In my scenario a quite hierarchical society has two particular groups of students:

  • An extremely intelligent well educated elite who are meant to lead the country some day
  • A not less intelligent group of lower class people or people considered a threat to the society.

The second group is used as a kind of intellectual slaves, multiplying the intelligence and education of the first group by studying for them.

Knowledge from every person in the second group is then transferred to everyone in the first group, so the elite students have the knowledge of like twenty people combined (the transfer process is quite harmful to the person the knowledge is taken from, which is why there needs to be a second group to be exploited).

Now I'm not totally inexperienced in how the brain works, but I can't figure out what a more or less believable mechanism for that transfer could be and why it's that bad for the person giving the knowledge. Does anyone have an idea?

  • $\begingroup$ Quite horrible idea ! the transfer would be harmful, like a life transplant ? In that case tip: search here worldbuilding.stackexchange.com/search?q=brain+transplant $\endgroup$
    – Goodies
    Sep 10, 2021 at 16:28
  • $\begingroup$ I was rather thinking of it being harmful over time. Like if you do it once it will be fine, but if you do it on a regular basis it will impair cognitive function and eventually leave you quite a mess. $\endgroup$
    – Ninke
    Sep 10, 2021 at 16:41
  • $\begingroup$ Like having to do repeated brain surgery to put electrodes in ? I'll read the answers, but to be honest I'm reluctant to go search for answers myself.. I don't want to have bad dreams about it! $\endgroup$
    – Goodies
    Sep 10, 2021 at 16:46
  • 1
    $\begingroup$ I think this is the amazing futuristic technology called a book :-) $\endgroup$
    – jamesqf
    Sep 10, 2021 at 18:43
  • $\begingroup$ unless you drastically enlarge the skull of your "haves" handwavium is your only option $\endgroup$
    – John
    Sep 10, 2021 at 20:13

8 Answers 8


I'd suggest reading How to Create a Mind by Ray Kurzweil. I also suggest re-reading [Gödel, Escher, Bach] by Douglas Hofstadter, specifically the part about network congruence (USA vs ASU).

Many stories simply propose your premise as a trope, giving the rules and ramifications without any detailed explanation of the technology.

In order to explain, I would suggest that the "elite" have nanotechnology-enhanced brains that can be "programmed" to make new connections. But the students with their "natural learning" must have their brains scanned to trace the connections which is not completely nondestructive. The nanites that feel their way along the axions to see what's connected to what will damage secondary cells, and characterizing the details of the synapse mostly puts it back the way it found it but is imperfect, and there is the issue of not reading everything simultaneously so a temporary disruption at one point can cause changes to start to spread from that point.


"Knowledge from every person in the second group is then transferred to everyone in the first"

A slight frame challenge in as much as only one person gets the skills from one of the 'underclass'.

So you want the transfer to be 'harmful' to the person it's taken from?

How about just going for a straight surgical transfer of the long term memory centres of the brain?

You'd excise the area of the brain associated with any long term memories or learned skills you want & graft them onto the recipients brain along with a course of stem cell injections along the join to help promote new connections to the transplanted tissue, obviously you'll also need to provide blood supply to the new brain sections & bone inserts to extend the skull.

You obviously add to the brain with each procedure or they'll lose as much as they gain from it.

Which means your elites will be easily identifiable by their now enlarged cerebral size of course.

![![enter image description here

Particularly after more than one or two such procedures.

So naturally big heads will become an ostentatious status symbol (the bigger the better) & fashions in head enlargement prosthetics will flourish .. right up until the inevitable revolution when it's all going to get a bit awkward having a big head of course ;)

Memory transfer wouldn't be immediate of course, it's going to take time for the graft to settle in & new neural connections to it to form.

For best results & to reduce or eliminate tissue rejection complications the donors will also need to be genetically compatible with the recipients so you're likely genetically engineering them for that.

Or you might 'inoculate' the recipients against tissue rejection from particular tissue types.

This is how you'd achieve that 'inoculation' > In utero hematopoietic cell transplantation: induction of donor specific immune tolerance and postnatal transplants > so it has to be done in the womb.

Families might have different breeding lines of donors they 'owned' if you use either option.

The rich person from group one subsidises the education & raising of an individual from group two then gets the memories & skills they have leaving them with no memories from before the transfer & unable to form any long term memories after the transfer.

So both Anterograde & Retrograde amnesia because you removed the structures that store long term memory, you take the structures that store learned skills & languages then they're gone too.

To cut down on the risk of revolution I'd recommend three social classes, your elites, the donors & the normal bulk of the population who are just happy they're not the ones being exploited this way.

So basically a slightly skewed take on The Island, probably.

Is that sufficiently dystopian for you?

  • 1
    $\begingroup$ @Ninke I like the Neuromancer option as a variation on this answer. Your "underclass" are clones or close relatives of your upper class, and thus the brain components are compatible because the unders and overs are HLA identical. The law allows the uppers to make questionable moral choices about their "relatives," for whom they have legal guardianship of. Then they grow a new 3d-printed memory center and surgically implant it in the under's brain to replace the lost one. Plus this way, you have a backup organ donor for the upper who might need emergency tissue replacement. Incest is best! $\endgroup$
    – DWKraus
    Sep 10, 2021 at 20:12

Every person has brain implant with removable storage cartridges, something micro SD sized. This implant can read or write data to mind. Actually, it acts as small part of brain taking functions of memory processing. When person learn something new, its recorded on this memory cartridge, but trick is - human mind is not computer, its believed to have quantum origin, so:

  1. Anything recorded to memory storage can be only recorded by live brain, while consciousness is learning something new. Computer generated knowledge or artificial memory is too simplified. No nuances, no reasoning. Its something suitable to low salary workers, not for children of elite.
  2. Storage cartridge can be read only once.
  3. if you plug away cartridge from somebodies brain, its like brain surgery - some neural links are broken, and, in most lucky case person whom undergo this procedure will have amnesia ("What have happened in previous few years with me?"), or will become comatose.

I would like to point out an interesting phenomenon that might not (at first) seem entirely relevant.

I am reminded of an experiment made with an FPGA a few years back. For those of you unaware, an FPGA is a special kind of computer chip that doesn't do anything specific. Most chips have paths etched into their silicon that makes them do one thing and one thing only... even CPUs are like this, their "one thing only" being really complicated (to load and execute programs according to the instructions those are composed of).

FPGAs are fabricated in such a way that you can essentially "reprogram" them to have different paths. And thus, you can tell it to be a ram chip, or a bunch of nand gates, or even a CPU. But all sorts of other things besides.

And the experiment was designed such that the FPGA would try to do human voice recognition. It was an evolutionary scheme, where it'd be randomly reprogrammed, tested, and if that scheme recognized the input correctly, it'd be adjusted. Thousands of iterations, maybe tens of thousands.

At the end, they had a program for this particular model of FPGA that could recognize that spoken word correctly, with very few false positives or negatives. But when they loaded this program into an identical FPGA, it simply would not work. At all. The hardware was identical, the software was identical. But it failed.

The story of how they went about debugging it was interesting to people who are interested in that sort of thing, but the explanation turned out that though they were ostensibly identical models, some slight manufacturing defect occurred in the one they used. Something microscopic on the wafer itself. Not enough to make the part fail manufacturing tests, not enough for them to notice. But the approach to coming up with the recognition code somehow incorporated that defect into its functioning... perhaps some signal was slightly delayed in the experimental unit, but faster in the rest. Something of that nature.

So, I ask you, the reader, what the chances are that loading software from one person's brain into another will result in that software working at all, when we are all very much "different models"?

It's one of the reasons telepathy is such a ridiculous concept.

Now, that out of the way, I contend that the OP needs to consider that with the current state of the art, any such attempt will require opening up skulls and implanting electrodes. While it may soon be possible to read from one mind, writing to another is going to be physically invasive into the foreseeable future. If human brain matter were magnetic in some way, then it could be possible to manipulate it without opening a skull... and if the story requires something similar, I would probably base the technology off of some sort of medical imaging (MRI). With large doses of handwavium.

In other such stories (Vernor Vinge's Rainbow's End), users sometimes suffer debilitating neurological issues from using/over-using similar technology.

  • $\begingroup$ A lil bit of MRI never killed anyone - exept the person with the microchip in their brains...That's quite interesting about the experiment, I might just apply the hadwavium there though. $\endgroup$
    – Ninke
    Sep 10, 2021 at 18:51
  • 1
    $\begingroup$ That FPGA story is awesome, do you have a link to the original source? $\endgroup$ Sep 11, 2021 at 13:36
  • $\begingroup$ Reprogramming cells connections in FPA have its limits, how much changes are guaranteed to work as intended. Ten of thousand is a lot (EEPROM in microcontrolers have guaranteed like 1 thousend overwrites) so some connections could simply wear out and regardless programming will be still open/still closed, so the FPGA would works other way, than programmed - so IMHO scanning each one connection could yeld the "real" program it was interpretting - slightly different from original $\endgroup$
    – gilhad
    Sep 11, 2021 at 22:33
  • 1
    $\begingroup$ @hamsolo474-ReinstateMonica Unfortunately, I do not have that one bookmarked, and modern-day Google is such a shitty search engine that it fails to turn anything up. I either read about it on reddit or Hacker News 4 or 5 years ago, maybe both. $\endgroup$
    – John O
    Sep 13, 2021 at 13:21

Since the process isn't possible (and, incidentally, probably never will be), you get to invent your own process, with whatever features you need.

For example, let us say that the knowledge (e.g. words in a book) need to be, for want of a better term, "digested" to become a series of neural weights and connections, that (this is the unrealistic part) are reasonably similar in all humans. The meat of the matter is this: while it is known how to imprint these engrams in a living brain (blindly projecting an electrochemical signal for long enough we can be confident that the knowledge has "taken"), we do not know how to create them. The only available way is to have the knowledge digested by a human brain, and then read and analyze the engrams. By checking which engrams were also found in the brain of an uneducated yokel, we can know which engrams are not useful.

But, here's the rub, reading the engrams isn't as easy as writing them. The only known way is to electrochemically map the brain. And the "chemical" part of the mapping cannot be done from the outside; the brain - or at least the relevant parts of the frontal lobes, plus the cerebellum for muscle memory - has to be flash-frozen, laser-sliced, and each slice sampled at the cellular level, a sort of 3D printing in reverse, to identify all the connections and content of each neuron. Then all the data need to be crunched by a supercomputer to reconstruct the memory engrams.

The first step of the process, as you can see, involves killing the donor.

Or, to keep the butchery to a minimum, instead of the "laser mapping" we could imagine other long-distance, no-contact measurement systems, that all involve hibernating the donor so that the brain isn't "running" during the measurement (you need to "take a snapshot" of the brain); thawing, then, involves a measurable risk of some cellular damage. In this case, the donor is not automatically dead, but can suffer significant brain damage (how significant, it's up to you to decide).

  • $\begingroup$ "Since the process isn't possible", Oh I don't know, I can think of some 'wetware' methods we practically have the tech for right now // well, almost, we kind of now how to do it, but it will also require developing enough expertise & experience with the procedures used b4 you could carry it off on a regular basis with acceptably low levels of failure for people to be happy with it as a regular elective procedure :) $\endgroup$
    – Pelinore
    Sep 12, 2021 at 1:35
  • $\begingroup$ @Pelinore it isn't possible - and very probably never will be - because knowledge is stored in "holographic" format and the "language" finest features change from person to person (brain architecture isn't completely innate, it gets heavily "programmed" during the first one to five years of life). Reading data, okay (probably lethal but possible). Writing would require recalibrating a vast neural network from the outside, operating with mechanisms we have only theorized. The only ways conceivable imply workarounds (e.g. David Brin's "soul gem"). $\endgroup$
    – LSerni
    Sep 12, 2021 at 11:00
  • $\begingroup$ I was only referring to the possibility of transferring memory from one person to another not your answer entire & by 'wetware' suggesting partial brain transplants of memory centres to add them to your own :) $\endgroup$
    – Pelinore
    Sep 12, 2021 at 12:46
  • 1
    $\begingroup$ Reading & implanting data with some sort of technology? probably impossible yes // reading the 'activity' (thoughts) without damage is possible (after a fashion) with something like Neuralink but that's really not the same thing as the underlying data (the actual 'hard wired' memory) of course. $\endgroup$
    – Pelinore
    Sep 12, 2021 at 12:47

There was an episode of Stargate SG1 like this (Season 3 Episode 5), Nanites. Everyone has nanites injected into their brain at birth as well as anti-rejection drugs, these nanites directly record your memories and synaptic pathways and everything, as a bonus they actually accelerate your learning speed, give you better recall, fewer mental problems etc... however the brain becomes overly dependant on the assistance provided by these nanites and when they are harvested they are physically removed from the person, In the show it was equivalent to a lobotomy with severe memory loss.

You can add new nanites to a person but they won't be able to fufill the role of the previous nanites, its like removing a tree from the earth where its roots have softened the dirt and found all the best directions to grow in, then transplanting that tree into a different area, the root network isn't laid out in the best way for this new terrain, it can eventually grow to adjust to it but it realistically has to begin testing the soil for nutrients, drainage ability, physical resistance, etc.. and make new pathways for this new area.

Adding someone else's nanites to someone of the upperclass would be like giving a person access to a new set of knowledge that isn't linked into any of their critical functions, they too will grow to integrate this new knowledge into their mind but they aren't at a disadvantage while doing it, they aren't functionally lobotomised because they maintain all the nanites that handle their core brain functionality.


Like today, skipping a few steps along the way

enter image description here

In the Ghost In The Shell universe, the process to enable this is called "cyberization"

As you are reading this, I am transferring thoughts into your mind. No, really, I am. What you are reading is my thoughts, in text form. And now that you have read it, that text has formed a thought in your mind, which is — ironically perhaps — the thought that I have transferred my thought into your mind.

"But that is just plain old human communication, right?"

It is. But what you have not realised (or maybe you have, I cannot be sure) is that this is a very primitive form of thought- and knowledge-transfer.

Let us go through the process...

  1. I have a thought, a feeling, a piece of knowledge in my mind
  2. I translate that thought into words and sentences
  3. I order my muscles to move such that they write words and sentences, that are transcribed as electronic data
  4. The data is sent over the internet
  5. The data is formed into characters on your screen
  6. Your eyes watch the screen and and send the image to your brain
  7. You brain interprets the image, recognising the characters as words and sentences
  8. Your mind translates the words and sentes into a thought in your mind

It is primitive, it is clunky, it involves a lot of really tedious and mundane steps. But it is knowledge-transfer, from one brain to another. I have transmitted a thought, a piece of knowledge, from my mind,to your mind.

In a more bare-bones description we can express it like this:

  1. I have a thought
  2. I modulate the thought into data
  3. The data is transferred to you
  4. You demodulate the data into a thought
  5. You have the thought

So, my ability to express myself, to write, my keyboard, the internet, Stack Exchange, your eyes, your ability to read, your reading-comprehension, are nothing but a very clunky modem for thoughts.

"Okay, but in the future we can do this more fancy, right?"

We probably can, yes, by cutting out steps in the middle, making the modem more streamlined and efficient.

For example...

  • Do we really need step 3? What if I can interface with a virtual "keyboard" in my mind, that "types" out the post as I think it?

  • Do we really need steps 5, 6? Why not have a virtual "reader" in your mind, that takes in data and instantly translates the characters to your mind

  • Do we really need steps 7? How about if what we send over the internet arrives as words and sentences in your mind?

"That is all fine, but how to do that?"

No matter how we want to do it, there is one vital piece of technology that is required, and that is a wetware-to-hardware interface, i.e. something that provides a bridge between the cells of the human brain and nerves, and computer hardware. When we have that, then these middle-steps will start to fall away, one by one.

For example, in the Ghost In The Shell universe, the process of adding this interface is called "cyberization". Without going into technical details, it simply means that — when that is done — your brain interfaces with — and meshes with — computer hardware.

So, how could transferring knowledge from one person's brain to another person's brain work in a futuristic society?

Exactly like it does today, only cutting out steps in the middle, when we have an interface between human cells and computer hardware.

  • $\begingroup$ But how do you get the limitations required by the trope? What you describe would allow for non-destructive one-to-many communication. $\endgroup$
    – JDługosz
    Sep 14, 2021 at 15:04

There is one BIG problem - back loop - many of the unprivileged would hate those privileged and many of them would hide it, to avoid repression. But as you said they are equivalently inteligent, so many would be able predict their fate and place (knowingly or not) some "bomb/virus/hidden agenda" in their brains (from "must help those poor" to "must help overthrone those bestial elites" to many more hidden schemas).

And the brain is not just sorted field of ordered chips with clean and specific content, but it is entagled net, where everything somehow interacts with everything other. So you cannot just took usefull content and left the rest there, you have to take more or less everything and use it as whole.

So when first elite took this enhancement, there was a lot of unwanted agenda, many which appeared only a way later ("I have to help the poors when I am in high enought position and hide my agenda before it, behaving arrogant toward poors to be trusted as elite"). It was discovered in a lot of first wave models, but this enhancements was still only for few and those was watched, so many cases were discovered.

Which led to preventive methods, where the added parts was used not as part of elite's brain, but as slave-sub-brains (more like books or other sources, which are not trusted) and to keep them in their place/positions was necesarry to first "edit and conditioned" the source. Which was cruel and brutal and put the subject to menthal and health/life risc - and so had to be done on the original keeper of the brain (not risking elite's brain) and then the subject (if survived in useable state) had to be tested and watched and maybe "edited" again. This was longer process and not all subjects passed it. But there is not so many elite and is a lot of testing subjects, so even whene there was losses, it was sufficient, if - say - 1 of 10 was "edited" enought and still survived. Then his brain could be used. (And brain-less body disposed - so there we have excluded reusability.)

And while this was done in secret, it was on such scale, that really inteligent people (= subjects) was able to find some traces here and there ("how it came, that nobody of previous classes was mentioned in anywhere in press, supposing that we are trained to serve at official places?") So those subject on the other hand was less willing cooperate fully and more prone to train their mind also in other, unwanted, ways. Which again led to need of more aggressive "editing and testing" - in positive loop of escalation. And while elites won, it was not easy and it could not be totally hidden from donating-subjects, so the war is going actively generation after generation, with a fatal endings for all subjects selected for potential use.

(And subjects could not be made less inteligent or less analytical, as it would render them useless by definition, so it had to be all applyed just after the subject manage all wanted learning and it means the subject must be mentally broken to submition afterwards. Also, as between elites are conflicts too, it is really udesirable (even if it would be possible, but it is not) to give potential rival access to part of your mind - so even elites do not try to "copy" those subjects minds, as it could be big hole in their personal security against each other)


You must log in to answer this question.

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