(Edited question to meet concerns)

The Problem

Can you create synthetic biology that can be installed temporarily in human DNA to give people skills at birth, or are there problems, risks and limitations such that this violates some fundamental principle?


There are reasons for believing skills can be genetic and in the form of instinct in the lower parts of the brain. Between DNA fragility and energy involved, a synthetic DNA strand may result in cancers or cell suicide due to failures at the genetic level. As no animal hard-codes the cerebral cortex or neocortex, there's no evidence this can be done genetically. Nor has anyone made synthetic DNA of this complexity. Not does anyone know if any arbitrary skill can be encoded this way (as all cells share DNA), if the extra chromosomes can be shed to avoid making these connections rigid or the impact a mutation might have. But we can reason if these objections are logically sound or logically unsound.


Some brains seem to be largely hard-coded and the autonomous nervous system must presumably also be. There is plenty of knowledge that is not subjective and learning it subtracts from the time spent learning things that must be experienced to learn.

This tends to be the hardest knowledge to learn, and the most stressful, causing many to develop an antipathy towards learning and the learned. This is not healthy. Skepticism is fine, but an active hatred harms everyone.

So let's say a race of transhumans, or beings on another world, reserved an extra chromosome for uploading a dictionary, of sorts, of such knowledge. This pre-wires the brain with this information. The chromosome is then ditched, because it would impede learning and wouldn't be valid by the next generation anyway.

The transhumans would then need to use some sort of brain simulator to build the next iteration based on new knowledge. You'd simulate what connections the brain would form with just this information, then reverse-engineer a chromosome that will generate these specific connections. This is an expensive operation.

The obvious risk of this method is that the brain relies heavily on the die-back of neurons and synapses in early childhood. There are all kinds of side-effects if this doesn't happen, but neurons and synapses in use don't die back, and this method will create a whole bunch of new ones on top of that.

There may be other unintended consequences. Abnormalities in the forming of connections would have a greater impact, for example. It might impact the forming of other abilities, as the brain would not be learning from a blank template but one that is partially constructed. Also, DNA seems to hold a 'grudge' over unexpected extra chromosomes that clash with existing ones.

On the other hand, it would free the mind to concentrate on things minds are good at and would increase the time the brain is at its most productive -- all reasons that transhumans and highly advanced societies might find the notion appealing. It would allow them to leapfrog limitations to progress caused by limited opportunity running into the latencies of learning.

There are other ways to achieve the same effect, but transient GMO learning that utilizes mechanisms that already exist could be argued as one of the simplest and requires the least technology to achieve.

So.. it's either extremely good or totally disastrous. Or maybe even both.

(From the perspective of SF/F writing, I like hard SF where you avoid breaking laws of physics. The Clarke Doctrine that you can break one with justification is OK, but I'd rather this not be it. Besides, it makes the idea hard to discuss.)

Clarifications to the question:

This system must:

  • Drop the encoding chromosome (a requirement that also makes free thought possible)
  • Program the higher brain functions
  • Use a whole brain simulator to turn a skill set into a set of genes

(It may be a one-way function such that you cannot derive any possible set of initial values. Conway's Game of life is like that, so are strong cryptographic hash functions. Neural connections may or may not be.)

  • Target the higher brain functions
  • Not result in any significant percentage of brains getting cancer, overheating, suffering massive cell suicide or any other terminal problem
  • Not be hard-wired, with no capacity for change, because what you have is an instinct and not a skill

In essence... Is this even possible? Are there any other limitations I haven't thought of?

  • 2
    $\begingroup$ @Imipak You seem to be describing a race of transhumans that are the same in every way as normal humans, but are born with an intuitive knowledge of many different subjects which does not prevent them from learning. Given that this only confers advantages seemingly without any drawbacks, I don't see how the answer can be anything other than 'good'. $\endgroup$ – walrus Dec 6 '17 at 17:32
  • $\begingroup$ @walrus: In a nutshell, you are correct. In theory. In practice, brains that over-develop get autism (which is why we with autism tend to be very good at whatever our specialty is). I have no proof that would happen here, though. There are human conditions in which people are born with extra chromosomes, but few survive for long. Can a brain simulator ever get the connections right? On the other side, if you have pure intuitive knowledge, are you less likely to question things that are wrong? So there are areas of uncertainty. $\endgroup$ – Imipak Dec 6 '17 at 17:43
  • $\begingroup$ @Imipak Thanks holywood you are mixing Autism, a disabiling condition with Sanvant Syndrome, a very rare condition not always connected with autism $\endgroup$ – jean Dec 6 '17 at 18:44
  • 2
    $\begingroup$ I think I've addressed most of the concerns in my reworded and reordered question. If this is insufficient, them I'd like to take this opportunity to thank everyone who has contributed to the discussion because their thoughts have been extremely valuable. I'll wait on the verdict, but it might be this is too complicated to ask in one shot. $\endgroup$ – Imipak Dec 7 '17 at 22:08
  • $\begingroup$ I think this is good now and voted to reopen. BTW: Welcome to WorldBuilding. If you haven't done so yet please take the tour and visit the help center to learn more about the site. Have fun and good luck with your question! $\endgroup$ – Secespitus Dec 7 '17 at 22:33

Arguably what you describe is what the body does. The body is built up from fantastically complex instructions that turn a single cell into the dominant form of life. We even see these complex structures built up in the brain. Consider our vision system. It is quite well accepted that we see the world using opponent processes. Somehow we all manage to process color into the same black/white red/green and yellow/blue channels, despite the fact that those colors don't even line up with the rods and cones in our eyes. Consider the complexity of the hypothalamus, or the tremendously important sex drive, a system so complicated that we're still trying to figure it out! Our brain is full of hard wired bits and pieces that make your hardwired skills look like child's play!

What is different about what you describe is that you're looking to do it synthetically. You're looking to shortcut a few million years and do rapid updates. To do that, you're going to need a language.

aaaacgatctagataatcgcccgcagagttgctcgatctgcggagtgaatgtgccgtgcgaacatcgtcatagaagagac agtttgatcgaattaaaagtcgtgcgccggcccctccaacttcgaccATGGTTGGCGCACCGCGCTTCACCCAGAAACCG TCCATTCAGCAGACACCTACTgtaagtttatcaaaaaaaacctttctgaaaatgttgtctaaaaatgttcaattgtttga cacttactattttatacatatctcacctatctctgttccggctctttctctctatatctcttgagcccaatgccctgtgc atagccacatggctcatggcaaatgattgcgggatcccgtgtatgtatctgcgcgtctgcgtctcgttcttccgatcatc gcaagtcctagtgtattgtgggggcgatgccccgaaagacaactggcagatgggggctcccgcgagttgccgcgctattt ctcgcctcggcatctctcatgatgctccccttgctccgtcgtcgctttacggctttcacgggctcaccaatgtgccgaga
gcagcggtggagccccacaggaatccgtagatgatgaagcacaaaaaacctaggagcaggccacagagaaacgagtagaa actaaaatctgaccgcaaaacacttactgaacataagaatagagcccagggggatttcgaaagatgaacagagatgaaca gagcagaatgattaaaaaaataagaagagcgtaaatattgactgtttgtagtttttgaattgtctcaaagaagaacaatt agctgaatcagccacttctacctcaaggacatcacataaaatttttgaatagttcgaatttccgctccacttttactggg acctcgttcgtttgtagagcttttagaaattcaaagaaaataattagttgaaaccgggtttttaaaaatttcgtttcgag tgttttcttttctcacgattttttccattgtatgcgttctgaaatttgacttgaaaattttctgcttcttgtagatgctc aaaaaacttgaaattctactgaaactgctcatcttctctaaaaactgtaaatctttcatctagtcagtctgaggtgtttt tatttctctctattttctctggaattcaaatgttgttctaatacataataatagtaaacg

It's not the most pleasant thing to try to pronounce, and its full 38,300 base pairs wont win any awards for brevity. But what is it? As it turns out, it's the genetic code for UNC-22, or Twitchin. It's a protein found in a nematode, Caenorhabditis elegans, the first species to have its genome completely sequenced. It's involved in sensing how much strain a muscle is under in C. elegans.

Why do I bring this up? It's because you didn't know what it was for until I told you. And neither does the body of C. elegans, if it did not already have a context which helped it express UNC-22 in the correct cells. Every cell in our body contains its own context, a soup of genes that are upregulated and downregulated. Each one shows how to interpret the genes in day to day life.

This context is optimized for the kind of genomic expression that has been useful for millions of years. It's not ideal for transcribing the code for how to play Beethoven's 5th, or how to know Kung Fu. It's designed more around resiliency than anything else. It's pretty astonishingly difficult to disrupt these systems. When they do get disrupted, the results are spectacular, like cancer.

So if you want to encode your skills, you're probably going to have to come up with another language which describes these skills in a way that is more convenient for your transhumans. Think of it like boostrapping. However, there's a catch. Now you have to encode how to interpret this language. This interpretation is very important: failure leads to "cancer" in these skills, so getting it right is going to be essential. The solution: a more robust language to describe your language. And so on, and so forth. It's turtles all the way down.

So this is the real challenge. How do you make that initial kernel of an interpreter which is bulletproof, upon which you build up the rest of your genome's meaning? Nature has their answer. It took them millions of years. What can you piggyback on that millions of years of evolution to construct your exacting structures?

Incidentally, C. elegans has 302 neurons (in its hermaphroditic form). Every single one of them is plotted out from birth to adulthood in the genome. We've actually built the tree of how every one of those 302 neurons differentiates, along with all of the "dead-end" neurons which are killed off along the way. They rely, in part, on combanitorial expression of genes to uniquely identify each neuron. If your transhumans wanted to become more than human, perhaps they should start by learning from the worm.

  • $\begingroup$ Boostrapping: When a ghost increases in strength using nothing but determination and an 80’s training montage. Oh, and +1 $\endgroup$ – Joe Bloggs Dec 12 '17 at 7:19
  • $\begingroup$ Magnifique! A superb answer. This allows me to go to the next stage. The high level language can be compiled to DNA, just as Pascal is compiled to assembly, and then you need supplemental error correction as you noted (I'm thinking nanotech containing error correction codes). The compiler would be the simulator. Clarke's Doctrine says I can break one law of physics as long as it's justified. So as long as one of these is possible, my world can just happen to have the other. DNA already has repair, turbo codes exist, therefore I can justify the second even if it's impossible. $\endgroup$ – Imipak Dec 12 '17 at 15:47
  • $\begingroup$ A metalanguage is going to be tougher, maybe I'll apply the get out of physics free card there. Ok, I'm satisfied. I can build a world with customized humans within Clarke's Doctrine, that meets Tolkien's demands that disbelief should never need to be suspended, and that contains both the best and worst of human nature (obviously important in a story to have conflict). Your answer is extremely helpful in doing this. $\endgroup$ – Imipak Dec 12 '17 at 15:51
  • $\begingroup$ @Imipak If I were reading such a story, I would want to see examples of where the process breaks down. You can actually use mathematics to prove that the process cannot be perfect, and the particular imperfections would impart a great deal of flavor. (Arguably the flavor of "life" is the imperfections in how our information is interpreted from generation to generation) $\endgroup$ – Cort Ammon Dec 12 '17 at 16:29
  • $\begingroup$ @Cort Ammon - BRILLIANT idea! That I'm using. $\endgroup$ – Imipak Dec 14 '17 at 21:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.