Oliver Knill, Harvard University, October 23, 2019

1. Sean Carroll lecture

The lecture ``Something Deeply Hidden: Quantum Worlds and the Emergence of Spacetime"
organized by the Harvard book store, the division of science and the science library was a packed
lecture in Hall C. Even the stairs and side walls were filled with people. Carroll is an excellent
speaker. Unlike others before, he was funny (one running gag was ``I have a book about that"
had some new things (the Schroedinger cat is ``asleep or alive" rather than ``dead or alive" or
when in the Q and A answer time, an audience member started with ``I understood everything you said in your lecture",
Carroll interrupted ``Good start, don't ruin it with a second part", to which end the audience member
ruined it with a second part. The last question was whether he ``would bet the life of his cat
on the picture he described", Carroll answered ``I would not bet on the life of my cat, but on the
life of your cat". The lecture was also clearly delivered, the slides were good and the talk was adapted
well to a general audience (this is an art not many research scientists can do). One thing which kind of
struck me was that Carolll switched from the ``many world paper" of Everett (written as a graduate student in the 50ies)
to the de Whitt and Wheeler wave function idea which was influenced by Everett (de Whitt and Wheeler invited Everett once
to give talks source)
(which is very reasonable and probably the picture most
mathematicians take towards the quantum theory: when considering the entire system, laboratory and observer together
as a quantum system, the wave function collapse picture disappears.

I was lucky to learn QM as an undergraduate from Klaus Hepp, who would cover quite well the
topic of the Kopenhagen interpretation, Einstein-Podolski and even discussed articles
like ``is the moon there if we do not look at it?" etc, He gave good advise about the particle-wave
interpretation of QM (you have to understand this on your own) and also made
clear that QM is a model and not an ultimate theory. It obviously
does not include special relativity (leading to quantum fields) nor gravity (needing to include a geometric
frame work). So, philosophical questions about QM have since then been a bit tiresome for me. Later
as a grad student, there had been seminars about Bohmian mechanics
at the Zürich university to which I went as I had been working on
random particle pictures (Isospectral deformations of random operators leading to Vlasov type integrable systems).
But also there, for Bohmian mechanics, it was not so much the philosophy
(it turns out that it produces the classical QM picture with more intuitive notions like particles)
which is interesting as it does not lead to new insight. But it leads for example to an interesting
existence problem: does the coupled
De Broglie-Bohm system
have global solutions? [It involves taking the log of the wave function which can be zero at some points which interested me
as a mathematician. I had been working on existence theorems for infinite integrable Vlasov particle systems at
that time]. And then there is the ``Shut up and calculate!" paradigm which I always admired: if you have a new theory,
it has to be able to get something new, or substantially faster than an other theory, otherwise it is not much worth.
Physics is always tied to experiments and a theory which can not relate to experiments eventually ends up
in the garbage bin or lives on as interesting mathematics (which is valuable but not necessarily physics).

3. A martingale picture

We use relatively crude models of equilibrium statistical mechanics to describe say a 10^{25}
particle system using macrosopic mean field theories or partial differential equations, this is a way
to DO stuff. The Kopenhagen interpretation also is a genius short cut to separate the observer from the
experiment. The wave collapse picture is a point of view, a laboratory person takes. Of course, one should include the observer
into the quantum picture as everything is entangled, but we do not do that because it usually does not
matter. QM is an extremely successful theory. Why ruin it with philosophy? One reason of course is that
we want to understand the merger QM+GR where it might be good to question old paradigms.
But we do not write philosophical essays about ``the interpretation of Newtonian mechanics". It would be tiresome because
it is obviously also just a model, an approximation. Early in the lecture
Carroll warned about trying to get into the field of ``understanding QM". This is good advise. Any model in physics has
a range of experiments, it can explain (which makes it valuable) and then boundary areas where it starts to fail.
Trouble comes with physical theories which can not be confirmed by experiments. The Everett theory was for me one of these ``worthless
theories". The lecture of Carroll a bit changed my point of view. He painted it not as a
multiple universe theory as popular articles
like this pointed out, but as a picture
in which a fixed universe is described by in time finer and finer partitioned space. It is somehow a Martingale picture of the
universe. The sigma algebra gets finer and finer. So, for me, it was the Martingale picture which was new
in that talk and not the many world interpretation which was Everett's original paper.