@waltercool as i read recently somewhere the ultimate turing test will be the ability to express politically incorrect opinions
@vbhide That's a good idea because of the complexity involved, but might violate some law of robotics.
If some AI with enough power have political opinions by itself, nothing would stop it from taking "the best decision" for everyone.
If some AI with enough power have political opinions by itself, nothing would stop it from taking "the best decision" for everyone.
It was likely trained to provide answers against Trump specifically, by the users or authors. Artificial Intelligence just learns from usage.
That's also why some AI can do racist or anti-Jewish comments too.
https://www.theguardian.com/inequality/2017/aug/08/rise-of-the-racist-robots-how-ai-is-learning-all-our-worst-impulses