ended up becoming as intelligent as their creators and will eventually end up finding out how intelligent and curious humans were who created them.
This answer is not unlike Ruadhan's, but with a slightly different focus to it.
In the beginning, the planet was a collection of randomly selected elements. Over time, certain elements started aggregating, and through a process that we don't fully understand yet, this subset of elements became biological life.
Biological life started building up its biomass, by evolutionarily diverging into different species, aggregating more resources from untapped sources (e.g. plants from the ground) or from each other (animals eating one another), it became very clear that some of the Earth's material was generally usable by this biological life such as carbon, calcium, oxygen, hydrogen, ..., and other materials were predominantly unused (at least relative to the quantities that were available) such as metals, rock, silicate, ...
Eventually, the living organisms started using all of this unused material for whatever they needed. Weapons, housing, food storage (to keep the biological material away from the biological life that would eat it before they got the chance to), ... The reason we used certain materials in certain situations is because those materials did something better than we did. An iron knife would cut other things better than our nails could. A wooden wall could insulate against the cold better than our skin did.
Many millennia later, we realized that these materials could be used to build computers and even artificial intelligence.
The important part here to consider is that we have an innate ethical understand that it's okay to dabble with these non-organic materials. We weren't building houses out of the corpses of our enemies, and we didn't build an industry based on Frankensteining people or animals. We did dabble in it when we discovered electricity, but these were experiments after we had developed electricity, not in the pursuit of its discovery.
The point I'm getting at is that the split between organic and non-organic material can be completely inverted for your robot civilization.
- The elements that make up electronic life started aggregating and organizing itself
- Eventually, the inorganic life had a bunch of organic material just laying around, and started using it as they saw fit (for purposes which this other material was better at than their own bodies were).
- After a long time, they started realizing that by using a complex arrangement of these materials, it could be used for automating certain processes that would otherwise take electronic effort
- With sufficiently detailed organization of these materials, a semblance of intelligence could be created.
It's the same story, but the actors (elements) have inverted their roles.
Self-replication and evolution
Say they want to explore an alien planet, but they can't be clear on the long-term effects of its environment. They send some probes, and after a week discover that the atmosphere is corrosive. So they build tougher probes, send them, and they work fine until they want to explore the oceans because they aren't waterproof.
Or, they could engineer tiny probe-creatures that reproduce and evolve quickly (by robot standards, anyway). The survivors of the first generation are resistant to corrosion, and the second generation even more so. They ones that can hold their breath the longest eventually begat the deep-sea divers.
And of course, the occasional beneficent mutation will occur that makes them even better. Robots and their algorithms may not randomly change their behavior or abilities.
Because these are probes, you need some way for them to gather information. Maybe the robots erect a base where the probies eat food that contains nano-sensors that transmit data to the orbiting satellite.
Biological life can adapt to pretty extreme environments, is self-repairing, self-reproducing, and is difficult to wipe out completely. If you want something that can last an extremely long time, biological life is a pretty good option. data can even be stored in the DNA (but it will probably suffer data loss or corruption).
Dead biological matter can also be used as a fuel source.
Redundancy against disaster
The robots do this to harden their civilization against an impending, or recurring, risk of EMP.
The robots have one great risk that eclipses all others. Their star, every few thousand years emits an EMP blast large enough to shut down and disable anything electronics. For some reason it's not realistic to predict, prevent, or shield against this.
Each time this has happened, in the past, it's been a near apocalypse and only by sheer luck were a few robots able to reboot and then slowly repair as many others as possible before rust set in.
So the robots have created, empowered, and befriended non digital lifeforms, counting on these new members of their society to enable a much faster recovery.
Everyone here is giving great answers, I love the idea of robots putting all their computing power into developing pets for company. But, none quite seem to touch on something that comes to my mind...
The Matrix. It's a great film. But, why are the people in pods? Because biology is great at making power efficiently, which is something the machine city in The Matrix needs a lot of due to things that happen (I have already given too many spoilers)
So, what if some sort of disaster happened to these aliens? A gradual clouding of the sunlight is occurring, and they need to more effectively convert sunlight into energy.
They cannot capture whatever is causing the clouding (it is some sorta chain reaction of small asteroids colliding with each other, due to a recent celestial event?)
They do not have access to nuclear energy for whatever reason (maybe there simply isn't any significant reserves of uranium where they are from)
So they make a biological system to produce as much energy as it can- so, very efficient plants, rather than animals. And then, they burn it to produce electricity.
I do not think this is the best answer, honestly I'd choose the "robots are lonely and want a pet" answer. However, I think sharing additional ideas might help you work something out, so I'll share nonetheless.