Predictability Of A Neuropathway: New Technology?

A few lines concerning an idea I had today. Bear in mind that this, in retrospect, is a ridiculous idea, but in case it's not so ridiculous I'd like to share it. Also, I'd like to mention that I have practically no education in neurosciences or neurobiology, and so there could be plenty of falsehoods, and of that I have no doubt.


Consider neuropathways associated to a given function. Said function is triggered by a well-documented process involving ionic discharge or inhibition of discharge. What if these neuropathways, which are presumably static (do not change greatly in a lifespan), become mapped out? Could we trigger intentionally a function by stimulating the neuropathway?

If indeed such a thing is possible, then using these mapped pathways (which, presumably --yet again--, are individually unique to each individual), an organized 'package' of data could be associated to the neuropathway. For example, if the aforementionned discharge is responsible for the word "potatoe", then that pathway, which has been mapped and recorded in such a way as to enable an outside stimulation, can be connected to a specific set of, say, technological data. Then, if having the ability to stimulate or inhibit the 'potatoe' neuropathway, it would be equally possible to stimulate other pathways with other labels. Compounded together with a more complex set of functions, thought patterns could then be engineered both by the poor individual having his brain programmed (information from the individual to the data receiving element) or to said individual.

Now, having established that any thought can be 'translated' to physical data, then an exchange between external data (from the brain's perspective) and internal data can be obtained. Knowledge could then be transmitted from an external data set and translated to the brain through either an already existing neuropathway, maybe even extending it further by incouraging more connections through other synapses, then knowledge can be bred artificially.
It would be possible to mindsearch elements, via the external data set, and acquire them from simple mental queries (queries themselves being neuropathway stimulations). A Google-mind, of sorts.

Limits and Assumptions
Now I understand a lot of what I just wrote is horribly untrue or inaccurate. In fact, most of it is probably violating current neuroscience research. One of the many erroneous assumptions is that information can be given simply by enabling specific neuropathways in the receinving individual's web of pathways. That, with the very little knowledge I possess concerning neurobiology, is a stretch.