The information paradox for black holes / Q&A

40:00 What was wrong with a semi-classical treatment, but one can do in a semi-classical treatment, one can compute, as you did, the radiation from the black hole for most of its life, it would seem that that would be a reliable computation until you got near the end of the black hole, but if that's so, then information has not been preserved. So are you saying now that the semi-classical All of these expressions for the radiation are wrong even when the black hole is very large and one would naively have thought that they were correct.

42:30 Can you say what's wrong? Say yes. I just want to comment on the question, just because I have a lot of information not lost. I would like to make a case, I will actually make a case for this in my talk and then again ask the question. The question was given was really, it seems to me that the argument is still at a natural level. The technical stone of this conjecture is how Green's functions behave, this non-sugar topology which got sorted out first. Now, if I did not focus on Green's functions, it would have worked out how the Green's functions behave. And while there are some exponential fall-offs, there are also tail terms which die only autonomically, basically b to the minus 3 when b or u to the minus 3 when u is the retirement time. Now they may well be very different. Situations may be dramatically different. In the case of anti-de Sitter, where there is a cosmological constant, there may be actual exponential overlap. Is that what you're saying? I'm asking just no questions.

45:00 Sorry, there's no other questions. Colin. I can ask you two yes or no questions. In this formula, you think about states which are connected to geometry. Can I think about the disconnected two geometry, two separate states? And if the answer is yes, then there's a second. The second question is, do you think the transition altitude between having the state of geometry and having the state of geometry is clear? Do you want to ask number two? I guess this is the answer to the first question, so that I can consider this correct geometry. And the answer is yes also to the fact that the transition altitude between 1 and 2 is over the 0. This is what you're trying to say.

50:00 You computed a kind of propagator or S-matrix or dollar matrix at infinity by a contribution of different topological sectors and in the trivial sector you said that the propagator would be unitary but in the non-trivial sector it sounded like everything would go to a single state which is very far from unitary evolution. So, are you somehow saying that if you add together those two propagators corresponding to the two different topologies, one with information loss and one without, that somehow the result is unitary? Is there a last question? Okay, so I have a last question. So, Stephen, are there topologies for which some information might be lost rather than other information would be preserved? Or is it necessary that if the topology is non-trivial that all of it is lost? Or are there some topologies which only some information is lost? So I'll bring the proceedings to a close and let's thank Steve again.