Are the risks associated with invasiveness fixed (eg, there will always be at least a 1/15000 chance that a patient dies within the first year of receiving a neural implant), or can they be reduced with better technology and more understanding of the brain?