Episode #37 is out! This time we talk with @GuannanWei to talk about his compilers research and we dive into the technicalities of his work. We touch on topics such as Staging, Futamura Projections, Symbolic Execution, Quantum Programming and more!
typetheoryforall.com/2024/03…
Couple of fresh Swift compilers one for Xcode and one for research -- both doing automatic first class differentiation. Ho ho! Merry Christmas! Santa Claus came early this year. github.com/apple/swift/blob/…
My 2024 shopping list: strong static typing, aot compiler, automatic first class differentiation and a smart array based rts that can put the data on the best device for the job. Dependent types a JIT and repl are a bonus.
Geometry grew out of the study of astrology and architecture, calculus from the study of mechanics. Could quantum theory also lead to a new kind of mathematics? Follow the discussion in the newly available #ThePrimeNumberConspiracy!
amazon.com/dp/0262536358/
Thinking how iterated functions that reveal (deterministic) chaotic behaviour and the resulting fractal maps of their attractors could be very interesting to explore w.r.t. encoding memory in machine learning. I'm already impressed with simple holographic (SDM) features.
An inspiring approach: science applied to the data - looking for causation in the signal and dealing intelligently with noise. x.com/quantamagazine/status/…
ICYMI: Black-box machine learning often isn’t adequate for understanding medical genomics. Computer scientist Barbara Engelhardt explains why.
buff.ly/2HOPpmN
“Progress here is not just about discovering a new particle. It is also — indeed, most of the time — being able to carve out the space of what might be possible in nature.”
buff.ly/2IJaaV1