A type of species is not defined yet. It can be the insect like bee, slime like oak moss or AI.

  • There is a hive-mind species, but its hive mind is not queen/king/mastermind. Each individual in the colony connect to each other using a protocol code (some kind of password that only allow individual in the same colony connect to the network). The protocol code is bounded by DNA (or anything with same functionality) prevent merge, exchange individual between colonies .The consciousness of each individual make up a hive mind for the colony.
  • The connection range is limited between individual. Thus, a colony can be distributed into many clusters when groups of them stay apart. (minimum 1 individual each cluster)
  • The connections are continuous and almost instantly (no delay).
    The species use its hive-mind to gather and share information (and knowledge) (Ex: when an individual see something, all others in the same cluster know.) and coordinate action.
  • The hive-mind may not require for the survival of each individual, but greatly depend on number of individual in a cluster.
    (ex: a whole colony which is big enough can become a civilization of sentient species, but when a group from that colony is separate, that group is no more than a wolf pack)

Here is my question.
1. I would like to know this kind of hive-mind sound reasonable.
2. Is there any works of fiction (novel, movies, lore, ...) or real-life example of kind of species I describe above ?

Thank you

  • 3
    $\begingroup$ That's what a hive mind always is to me. The classic example of this type is probably Gaia in Asimov's Foundation series. $\endgroup$
    – Separatrix
    Jun 17, 2016 at 6:26
  • 1
    $\begingroup$ There are no species that share information on the level that you're describing - that would be telepathy. That being said, does it sound "possible"? I'm going to say that it's your story, and if that's the way you want to play it, you can. Is it reasonable? From a biological stand point, I don't think so. From a sci-fi stand point, a lot more extravagant concepts have gone over very well with readers/audiences. $\endgroup$
    – AndreiROM
    Jun 17, 2016 at 10:33
  • 4
    $\begingroup$ Ants do everything you've mentioned except for "be telepathic". It's worth noting, though, that distributed computing isn't reliant on instantaneous, continuous communication between all nodes and nodes knowing the full state of other nodes, but rather on passing as little information as is necessary between nodes to complete the task at hand. Colonial animals do this as well, such as with huge schools of fish where small details aren't communicated, but danger is almost instantly communicated across the entire school. $\endgroup$
    – ckersch
    Jun 17, 2016 at 15:05
  • $\begingroup$ I don't think I need to add another answer, but one of the terribly interesting bits that happens when you make living things out of algorithms is that you have to become very aware of the persistence of your data. What happens when an individual loses their protocol code... or worse, through clever infection, a virus steals a protocol code. When I look at computerized hive minds like this, its solutions to questions like that that make such an AI come alive in my mind. $\endgroup$
    – Cort Ammon
    Jun 20, 2016 at 1:36
  • $\begingroup$ you are describing humans and the internet... almost... $\endgroup$
    – Keltari
    Jul 1, 2016 at 20:50

4 Answers 4


After thinking about your post, I came to the conclusion that the particle swarm fits your needs best. Therefore I will describe a version of this approach.

The particle swarm algorithm is similar to the evolutionary programming paradigm. At first one needs to specify the maximum size of the population and an initial population size. The next step is to define a world. The world is a n-dimensional array that is at least as great as the maximum number of individuals in the population. Then each individual is assigned to a slot. The slots contain information to be processed by the individual. Adjacent slots are similar and build regions. The simplest case is that all slots contain the same information and build one big region. In the most complex case, all slots will contain different information, so that a region comprises one slot (this means there are as many regions as slots in the array). In addition, one can define continuous or quantised regions (the region will change suddenly and adjacent regions will not be more similar than more distant regions). Since each individual lives in a particular region and is adapting to that region, the individuals will become specialists. The better an individual can handle some region the higher is its fitness and the more attractive it is for others. Thus the individuals in a region are converging (become similar), while the individuals of different regions will diverge. In addition the individuals stay not constant, they mutate and mate with other individuals and produce offspring. Now there are two possibilities:

  1. If there are empty slots in the region, the offspring will inhabit this slots.
  2. If there are no empty slots, the weakest individuals (the individuals with the smallest fitness value) will die and their slots will be inhabited by the offspring.

As I said, it is somewhat different from your description. However, your first point is accomplished by defining a velocity that specifies how fast information is spreading between the individuals. The second point can be satisfied by defining a maximum distance at which information is still receivable (beyond that distance the information will be too noisy or will be faded). Your third point is a problem because the information is spreading with the defined velocity and not instantly. However, the individuals of a region do coordinate their actions by adaption.

Here is a link that offers a more detailed description of the swarm intelligence approach:


The previous mentioned approaches (Tierra, Conway's Life, Amoeba and Avida) are less adequate because they address the evolution of self-replicating individuals. And I think that is not what you thought of. However, if you are interested in that systems, here are the links:

  1. Conway's Life:

  2. Tierra: http://life.ou.edu/tierra/
  3. Avida: http://avida-ed.msu.edu, http://avida.devosoft.org/
  4. Amoeba: http://thirdworld.nl/digital-life-behavior-in-the-amoeba-world
  • 2
    $\begingroup$ Can you also provide a description? The problem with link-only answer is if the links get broken, the answer become useless for future users. $\endgroup$
    – Vincent
    Jun 18, 2016 at 22:04
  • $\begingroup$ I will recommend to not delete, but if it's not edited I'll downvote later. $\endgroup$
    – JDługosz
    Jun 19, 2016 at 4:33
  • $\begingroup$ No problem. I'm going to edit it later this day. $\endgroup$
    – BobbyPi
    Jun 19, 2016 at 9:45
  • $\begingroup$ @BobbyPi it will be useful if you put links too, as reference section or something - no need to delete them completely. Nice links, good answer. $\endgroup$
    – MolbOrg
    Jun 30, 2016 at 1:14
  • $\begingroup$ @MolbOrg Okay no problem. $\endgroup$
    – BobbyPi
    Jun 30, 2016 at 15:41

Vernor Vinge's A Fire Upon the Deep book series has a similar species known as Tines.

In the book they communicate using an ultrasonic communication method, and each soul is made up of 4-8 members, with the ability to add members or merge souls to give lifespans longer than the individual members. At low numbers they are non-sentient similar in intelligence to dogs, at higher numbers they are much more intelligent than humans.

One problem I think your scheme might encounter is bandwidth through the entire network. If every node in the network is going to receive all the data from every other node this would require a huge amount of bandwidth between the individual members, this could be mitigated with local processing i.e. only passing certain information between the individuals or you could come up with a communication method that allows for lots of information to be transmitted.

An interesting outcome of what you propose would be that geographically isolated populations would effectively form independent hives. Given enough time would they change their protocol code via evolution? Would they fight each other when reintegrated or would they just become part of the greater being or interact as separate entities?? depending on the technology level you could have a planetary or interplanetary being or a group of colonies somehow coexisting.

Interesting stuff.


The call for references might better fit Science Fiction & Fantasy than Worldbuilding, but since others are answering here...

Star Trek has the Borg. They almost meet this, although they are still intelligent even in small groups or as individuals. They do have near instantaneous communications with a limited (but high range). It's unclear how they keep non-authorized individuals out of the hive-mind.

James Schmitz wrote a short story called "The Searcher" with a creature called the goyal. The goyal could split off part of itself. If it split off a lot, the smaller part could take intelligent action. If it split off just a little, it was dumber. The goyal was composed of smaller entities that normally worked in conjunction.

Schmitz also wrote about vatches in The Witches of Karres. Vatches could peel off part of themselves and leave that part with tasks to perform. Small enough pieces were not intelligent in their own right. Schmitz didn't explain how that worked in detail, so we can't know if they were a hive-mind or not.

Alan Dean Foster's Bloodhype has the Vom, which can lose communication with parts of itself. The Vom also has the ability to delegate tasks to smaller parts of itself.

The television show Revolution had nanotech that achieved sentience as a group despite each individual member being little more intelligent than a single neuron.

Replicators in Stargate SG-1 consisted of smaller pieces (think Lego-sized). The small pieces could gather together to form robot-like assemblies. Pieces in separate assemblies could communicate with each other but pieces in separate solar systems couldn't.

Replicators in Stargate Atlantis consisted of nanotech-sized pieces that could form human-shaped entities.

Keith Laumer's Retief series includes Retief's Ransom (also published as "The All-Together Planet") which has the Lumbaga. The Lumbaga are made up of smaller pieces (like kidneys and fingers). When broken down or in small groups, the individual pieces are not intelligent while humanoid-sized collections have similar intelligence to humans. Merging two humanoid-sized collections results in a super-intelligent entity.

Hive minds are rather common in science-fiction. Not all of them meet all of the criteria. Some of them may meet the criteria but we wouldn't know, as the explanation of how they work is limited.


The insect hive mind is the basis of a Frank Herbert book The Green Brain.

The same author also touches on this in Hellstrom's Hive.

  • 1
    $\begingroup$ Maybe you could add what it says in the green brain and hellstroms hive? $\endgroup$ Jun 18, 2016 at 0:23

You must log in to answer this question.

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