6L6 wrote:Silverwall wrote:
"Interesting reply but the complexity of tapping into existing memories and the gaps in "Nharman" that resulted stongly support the theory that infancy would be the best time medically for such a procedure.
While we have a lot to learn about the brain a lot of what we are learning though things like concussion and related issues is how unbelievably complex the brain is.
Personally I think that we will find that such incredibly complex brain-interfaces will be theoretically possible in the same way star trek style teleporters are. Mathematically possible but would require more processing power than the entire universe holds to actually make it function safely.
Neural education devices are a staple of sci-fi and probably one of the elements we are getting further from as we slowly learn more about how the brain actually functions and we start discarding our primitive Cartesian thinking about consciousness and memory."
Good point Silverwall, but look at the progress made in Genome Sequencing machines in the last ten years.
The problem is that unlike genetics brains are not all wired the same, we see some of this when we look at how hearing works. When you take a recording from one persons eardrum and play it to a second person it is gibberish because each persons brain decodes things differently and has learned the unique ears that belong to that indivudual. This makes it orders of magnitude harder to train someone in the way suggested by the safehold books or the nanovirus assassins in Honorverse.
So much of memeory is learned rather than evolutionary in the way genetics are. Where in the cortex your muscle memeory of an action is will be different from mine and different from RFCs even if we all perform the same action to the same skill level.
The other problem compared to genetics is that we still lack a good theoretical model of how memories work and are formed. We know how DNA works but memory is much less well understood. Despite the common sci-fi concept of engrams or identifiable memories all the most recent research suggest that memory and learning is not formed in discrete identifiable packages we could target with technology.
This quote via wikipedia below sums up our current level of knowledge quite well and it is well short of being operationalised. Given what we suspect in the quote below a neural interface device would need to be able to access every neuron in the brain to be able to lay down memories without damage as unlike DNA there is no "machinary" that writes memory like there is for building DNA. This is why I consider it in the theoretically possible but practically impossible category. If we come up with a better model I may change my mind.
"One question that is crucial in cognitive neuroscience is how information and mental experiences are coded and represented in the brain. Scientists have gained much knowledge about the neuronal codes from the studies of plasticity, but most of such research has been focused on simple learning in simple neuronal circuits; it is considerably less clear about the neuronal changes involved in more complex examples of memory, particularly declarative memory that requires the storage of facts and events (Byrne 2007). Convergence-divergence zones might be the neural networks where memories are stored and retrieved. Considering that there are several kinds of memory, depending on types of represented knowledge, underlying mechanisms, processes functions and modes of acquisition, it is likely that different brain areas support different memory systems and that they are in mutual relationships in neuronal networks: "components of memory representation are distributed widely across different parts of the brain as mediated by multiple neocortical circuits."