Our sense receptors send electrical signals to the brain and it is possible (although complex) to use these communication lines to send signals to the brain. As an example, significant progresses have been made in creating an artificial eye that would take over a no longer functioning one. A chip is receiving the light and connects to the eye nerve in the retina sending signals. However, the retina is not just a converter from light (photons) to electrical signal. It makes quite a bit of processing and it codes the result to send it over the nerve fibers. Scientists have worked out the coding scheme used in mice eyes and are pretty close to understand the coding scheme of primates, that seems to be pretty similar to ours.
By 2050 we can expect this work to be long completed and to have implantable chip that could restore vision. Notice that at that point a chip may be receiving signals from any other place, so that one could switch vision from his surrounding to the other side of the world. This would be the ultimate virtual reality device!
Retinal implants have been experimented in the past few years and have resulted in restoring (partially) the perception of light (Argus 11 has been approved by FDA as retinal implant for patients with retinitis pigmentosa, up to 4,000 per year; it costs over 100,000€). However, this requires the brain to learn “vision” all over again, since the retinal coding function is missing.
Conceptually similar chips have been designed to restore hearing (cochlear implants). They convert sound into electrical signals that stimulates the aural nerve fibers and the brain “learns” to decode these into perception of sound.
Skipping the nerve fibers to bring the signals directly to the brain is way more difficult because of several issues.
Although there are areas in the brain were information is processed (e.g. the occipital part of the cortex processes vision) these areas are not a one to one match with a specific function. They are also serving other functions and other areas contribute to that specific function. As an example, vision is processed in the occipital cortex area and also in the amygdala (and possibly in several other areas). It is the concurrent operation of all these areas (neuronal circuitry) that results in a certain perception. A brain processing visual signals only in the occipital cortex area “sees” the image and the person can describe it, but does not understand its implications (like seeing a truck coming your way and not realizing that you need to hurry back on the sidewalk otherwise you would be hit by it: “blindsight” is a pathological condition resulting from a stroke in the occipital areas that makes the person blind but still aware of danger. He would not see a truck approaching but will move away from it sensing danger).
Electrodes can be placed on the surface of the cortex or can be embedded in the brain. Very thin electrodes are available that can send electrical pulses to very tiny areas, involving just few neurones. Hundred of electrodes arrays are also available to stimulate cortical areas. The stimuli are controlled by a computer and can be shaped and timed to achieve the desired results. Additionally, optogenetics has been providing in the last ten years a new and very specific way to activate (or disable) neurones letting researchers peer into single neurones or single neuronal circuits. These approaches, however, have a very limited time window of effectiveness since a brain is continually changing itself by rewiring its connections. A neurone that was involved in a specific processing may no longer participate in that after a few hours, as new data are processed by the brain. Hence, the placement of electrodes -very difficult to start with- will be made useless by the dynamic rewiring of the brain. This is why researchers have been looking at electrodes that can fade away once their "time-limited" use is over.
Stimulation of cortical areas to “communicate” with the brain has been experimented and has proven that can lead to partial restoration of vision, as an example. However, a full communication does not seem within the technological possibility in the next decades.
However, although it seems impossible to have a detailed communication to the brain, experiments have shown that coarse communication is possible. A team of researchers showed in 2015 the effectiveness of information transmission between two rats brains. A trained mouse could “send” its experience on accessing food to another mouse via wireless communications (the two mice were in cages thousands miles away from one another). That experiment was soon replicated between two persons and again it was shown that it is possible to “influence” a brain via wireless communications to take decision in a certain way (for human the experiment was based on gaming).
A true brain to brain communications is far beyond foreseeable technology evolution and that might just be good. Just imagine if other people could go beyond “reading your lips” and just enter inside your brain!