A bidirectional neural interface is a device that is implanted in a biologic nervous system so that it is able to send and receive inputs and instructions.
In other words, instead of just having a futuristic prosthetic arm that is controlled by the human brain like a living limb, modern science is devising new technologies that are able to send back instructions into the subject's brain as well.
The potential of these inventions is immense, and ranges from smarter prosthetics that allow the user to get used to them in no time, to neurorepairing of damaged brain areas, neurotherapeutic systems that could prevent degenerative conditions such as Alzheimer's, and increased intelligence and memory.
Brain-computer interfaces (BCI) are a technology that has already been available for a couple of years. These devices are implanted in human brain tissue, and stimulate neurons for various medical purposes.
However, a bi-directional neural interface truly creates a deep connection between human and machine data and information that now only one of the two sides can understand will be shared and decrypted by both the biological and artificial interfaces.
Cognitive enhancement will likely result from the ability to integrate different sources of information into the human mind.
Obviously enough, the mere prospect of achieving such an ambitious technological goal pulled the whole scientific community into a heated debate about the ethical implications of implementing these devices.
SpaceX's CEO Elon Musk already described the need for humans to become "merges of biological intelligence and machine intelligence" in order to stay relevant and competitive in the age of Artificial Intelligence.
He already founded Neuralink, a company that aims at developing technologies that will integrate AI with the human brain.
However, many experts still consider these projects as just pie in the sky goals that cannot be achieved with our current instruments.
Both our knowledge of advanced AI and our understanding of the human brain are allegedly far too backwards for us to produce reliable results in machine-powered cognitive enhancement, and a true bi-directional neural interface may require at least another 10 to 15 years before it becomes a reality.