Given the Chinese room argument, and given the development in chatbots and machine learning, isn't Turing test superseded by some other way of evaluating AI's inteligence? Would an positive result of a Turing test provide any value, besides telling that a machine is good at conversations (but possibly nothing more)?
The Turing test is a good test for AI applications like Siri and Alexa (or, in general, intelligent personal assistants), but it doesn't test a lot of features (e.g. vision) that an artificial general intelligence requires. There are several arguments against the Turing test (and it is definitely not flawless), but is it a necessary (but insufficient) test for AGI. The Turing test is still quite relevant: e.g. if IPAs passed the test, people would use them more often and they would be a lot more useful.