The replay is not conscious. To get the replay to work, the Turing Test administrator has to do the exact same test. If the question is delayed a second, the "brain" will answer a question that was not asked. If the speaker to the "brain" burned out, it will answer without hearing the question. If the questions are not identical, the wrong answers will be given. The really-conscious brain would correctly react to these situations. The fake-conscious brain will fail to react.
I would say that the second Turing Test is invalid, because the tester must cooperate to make the test work. The whole point of the Turing Test is that the testor cannot tell that the subject is not human. In the replay example, the Testor not only can tell, but must carefully craft the questions to make the subject appear to answer. Any failure on the part of the Testor will make the fake-brain fail. But the same failure on the part of the Testor to a real brain will not make the real brain fail.
Your example proves nothing. A tape-recording of the answers would also pass your test and would appear to give the proper answers. No tape recorder can really pass a true Turing test, because the testers would quickly realize it is a tape and fail the device. Your example requires the Testers to cheat and pretend not to notice that the answering replay is not really conscious. You are invalidating your own test of consciousness.
I think the first run is conscious because it can adapt to any changes in the test. No matter which test is given, the real brain will pass. I think the second replay run is not conscious because it cannot adapt. It will not respond to any changes in the test. It also is dependent on the replay machine to provide the answers. The brain by itself is not replaying the answers. The requirement for the Testor to do the test in an exact way invalidates the test. A real Turing test does not have this requirement, hence I would claim that the second run cannot pass a real Turing test while the first run can.
If I am not right, why have tape-recorders not already passed the Turing
test and been declared true AI devices?
Harvey Newstrom <mailto://firstname.lastname@example.org> <http://newstaffinc.com> Author, Consultant, Engineer, Legal Hacker, Researcher, Scientist.