Bruno Marchal writes:

I believe that the quantum theory does not allow cul-de-sac branches.
I also believe that the Godel-Lob theory of self-reference not only allow cul-de-sac branches, but it imposes them everywhere: from all alive states you can reach a dead end. The Universal Dovetailer Argument shows that the physics (which has no dead ends) should be given by the self-reference logics (with reachable dead end everywhere).
I have been stuck in that contradiction a very long time ...
... until I realized the absolute necessity of distinguishing the first and third person point of views. That necessity is implied itself by the incompleteness phenomena, but that is technical (ask me on the everything-list if interested). The intuitive point here is that you cannot have a first person point of view on your own death: 1-death is not an event, and should be kept out of the domain of verification of probabilistic statements. Another intuition: the finite histories are of measure null among the collection of all histories (the continuum).

Can we clarify what is meant by "dead end branch" here? I assume that we are only talking about branching from the first person perspective. It is obvious that dead end branches exist from the third person perspective, because we have all known people who have died while the rest of the universe apparently continues. But the person who has thus "died" does not include that death and subsequent branchings in *his* tree, any more than he includes those branches in which an asteroid hit the Earth millions of years ago and modern humans never evolved. His tree by definition includes only those branches in which he survives, and in the multiverse, those branches go on forever, even if sometimes it takes an apparent miracle to achieve this. In Bruno's words, "you cannot have a first person point of view on your own death".

Alas, when I put the above argument to someone hearing about QTI/QS for the first time, they usually will have none of it. I explain that the brave physicist who will be killed, or not, on the basis of some random quantum event will experience only those multiverse branches where he is not killed, and therefore from his perspective will always survive. The usual response is: Yeah, but what if he ends up in the branch where he gets killed? I then get bogged down in a long discussion about the philosophy of personal identity, which isn't nearly as interesting as physicists volunteering to get killed. Does anyone have a more elegant way of driving home the point?

--Stathis Papaioannou

Search the Web for a new car

Reply via email to