My feelings when working with tensors:

Posting Tits on the timeline

Been listening to this on repeat for 3 days. I think they put crack in it. open.spotify.com/track/6r2zoEGc…

You just get used to it

The algorithm runs on O(N^2)? No one’s got time for that!

I foresee a need for a family of algorithms that used to be run globally to be done locally. Specifically in analyzing long time series. No one’s got time for that O(N^2) BS. Looking at you, Synchrosqueeze Transform.

Fortunately, algorithms do what I tell them to do. Unfortunately, algorithms do what I tell them to do.

Good to read this once in a while hep.upenn.edu/~johnda/Papers…

The Morning Show Season 3 can be summarized as follows: “No, Elon Musk, you can’t buy out legacy media!” Brought to you by CBS and Apple

I asked ChatGPT why it’s so hard to find synergy in a particular collaboration. It answered. And then I asked it why it’s so easy. It answered. Totally changed my perspective

You know that moment when you write a more serious tweet and then delete it coz there’s already so much noise out there? Yeah, me neither

RIP Charlie Kirk

RIP GM Daniel Naroditsky

Common sayings I used when tackling grad analysis problems: - when in doubt, integrate by parts (IBP) - swapping limits and integrals? Dominated Convergence Theorem (DCT) - contour integration? Jordan’s Lemma & Residue theorem - Solving non-homogeneous PDEs? Green’s function

AI got a feeling!