On Tuesday, October 15, 2019 at 12:26:55 PM UTC-5, Philip Thrift wrote: > > > > On Tuesday, October 15, 2019 at 10:26:15 AM UTC-5, Lawrence Crowell wrote: >> >> On Tuesday, October 15, 2019 at 6:02:15 AM UTC-5, Philip Thrift wrote: >>> >>> >>> >>> On Tuesday, October 15, 2019 at 5:48:58 AM UTC-5, Lawrence Crowell wrote: >>>> >>>> On Tuesday, October 15, 2019 at 2:24:10 AM UTC-5, Philip Thrift wrote: >>>>> >>>>> >>>>> >>>>> On Monday, October 14, 2019 at 6:52:24 PM UTC-5, Lawrence Crowell >>>>> wrote: >>>>>> >>>>>> On Monday, October 14, 2019 at 4:44:42 PM UTC-5, Bruce wrote: >>>>>>> >>>>>>> On Tue, Oct 15, 2019 at 5:38 AM Philip Thrift <[email protected]> >>>>>>> wrote: >>>>>>> >>>>>>>> On Monday, October 14, 2019 at 1:20:39 PM UTC-5, Brent wrote: >>>>>>>>> >>>>>>>>> Part of the dislike of the MWI is that its proponents assume a >>>>>>>>> purity that is not an evident virtue of the intepretation. For >>>>>>>>> example, >>>>>>>>> interpreting the squared amplitudes as probabilities seems to be >>>>>>>>> assumed, >>>>>>>>> along with the existence of the preferred basis in which the >>>>>>>>> amplitudes are >>>>>>>>> defined. Together these are almost the same as CI. If you ask >>>>>>>>> "probabilities of what?" in MWI the answer can't be probability of >>>>>>>>> existing >>>>>>>>> because MWI has committed to all solutions, however improbable, >>>>>>>>> existing. >>>>>>>>> So it becomes probability of finding yourself in a particular >>>>>>>>> world...which >>>>>>>>> depends on a theory of consciousness and seems to regress to von >>>>>>>>> Neumann >>>>>>>>> and Wigner. >>>>>>>>> >>>>>>>>> Zurek's envariance attempts to answer these questions and provide >>>>>>>>> a justification for preferred bases and what probability refers to. >>>>>>>>> But >>>>>>>>> notice that to the extent he succeeds he is justifying taking a >>>>>>>>> simple >>>>>>>>> probabilistic view and saying one of those preferred states happens >>>>>>>>> and the >>>>>>>>> others don't. >>>>>>>>> >>>>>>>>> Brent >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>> In the single-particle double-slit experiment*, an observer could >>>>>>>> see a dot appear anywhere on a screen where path interference does not >>>>>>>> reduce the probability to zero. So with the literal >>>>>>>> many-world-branching >>>>>>>> theory, how many different worlds are produced, each on with its own >>>>>>>> observer seeing a dot on the screen? >>>>>>>> >>>>>>> >>>>>>> According to MWI, an infinite number. Each world will have the dot >>>>>>> at a different place on the screen. >>>>>>> >>>>>>> Bruce >>>>>>> >>>>>> >>>>>> What you say may open up a bit of a hole or snag in MWI. This is >>>>>> something I have been pondering some since Carroll's popularization. If >>>>>> MWI >>>>>> fundamentally preserves unitarity by splitting off worlds then >>>>>> localization >>>>>> of a measurement is an illusion.Consider a particle measured somewhere >>>>>> on a >>>>>> path from x and x'. The path integral and the nonlocality of paths is a >>>>>> sum over all possible measurements in all space containing x and x', >>>>>> then >>>>>> there must be a continuum of possible worlds splitting off. If the >>>>>> operator >>>>>> has a continuum of eigenvalues *x*|x> = x|x> there must then be a >>>>>> continuum of possible worlds if there is indeed no fundamental >>>>>> localization >>>>>> with a measurement. This is not just infinite, but uncountably infinite. >>>>>> >>>>>> This is different from how decoherence maintains unitarity and >>>>>> conserves qubits. There a local interaction occurs that induces quantum >>>>>> phase to enter into a set of ancillary states or reservoir of states. >>>>>> Then >>>>>> we can consider quantum states as finite, but unbounded from above, so >>>>>> that >>>>>> local observations and measurements are possible. >>>>>> >>>>>> This does seem to run into some oddities that either need to be >>>>>> worked out or that might indicate some gap in MWI. The persistence of >>>>>> nonlocality in MWI is interesting for possible quantum gravitation work. >>>>>> In >>>>>> that case I can think of maybe a way around this, where this uncountably >>>>>> infinite set of g_{ij} configurations, or Ψ[g_{ij}], can be identified >>>>>> with >>>>>> "exotic" manifolds that are removed. It is less clear how this can >>>>>> happen >>>>>> with ordinary quantum fields that have local realizations. >>>>>> >>>>>> LC >>>>>> >>>>> >>>>> >>>>> >>>>> To mix an analysis (or a theory) of the path integral with an analysis >>>>> (or a theory) of MWI is mixing two fundamentally contradictory frameworks >>>>> that only leads to confusion. >>>>> >>>>> @philipthrift >>>>> >>>> >>>> I am thinking of a path integral as most physicists do, which is an >>>> action principle that is a sum over amplitudes or histories. You are >>>> thinking according to the quantum interpretation of Dowker and others, >>>> which has auxiliary postulates or assumptions. >>>> >>>> LC >>>> >>> >>> Path integrals or histories are not eve brought up in Sean Carroll's >>> book (a search of the text shows). >>> >>> So they not present in any way in MWI. >>> >>> MWI (in Sean's mathematical formulation) is contrary to the path >>> integral, because histories (as you mention above) are simply not worlds >>> (in Sean's formulation). >>> >>> @philipthrift >>> >> >> Path integrals are just methods. Three is nothing any different from QM >> or QFT without them other than a methodology. Dowker et al are jumping off >> into an interpretation based on path integrals. >> >> LC >> > > > There is no interpretation outside of methods. > > Physicists who pursue "interpretations" without method (without useful > application) are wasting their time in fantasyland, as Sean Carroll is > doing. > > See: > > https://arxiv.org/abs/1807.10749 : > > *Quantum Supremacy Is Both Closer and Farther than It Appears* > > Our algorithms can be characterized as *Schrödinger-Feynman hybrids*. > > Our simulator combines highly-optimized Schrödinger-style simulation > within each > qubit block and simulates xCZ gates with Feynman-style path summation, to > limit memory use. Unlike in Feynman-style simulation, runtime scales with > the number of xCZ gates, which is very limited in planar qubitarray > architectures with nearest-neighbor gates. Unlike traditional > Schrödinger-style simulation, the resulting algorithms are depth-limited, > and supercomputer simulations may hold some advantage for very deep > circuits. However, near-term quantum computers rely on noisy gates that > also limit circuit depth. > > > @philipthrift > > >
Or: I have yet to see any Many Worlds (Carroll's, Everett's, ...) used in the actual programming in quantum modeling software. Where is it? (Surprise me.) @philipthrift -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/257b5a5f-b09f-4f87-8dbf-d57961740a10%40googlegroups.com.
