Magos Logis Logan @jmw150@qoto.org

I think my family just got used to these things. Lol

This paper is great. It is like a good slice of what is going on in formal methods and machine learning right now.

- Methods of space convergence from AI, can, and do, outperform the brightest algorithm writers, especialy when the time complexity is high. And innacuracy is easy to control for.

- It is all just programming. It is research level, but still. ML algs can be used easily in code logic, if there is a layer to handle uncertainty.

- Many of the referenced papers are about software that generates code without programmers. Obviously computer vision is going to beat any human, but the best tech for other domains are referenced here. Much of it is about indirect code reasoning.

- Real data is not needed for ML (supervised) research to progress. This data, CLEVR, is synthetic. Sometimes richer synthetic datasets make a better playground.

http://pages.cs.wisc.edu/~aws/papers/icml20b.pdf

I hope I can be as creative as Nikola, dreaming up cellphones before we even had digital computers.

Occam's razor does not apply to hypothesis spaces.

Study broadly.

Now this is a proper use of tenure.

This whole lab is a big middle finger to USA government, and large company, censorship and spying.

https://freedom.cs.purdue.edu/

That is more like it. Make the computer figure out security.

https://link.springer.com/article/10.1007/s10703-017-0296-5

So... encrypted LVM on Debian takes 30+ hours to set up.

This is a fresh SSD. Why?...

Math people either love or hate statistics.

What I have found that helps me is to think of the various theorems from the perspective of functions on spaces. Instead of the messy pseudo-math that intro courses present it as, there are underlying measure spaces. Past all the word salad, spaces are what is tossed around and modified to reason about reality.

My computer colleges and their macbooks and windows machines spying on them... Do you even live bro?

yeah, coffee crash. My face hurts from pure tired.

Hmm... 3x

Hmm!

Science vs engineering research really has that old "created vs discovered" vibe math has.

His erdös number is either 2 or 3

A research buddy cannot figure out video conference software. So time to drive an hour and a half.

You know, discrete geometry is actually pretty great.

Many regular geometry things made it in. It is super easy to program with. And it has some of its own nice things too.

Wish I had some of that spice from Dune. Caffeine is only so good.