J'apprends qu'elle est décédée aujourd'hui (à 101 ans, donc).

A fun fact: diamond melts at 4500 C, while graphite melts at 3600 C I'd be curious what happens when you bring solid diamond at (say) 3900 C into contact with liquid graphite at 3900 C. Does the graphite crystallize into diamond? Do any chemists happen to know? (Real life

(2/2) -Explicit binary linear codes over matrices that tolerate any δ fraction row/column erasures at optimal rate (1-δ)^2 -Explicit codes approaching the Singleton bound arbitrarily closely over a constant alphabet (similar to AEL) -Improved erasure codes on undirected graphs.

i have been dealing with a nagging dread that someone needs to 10X-100X the capacity of the ArXiV asap... when it hits (and it's gonna hit), we are at risk of preprint publication grinding to a halt. (for-profit journals are ngmi)

China builds world’s first working thorium reactor using declassified US documents | Aamir Khollam, Interesting Engineering In the Gobi Desert, China has powered up the world’s only working thorium reactor. Chinese scientists have completed a major breakthrough in clean energy

is there a standard name for the representation of U(n) on the n-mode Fock space that associate to U \in U(n) the unitary {a_i} --> {\sum_j U_{i,j} a_j} ? "metaplectic representation"?

High-performance local automaton decoder for defect matching in 1D, happy to share that our new work in with Anthony Leverrier, Mazyar Mirrahimi and Dr Christophe Vuillot is now available on the arXiv ! 👀 (1/6)

#newpaper Bosonic codes can dramatically reduce the cost of fault tolerance and many codes look like serious contenders: dual-rail qubit, cat qubit, GKP. But the Fock space is huge, esp. if you consider multimode encodings, and there are many great codes waiting to be found (1/4)

I'm often asked if I'll redo the 2019 quantum factoring estimate. Denser storage by yokes, smaller magic factories by cultivation, slimmer approx arithmetic by Chevignard et al… surely the cost is lower now? Yes, it's lower now. security.googleblog.com/2025/05/tracki… arxiv.org/abs/2505.15917

What a great initiative !

also, here is one part that people not interested in the rest of the post might still be interested in:

Great work by Yiyi Cai that improved the analysis of robustness for the local codes used in quantum Tanner codes. (To be fair, the bounds remain quite pessimistic if one wishes good quantum LDPC codes with linear distance.) thesis.library.caltech.edu/17412/

Anthony Leverrier Relates to the fact that the strength of the architecture is also a weakness--data and instructions stored in the same memory. CPU is fast so has to wait around for data. en.wikipedia.org/wiki/Von_Neuma…

might apply to quantum computing startups as well...

GPT-5 didn’t live up to OpenAI’s hype, but it is *exactly* in line with extrapolations from prior AI advancements. Go ahead and discount future statements from OpenAI/Altman, but you should still expect the fast AI progress that we’ve been seeing to continue

ha! you meant *those* nails 🤦