Tsirelson Memorial Workshop

(thanks to Flavio Baccari for the photo)

After more than two years of the world conspiring against me and Miguel, we finally managed to pull it off: the workshop in honour of Boris Tsirelson happened! In my opinion and of the participants that I asked it was a resounding success: people repeatedly praised the high level of the talks, and were generally happy with finally having a conference in person again. An often-overlooked but crucial part of every conference is the after-hours program, which is rather lame in online conferences if it happens at all. I didn’t really take part here as I have a young child, but the participants told me that it was quite lively. We did have problems because of the recent corona wave in Austria and the war in Ukraine1, but still, it happened.

Initially we had planned to have a small workshop, but we ended up with 78 registered participants (and some unregistered ones). It was great having so much interested, but it did create problems. The idea was to have a small amount of long talks, where the authors could go through their results in depth, and people would have plenty of time for discussion. We kept the talks long (45 minutes), but we ended up with a completely packed schedule (9 invited talks + 22 contributed). We thought this wouldn’t be a problem, as people could simply skip the talks they were not interested in and use that time for discussion. That didn’t work. It turns out students felt guilty about skipping talks (I never did), and there wasn’t a good place for discussing in our conference venue. We apologize for that. Another issue is that we had to review a lot of contributions (44); thanks to a small army of anonymous referees we managed to get through them, but we still had to do the painful job of rejecting good contributions for lack of space.

A curious feedback I got from some participants is that the talks were too long. The argument was that if you are not familiar with the topic already you won’t be able to understand the technical details anyway, so the extra time we had to go through them was just tiresome. I should do some polling to determine whether this sentiment is widespread. In any case, several long talks on the same day are indeed tiresome; perhaps we could reduce the time to 30 minutes. What I won’t ever do is organize a conference with 20-minute talks (which unfortunately happens often); most of the time is spent in introducing the problem, the author can barely state what their result was, and there’s no chance of explaining how they did it.

There were two disadvantages of organizing a conference that I hadn’t thought of: first, that even during the conference I was rather busy solving problems, and couldn’t pay much attention to the talks, and secondly that I couldn’t present my own paper that I had written specially for it.

As for the content of the talks, there were plenty that I was excited about, like Brown’s revolutionary technique for calculating key rates in DIQKD, Ligthart’s impressive reformulation of the inflation hierarchy, Farkas and Łukanowski’s simple but powerful technique for determining when DIQKD is not possible, Plávala’s radically original take on GPTs, Wang’s tour de force on Bell inequalities for translation-invariant systems, Sajjad’s heroic explanation of the compression technique for a physics audience, among others. But I wanted to talk a bit more about Scarani’s talk.

He dug out an obscure unpublished paper of Tsirelson, where he studied the following problem: you have a harmonic oscillator with period $\tau$, and do a single measurement of its position at a random time, either at time 0, $\tau/3$, or $2\tau/3$. What is the probability that the position you found is positive? It turns out that the problem is very beautiful, and very difficult. Tsirelson proved that in the classical case the probability is at most $2/3$, but ironically enough couldn’t find out what the bound is in the quantum case. He showed that one can get up to 0.7054 with a really funny Wigner function with some negativity, but as for an upper bound he only proved that it is strictly below 1. Scarani couldn’t find the quantum bound either; he developed a finite-dimensional analogue of the problem that converges to the harmonic oscillator in the infinite limit and got some tantalising results using it, but no proof. The problem is still open, then. If you can prove it, you’ll be to Tsirelson what Tsirelson was to Bell.

This entry was posted in Uncategorised. Bookmark the permalink.