“Conscious AI” would have to be “conscious like us”: having been assessed to be fundamentally human in self-awareness, agency, thinking and understanding (much

“Conscious AI” would have to be “conscious like us”: having been assessed to be fundamentally human in self-awareness, agency, thinking and understanding (much more than octopi or great apes for example), shared cultural baggage, ways of learning, and so on. This is because between “conscious in a different way” and “conscious like us”, for all intents and purposes, the latter is the only one we can recognize and reason about. (As a close-to-home example, modern human is known to not care that e.g. animals or plant systems could be conscious in some different ways, it was not a concern so far so it’s not like we are going to start to care.) “Conscious in a different way” can matter, but some pretty mainstream philosophical assumptions would have to be turned upside down first. Personally, in part due to the above, I don’t think “conscious AI” is a theoretical possibility in foreseeable future (and while it remains a simple inhuman tool, incidentally and tangentially but importantly, the excuse of “learning” cannot waive copyright protections: if there is no human doing the learning in there, it is nothing more than scraping original works in order to create derivatives). However, if it were to happen, bad or not, “conscious AI” should spell the end of the industry which hinges on its exploitation (for all our faults, we tend to abhor blatant abuse of our fellow thinking human beings).