"An implanted memory prosthetic would have electrodes to record signals during learning, a microprocessor to do the computations, and electrodes that stimulate neurons to encode the information as a memory."

or in other words -- we record "signals", then compute something, and finally "stimulate" the neurons. I appreciate that this isn't a scientific report but this borders on parody.

The magnitude of technical, computational and biological problems to be solved here in order to achieve anything even remotely close to the headline would require 3-4 major scientific breakthroughs. With all respect, you don't publish them in IEEE proceedings.