Brain to Computer Interfaces (BCI) have progressed quite a bit both in the “hardware” part increasing sensing resolution and capabilities and the “software” part making it possible to make sense out of the electrical signals detected.
Today it is possible to control a robotic arm with thoughts helping people with motor disabilities, there are a few BCI demonstrations for playing game on a computer and several experiments have been performed using animals. BCI are transitioning from the labs to actual use at home, signalling the transition towards maturity although a lot remains to be dome to make them really usable. There are two crucial points: detection of brain activity shall be done without needing an invasive implant (today this is required to get better signals and decrease the noise created by the thousands of concurrent activities going on in the brain) and to move to a wireless connection from the detecting sensors to the processing unit.
Research has also progressed in the other direction (CBI: Computer Brain Interface) sending signals to the brain to activate specific reaction. So far most experiments have been run on animals and have been focussing on the activation of motor neurones.
It is much more difficult to move into this direction, computer to brain, since it is both difficult to influence what is going on in the brain and do so in the very specific areas that are to be involved.. Technologies like deep brain stimulation, DBS, and transcranial magnetic stimulation, TMS, are being tested, with the former providing much more precise control but requiring an invasive procedure and the latter being less accurate.
Obviously, by linking together a BCI with a CBI a direct brain to brain communication should become feasible. The hurdles are big and they are not just of technical nature. The most basic issue today is that we still don’t understand what is going on in the brain and the more we understand the harder establishing a brain to brain communications seems to be.
Whilst intercepting motor signals (ie the intention to move a finger of the left hand) is now -almost- straightforwards and conversely sending a signal to move that finger is becoming feasible by stimulating the appropriate motor nerves, the manipulation of thoughts is still in dreamland. We cannot, as an example, transfer the thinking of moving the finger and have the brain activating the appropriate motor nerves to execute that action. We do not know where the stimulation should take place (probably in different parts of the brain and at very specific moments…). Also, notice that by stimulating a motor nerve we can indeed move the finger but the brain would be unaware that the finger is moving because of a decision it has taken!
The problem is our brain (as well as a fly brain) is complex. We can manage complication but to manage complexity you need to tackle the full brain, to interact at the same level of complexity and we are still far from that point.