Strategies developed by the artificial intelligence (AI) community for dealing with ethical issues should be useful in dealing with neurotechnology (neurotech). Some ethical issues associated with both AI and neurotechnology are illustrated by an application: ensuring a representative group of patients when testing; including a diverse set of engineers when designing, mitigating bias in the AI algorithms; verifying that certain populations are not disproportionately impacted by side effects; and obtaining input and feedback from a diverse set of participants. The AI community has wrestled with these issues for some time and has developed insights and strategies. (The application described aims to mitigate epilepsy by both applying signals to the brain and measuring brain voltages. AI is used to analyze the voltages and to design the signals applied.)
Of course, some ethical issues inherent in neurotech do not completely correspond to AI issues. One example cited is long-term influence on a patient’s sense of identity after prolonged neuromodulation. Another example concerns “fairness,” which can be much more significant in neurotechnology situations than in AI because of increased intimacy. Fairness also includes issues of expense and access. The article lists 14 core neuroethics issues and suggests a multi-stakeholder effort to identify corresponding AI issues and to transfer AI knowledge and capabilities to neurotechnology.
The article includes a review of core AI ethical issues, neurotechnology, and how AI is being used in the neurosciences. The basic premise of the article appears promising but it is somewhat hard to read, perhaps because it describes so much.