Post AcPjhomLUtOk7pfTvs by dogzilla@metrobus.masto.host
 (DIR) More posts by dogzilla@metrobus.masto.host
 (DIR) Post #AcPjhhpzIx9Ub6Jn8a by mpesce@arvr.social
       2023-12-02T21:47:40Z
       
       0 likes, 0 repeats
       
       btw i'd gently suggest that any 'turing test' is going to have massive methodological issues that go to definitions of what human behavior is, what machine behavior is, what interaction is, what interiority is etc etcyou know, just the basic philosophical questions we've been pondering for a few thousand years.so when someone says 'oh this or that passed/failed the turing test' ask them to unroll all of their comfortable assumptions and biases in that testing
       
 (DIR) Post #AcPjhomLUtOk7pfTvs by dogzilla@metrobus.masto.host
       2023-12-02T21:54:30Z
       
       0 likes, 0 repeats
       
       @mpesce I thought the Turing test explicitly sidestepped all those concerns by making it a measure of whether a human believed it was a human on the other side
       
 (DIR) Post #AcPjht2ndew7M1Og8O by seanmcbeth@arvr.social
       2023-12-02T23:22:26Z
       
       0 likes, 0 repeats
       
       @dogzilla @mpesce that's the problem. When you consider the case of sociopaths, they don't see anyone as "people", other than just a word to refer to other entities that are not themselves.A person is more than just an answer-providing machine. The fundamental flaw of the Turing Test is that it reduces people down to just that.
       
 (DIR) Post #AcPqPyrRyKZtrUUt6W by dogzilla@metrobus.masto.host
       2023-12-03T00:37:42Z
       
       0 likes, 0 repeats
       
       @seanmcbeth @mpesce But think about what you’re saying : your comparison is to a human. Possibly one that we describe as sociopathic, but still human.The Turing test doesn’t measure morality, but rather capability
       
 (DIR) Post #AcQ9yWoe8Lms95NB3Y by seanmcbeth@arvr.social
       2023-12-03T04:16:51Z
       
       0 likes, 0 repeats
       
       @dogzilla @mpesce No, my point is that the Turing Test is not scientific, because it's not repeatable. The example of the sociopath was just an extreme.The results of the test fundamentally rely on the background of the tester. Indeed, the test probably says more about the tester than the testee.There are some people who have been convinced by the humaness of ELIZA-like AIs since the 1970s. It doesn't mean anything. The Turing Test is crap.