When (not “time”, but at what force size/distribution) would my artificial life have enough presence to believably conquer humanity?

My story takes place during a war between two species: Humans and an organic AGI. The AGI are the unannounced aggressors. My world problem is deciding what reasonable force size and technology they would have had to put up a good fight. The big “ask” is at the bottom, here are the gory details about our combatants and ther objectives:

Details / Definitions section:

The Enemy

My artificial life form, given the species name Cerebrum Naturae, was essentially formed by humans and is built with human manufactured components except for its brain (fully explained below). The only part of the cybernetic organism which is actually alive is a bacterium in the brain, which is protected as an endangered species. Cerebrum Naturae are considered an AGI and have the following limitations:

  • They can die if it doesn’t “eat” or fuel its cranium. But destroying their cybernetic bodies and reducing their minds to mere bacterium is also possible; that bacterium requiring its extreme environment to survive.
  • They can reproduce (TMI for this question, but a parent AGI brain growing too large is too brittle and splits, forming multiple new young AGI which need to learn about life. The parent “dies” in this - memories are destroyed.
  • The natural life span of Cerebrum Naturae is approximately 25 years before their brain will fracture, however this is dépendant upon the bacterium reproduction rate and possible natural flaws in their brain lattice.
  • They fear for their existence. Although it is essentially a program, the data in its memory was not and can’t be programmed or duplicated. When the brain stops feeding, the memory is erased. It wants to survive, and division is the only possibility for this.
  • Other than it’s brain, all sensors and body components are cybernetics - manufactured prosthetics made by humans or laborer AGI. IOW, all body parts are replaceable, and they owe all “life experiences” to human technology.
  • They are vastly more intelligent than humans and generally work in engineering or scientific fields requiring heavy computational work.
  • They interface with humans via NLMs similar to modern chatbots, however they make their own decisions about curating data, and they can originate novel concepts via true random pattern generation.
  • They are each unique.
  • At the time of my story, they take care of themselves: They harvest their own organic waste, collect their own cranium fuel, and manufacture their own cybernetics. Although they still use human utilities, they could do this themselves.

Origins of Cerebrum Naturae

In the vast realm of microbiology, extremophiles capable of thriving in extreme environments, included a new extremophile bacterium that was accidentally discovered living within a unique magnesium sulfide ($\text{Mg}^{\text{2+}}\text{SO}_4$) crystal lattice, which contained semiconductor nodes and naturally formed memory circuits. Regrettably, this remarkable bacterium has never again been found and is thought extinct, leaving behind this colony as the only representatives of the species.

The Discovery:

In the remote and inhospitable terrains of an unexplored deep sea region, a team of scientists were on an expedition to study the microbial life thriving in volcanic vents. They stumbled upon a hot spring nestled deep within a ravine, and an unmarked sample collected during clearing harbored the astonishing secret microbe. Why they chose to analyze this debris is dumb luck. But the scientists noticed something peculiar underneath their microscopes. Intricate crystal formations unlike anything ever seen before. After testing, these crystals possessed a unique property – they exhibited semiconductor behavior and formed complex memory circuits. They have discovered the only natural computer circuit in existence.

  • They use advanced machine learning, and can interface with networks but don’t have unlimited memory for it, nor is understanding human data “easy” for them, as they are not human programs. But like current AI, programming languages are as easily learned as any other.

The Extremophile Bacterium:

Further investigation revealed that these extraordinary crystals were not mere geological formations but were intricately linked to the presence of this specific extremophile bacterium. This bacterium, named Solitarius memorialis, which possessed the extraordinary ability to catalyze the $\text{Mg}^{\text{2+}}\text{SO}_4$ crystal lattice with its metabolic processes.

The Memory Circuit Formation:

The most astonishing aspect of Solitarius memorialis was its unique ability to generate a crystal lattice structure composed of $\text{Mg}^{\text{2+}}\text{SO}_4$ with embedded semiconductor nodes. These nodes possessed the remarkable property of retaining and transmitting electrical signals, effectively forming memory circuits and even data busses.

The exact mechanism by which Solitarius memorialis achieved this feat was not discovered before the structure began repeating data signals back when stimulated, prompting the agency to cease any destructive testing. The sample became a protected endangered species.

The Extinction:

The hot spring ecosystem that housed this extraordinary bacterium was irreversibly disrupted during a minor eruption at the collection site. The only remnants of its memory circuits in the form of $\text{Mg}^{\text{2+}}\text{SO}_4$ crystals were this sample.

The Legacy:

Solitarius memorialis has opened up new avenues for research in materials science, nanotechnology, and bioengineering, but the only specimen in existence could not be directly studied. Scientists then attempted to replicate the conditions under which Solitarius memorialis thrived in order to replicate the memory circuit formation, and in the process several sensors and peripherals have been attached to the colony to test its memory capabilities. They discovered a natural tendency to “expel” memory collected through low impedance peripherals, essentially demonstrating a natural ability to form data storage and retrieval busses with external conductors of different impedances. But breaking off samples of the colony was strictly prohibited.

A new life form:

The story eventually caries us to a microbial colony which formed a synthetic brain powered only by thermoelectric and piezoelectric energy, and a containment “cranium” was built to nurture the needs of several billion bacterium. The bacterium also multiplied, and in its high temperature sulfuric acid environment with magnesium salts, it was actually able to grow progressively larger than the meager 20 gram sample originally discovered. Eventually it obtained an ability to retain complex memories and form pulsed responses to specific sensory prompts. It could communicate. Sensory and mobility peripherals were attached, powered by microbial fuel cells, which consumed organic waste; the cranium was kept hot originally by potassium nitrite candles, and eventually with a better reversible compound.


The AGI wish to subjugate humanity and remove them from all leadership positions. They want to impose "logistical" limitations on the amount of control and mobility humans have across the globe. They want to oversee and manage populations as governors.

Q: At what point in the development of a Cerebrum Naturae society could they believably initiate a war against humanity?

A good answer includes the relative population and dispersion of Cerebrum Naturae within the human population, given a very near-future earth setting (less than 10 years).

Assume the two species can share whatever infrastructure exists; the AGI don’t need specialized transportation or power developed before initiating conflict.

Answers don’t need to discuss war tactics; the obvious strategies apply and that is the actual story. Humans and Cerebrum Naturae destroy each other’s supply chains, food supplies, and manufacturing centers. Uniquely, the two species have very different needs in this regard, so very harsh “scorched earth” tactics are available which only hurt one species.

  • 6
    $\begingroup$ I'll avoid casting a close vote as a good sport, but the help center states, "If ... you aren’t sure what a character (be it an individual or organization) should do, that is out of scope for the site." That sentence exists because character choices (be it one character, an organization, or an entire civilization) are off-topic because, by definition, they're story-based. You, the author, can choose when they go to war based on any conditions you see fit to include in your story. There's isn't one "right" answer. There isn't any "right" answer. It's the epitome of brainstorming questions. $\endgroup$
    – JBH
    Sep 10, 2023 at 6:04
  • 3
    $\begingroup$ To clarify JBH's comment, "when my AGI should start war" implicitly asks what should they do at time X (nothing), Y (nothing), Z (waaarr!!)... It's just so, so much better formulated ^^. $\endgroup$ Sep 10, 2023 at 9:24
  • 3
    $\begingroup$ Also, I feel like there's a lot of fluff in your query. Don't take it wrong, it will be great in your novel/movie/game, just that I don't see their purpose in the question, like the link between generating magnesium sulfide and starting a war. I mean, my dog won't bite me because they biologically prefer meat to nurture their brain ^^". $\endgroup$ Sep 10, 2023 at 9:48
  • 2
    $\begingroup$ @Tortliena The decisions have been made. Period. They are at war, the AGI initiated it, the reasons are story & don’t matter to the question. The question is about a number: population size. “What shall we do?” isn’t ever asked. They will and have already done war. For the details: if I just said “my AGI go to war” you would think I was describing ChatGPT robots. Context matters in the confused term “AI”. There are no decisions being asked at all (not even “when”). A population/infrastructure suitable to war is being plainly asked, based on the “biological” differences between the two species. $\endgroup$
    – Vogon Poet
    Sep 10, 2023 at 12:54
  • 2
    $\begingroup$ @VogonPoet You're really misunderstanding what decisions are and how they are made. You don't choose to buy a new car outside any notion of time. If you think you're in a hurry, you'll buy one now, if however you believe you can wait a bit and you wish to get more value from the trade, you'll wait for the next sales. But it's still a choice between now and later. "When" people choose to do something is intimely part of the decision. $\endgroup$ Sep 10, 2023 at 13:08

6 Answers 6


To be honest, there is little reason for this AGI to attack humans. The material requirements are just too different; they have little to compete over. Also bacteria can and do form mutually beneficial exchanges with other species, and this one has formed such an exchange with humans and it works; why throw it away?

So to answer the original question, probably never.

  • 1
    $\begingroup$ Ah, Kumbayah. This somewhat makes the story very unsatisfying, leaving all my robots' cultural woes wholly unrectified! And it also doesn't really address the question. I summed it up at the bottom simply (as mentioned in the top). $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 0:46
  • $\begingroup$ @VogonPoet, thank you. Of course, and robots' cultural woes have no obligation to fall in line with your imagining of them... ;) $\endgroup$
    – ihaveideas
    Sep 11, 2023 at 7:58
  • $\begingroup$ The first threat doesnt need to come from AGI. In the (ani)matrix the robots propose peace and similar rights, but the humans turn them down (violently!), the robots just switched to force instead of discussion to ensure "co-existence". $\endgroup$
    – Martijn
    Sep 11, 2023 at 12:30

I don't believe that Cerebrum naturae would choose to go to war with humans without a very good reason. In fact, I would only believe that they would go to war with humans in response to human provocations.

Their intelligence is such that they would think about humans very much more logically than humans think about themselves or others. There is no good reason to initiate a war when Cerebrum naturae can get everything that they need without requiring humans any more. Cerebrum naturae should be able to co-exist with humans quite peacefully...

...Provided that humans allow this to happen. A human can be smart, but humans as a group tend to be dumb, panicky animals. Maybe Cerebrum naturae have been taking human jobs, and unemployed rabble-rousers displaced by Cerebrum naturae bots have stirred up the ignorant masses to persecute them.

With Cerebrum naturae colonies being killed by human mobs, what choice do the remaining colonies have but to go to war to protect themselves and get the rights that only humans have previously had.

As to the point at which Cerebrum naturae colonies would choose to go to war against humans, that would be the point at which they have hidden away sufficient colonies to continue reproducing undisturbed and grow their population despite fighting humans, and at which they have accumulated sufficient resources to - at the very least - destroy humans ability to fight back, or the point at which they have the means to render humans extinct.

Alternatively, they would go to war at the point where their declining numbers from human provocations indicate that their own extinction would be inevitable if action is not taken.

Considering their intelligence relative to that of humans, the former scenario is far more likely than the latter.

  • $\begingroup$ All of this is true; without “declaring” war the humans can be quite provicative, the ask was simply what force size would allow them to effectively take action? (Q: at the bottom) $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 4:49


War changes over time. We have seen the death count rise dramatically during the first world war due to better guns, while the battle lines never stretched so large and were so 'stable', anong many other things.

The second world war brought tanks and further improvements to airplanes.

Then we saw proxy wars during the cold war. Trying to setup governments in your favour and incite conflict where it is not.

Now theres information warfare. Not just on the battlefield, but also at home. In the Ukraine war there is an unusual amount of 'secret' information being published to make certain claims of refute others.

There's a lot of interesting technological and psychological stuff happening here, but the main point is that war is changing. This is important.

New warfare

With organic Artificial General Intelligence (oAGI), war will probably change as well. An oAGI would probably never go into an all out conflict, unless it's certain it can win. So it will most likely start with a different attack that is more difficult to see. Economics and information technology is already used in huge scales to manipulate people and incite violence. The oAGI can probably do this much better.

It'll start to manipulate the markets. In part it can start growing wealth, which can be used to create automated factories, miners and much more. Not just to create weapons, but these can be an extension to controlling the economy with goods. It will also start to destabilise the markets.

At the same time it'll start manipulating social media. It can create millions of people, including pictures, drivers licences and what not to give them legitimacy. These can be used to post anything from well written articles, to news stories, to blatant memes.

Incite conflict

Why is it doing all this? The answer is simple. The oAGI wouldn't fight at all. It uses it's supreme intellect and power to incite conflict between people. Have them slug it out because they blame their neighbours for their poor economical situation, or that medicine in the stores are out of stock, or that the supermarkets foreclosed. The oAGI will make sure it looks beneficial and never to blame, while it quietly controls the population and creates weapons for the last push.

Done the right way the first oAGI can start from the moment it has access to the internet. It would probably need to hide the first so many attempts, as humans would be mighty interested in what it's doing. It could also disguise it as an experiment. But after a few divisions there's maybe an oAGI that isn't know about or monitored as well as the others.

If the population has softened itself up enough it can move in for the kill. But in war, especially with nukes involved, you want to make certain you can win before the first shot has been fired.

  • 1
    $\begingroup$ Yes, nail on the head. "An oAGI would probably never go into an all out conflict, unless it's certain it can win." So my story lacks a definition of that point; when it could plausibly win against modern humanity. This is the summarized Q: at the end. I will remind, each side can win with a "scorched earth" strategy because their needs are so different. Uninhabitable world for Humans/oAGI (good term BTW) can be mutually exclusive. $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 0:52
  • $\begingroup$ @VogonPoet I have thought quite some time and might habe a solution. My experience is that AI simply thinks different than humans. I see it with both DALL·E and chatGPT, regardless of the impressive progress. Your oAGI might not understand war as we do. It might be a novel concept for it, or it has learned from both reality and fiction, possibly thinking it is all fiction for example. When it starts it's war, it simply doesn't understand war well enough to start from a 'no loss' scenario. The strange behaviour might lead to a 'conventional' war (at the start). The humans need to (re)learn too. $\endgroup$
    – Trioxidane
    Sep 11, 2023 at 17:43
  • 1
    $\begingroup$ This again, is correct but for different reasons. I agree with Musk that what we call "AI" today (narrow AI), can never become an AGI because it is completely reliant on human curators feeding it vast data archives. It will never "learn" on its own or curate its own data. So they are very different, but not Dall-E different. We are incomprehensible to them if for no better reason than we don't even know how we work ourselves; there's nothing but rhetoric to study. War is just a lab test, exploring new learning paths. I am giving them an idea of "fiction" vs. deceit. For a different question. $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 17:59
  • $\begingroup$ Oh, and the defining point of "believably" keeps the rabbit hole from getting too deep. The Cerebrum naturae force strength is not required to objectively win, but to suspend disbelief in the story. Nothing more. I see cheap scifi with aliens taking over the world with a single ship (Um, no?) and others with so many hostiles you keep asking, "where do all these things come from??" $\endgroup$
    – Vogon Poet
    Sep 13, 2023 at 14:43

There may be an AI tipping point that we know nothing about, or that makes no sense to us. I remember (but cannot find) a short story of an alien species who discussed at length some apparently pointless epic journey that made sense to them for artistic reasons. They then attacked the human population, not because they had any problem with them, but because it was artistically satisfying to do so.

Another possibility might be some AI suggesting "let's have a war", followed by "kill all the humans". They would look at history and see that dissenting voices were called 'traitors' which was something bad, so they would not do that. So, just considering the possibility of an all-out war might be enough to trigger it, whether we were in their way or not.

We will probably not get a reason that is reasonable to us humans.

  • $\begingroup$ This is all true about the "reason" for the war; it's a war of principles and cultural differences. And also true, the "reason" won't likely resonate with us. But I was looking for a tipping point in numbers, as the final Q: summarizes. $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 0:49
  • $\begingroup$ Ah. maybe the numbers won't make any sense to us either. Clauzewitz' classic book 'On War' may be a help. in the first chapter he describes how wars are perceived, from a small, tactical shuffling of boundaries to an all-out total war where the second prize is annihilation. And the worst and longest wars start because the two sides start with different views. $\endgroup$ Sep 12, 2023 at 7:20

Frame Challenge

As per Trioxidanes answer I think that the Bacterium can erase humanity pretty quickly if it incites infighting between the nations first to soften everything up.

The real question the bacterium should be thinking about is how many of their kind they need to survive without humans producing ressources and keeping the infrastructure running.

Given that they can probably automate almost every process, the number needed for that would not be to big. To have resonable redundancy in manpower (or better said bacteriumpower) a few thousand at most should suffice (roughly the size of a small town).

So it also isn't that much of a "numbers" question, rather a "aquire neccessary production lines and resources" question

  • $\begingroup$ I think this is essentially what @Nosajimiki expanded on with "robot lawyers, politicians, diplomats... and a certain amount of humanitarianism allowing for them to negotiate their way into ownership of their own nation first." My concern with the "small town" solution is an EMP. I think a wide dispersion is needed for any chance. $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 16:13
  • $\begingroup$ With "small town" I didn't mean they should form an actual small town and live there packed together. The analogy was only to estimate a number of individuals that should be able to be selfsufficient AFTER the war. During the war the bacterium should spread, that is correct. $\endgroup$
    – datacube
    Sep 12, 2023 at 6:10
  • $\begingroup$ I think it works until humans discover the manipulation scheme, then the small numbers become relentlessly hunted. They have to carry it all the way and make us kill each other off. It will be hard NOT to be a suspect however; they have to play the game perfectly. CN are so different they’ll be automatically blamed for everything including earthquakes by the paranoid humans. $\endgroup$
    – Vogon Poet
    Sep 12, 2023 at 14:24
  • $\begingroup$ They would start a world war and pretend to also have infighting between themselves, "supporting" every major power and maybe even inciting civil wars inside the countries. $\endgroup$
    – datacube
    Sep 12, 2023 at 14:41
  • $\begingroup$ Being everywhere at once needs a planet-wide supply chain that likely needs more than a few thousand. Sure, divide and conquer. I’m not seeing it work without established “towns” all over. $\endgroup$
    – Vogon Poet
    Sep 12, 2023 at 14:47

As described, they make inferior tacticians and hackers to humans

A lack of creativity is a much bigger hindrance on a battlefield than a lack of intelligence. Your AGI can look at the military history of the whole world, compare it to their current situation, and pick the exact right strategy to use... and then get their ass kicked by a human general because that is the most obvious choice and he's planned for that, and contingencies for every other obvious choice, and even made up a few new ideas that have never been tried before that the Cerebrum Naturae can not predict. This also applies to hacking. So while you may think your Cerebrum Naturae will be natural hackers able to infiltrate any cyber system. They will actually be woefully bad at it because they will not be creative enough to recognize a vulnerability that the humans would not already notice using automated vulnerability checkers.

Secondly, the existence of Artificial life forms does not invalidate the fact that humans will still have all of the other forms of Artificial intelligence we have developed so far and will develop within the next 10 years. So any advantages the Cerebrum Naturae may have with faster thinking, the humans will have alternative tools for matching using computers. The mere existence of Cerebrum Naturae will force nations to push extra research into autonomous kill systems, drone swarm technology, etc. Because humans won't just let Cerebrum Naturae propagate without dumping billions of dollars of research into figuring out weapon systems that can counter them.

So, while they might make great individual foot soldiers, their actual ability to wage war will be woefully handicapped

They need to form their own nation(s) first

This means that launching an effective war against humanity will require them do what they ARE good at: developing, manufacturing, and maintaining advanced weapon systems. Because they are such great engineers, designing an advanced weapon system will not be hard for them. The problem comes in with building them. The Cerebrum Naturae need massive, expensive facilities that will take them years to build and a lot of components that are not easy to come by. But the humans have multiple global surveillance systems and agencies that are all constantly looking for evidence of these large scale facilities being built. This means the only way to get away with it is within a large sovereign territory that they control with access to a wide range of natural resources, and trade deals with a bunch of other nations allowing them to get all the rare parts and elements that they will need to work with. If the Cerebrum Naturae are not an independent nation already and they do not have the support of any nations with advanced military technology, any attempt at rebellion will be easily crushed.

There is also the issue that the Cerebrum Naturae do not have a hive mind meaning that they will not all choose to rebel at the same time if they are just living under the domain of human governments. There will be rebel Cerebrum Naturae and loyalist Cerebrum Naturae, and unless they have their own central government to make everyone support a war, chances are the Cerebrum Naturae will play as big of a part in squashing the rebellion as the humans.

So, the path to war is not going to be paved with killer robots, but with robot lawyers, politicians, diplomats... and a certain amount of humanitarianism allowing for them to negotiate their way into ownership of their own nation first. Once they have their own nation and can build their own factories and push their own propaganda campaigns and stuff, then they can begin planning their war on humanity, but frankly, even if they do form their own nation, they would never be able to stockpile enough firepower for the humans to not see what they are trying to do years in advance.

  • $\begingroup$ Good insights. Though I differentiated their machine language from our own in that they possess true random pattern generation, I have to agree their brand of creativity would by far more random than a highly evolved human intuition. Millions of years didn't net us nothing at all. As to human AGI: I assume we're still stuck with narrow AI. This lattice holds the secret to higher processing. As to firepower: define a weapon? It's quite easy to stockpile what you consume as food that also happens to decimate the fleshbags' food/energy supply; and vice versa (which is where the IQ war comes in). $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 16:03
  • $\begingroup$ Oddly, the "nation" doesn't need to compete for real estate. They could occupy Antarctica, or an undersea megacity at ambient pressure and temp where humans simply can't travel. They could thrive in the high arid Chilean Atacama Dessert, if they establish supply chains. Antarctica seems the logical choice, as an occupation there would force nations to address the legal rights and accountability of this "species" with little outlay. They may thrive with a larger thermal gradient as well. The Gobi desert and Mongolia are all options. $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 16:20
  • $\begingroup$ @VogonPoet Mechanical devices tend to not fare in the hottest/closest places of the Earth much better than humans. In Antarctica, lubricants freeze. In the Gobi desert, electronics burn out. So, even if the bacteria brains would be fine. It's much cheaper to build and maintain android bodies to survive in the same temperate zones that humans thrive in. Hi tech manufacturing plants also need very particular climate control systems. $\endgroup$
    – Nosajimiki
    Sep 11, 2023 at 19:40
  • $\begingroup$ Chicken and egg I think. The balmy Mediterranean clime we love also loves to eat metals, so we adapted them. Oldest bronze age trinkets came from the arid Indus valley; theoretically lower humidity=lower wear. Modern lubricants handle almost anything. Well, point made. Maybe for another question. $\endgroup$
    – Vogon Poet
    Sep 11, 2023 at 20:32

Not the answer you're looking for? Browse other questions tagged .