Gay Maths Major

jenny slate:

me:

• updates on how much I’m procrastinating

• pictures of my ridiculously disorganized study spaces

• my messy fucking notes that no one but me can read and bad grades/embarrassing comments from professors

I’ll put a ton of my own stuff in the queue but I’m really hoping this can be a community-oriented thing where people can submit their own ridiculous anti-studyblr stuff

There’s an application for neural nets called “photo upsampling” which is designed to turn a very low-resolution photo into a higher-res one.

This is an image from a recent paper demonstrating one of these algorithms, called “PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models”

It’s the neural net equivalent of shouting “enhance!” at a computer in a movie - the resulting photo is MUCH higher resolution than the original.

Could this be a privacy concern? Could someone use an algorithm like this to identify someone who’s been blurred out? Fortunately, no. The neural net can’t recover detail that doesn’t exist - all it can do is invent detail.

This becomes more obvious when you downscale a photo, give it to the neural net, and compare its upscaled version to the original.

As it turns out, there are lots of different faces that can be downscaled into that single low-res image, and the neural net’s goal is just to find one of them. Here it has found a match - why are you not satisfied?

And it’s very sensitive to the exact position of the face, as I found out in this horrifying moment below. I verified that yes, if you downscale the upscaled image on the right, you’ll get something that looks very much like the picture in the center. Stand way back from the screen and blur your eyes (basically, make your own eyes produce a lower-resolution image) and the three images below will look more and more alike. So technically the neural net did an accurate job at its task.

A tighter crop improves the image somewhat. Somewhat.

The neural net reconstructs what it’s been rewarded to see, and since it’s been trained to produce human faces, that’s what it will reconstruct. So if I were to feed it an image of a plush giraffe, for example…

Given a pixellated image of anything, it’ll invent a human face to go with it, like some kind of dystopian computer system that sees a suspect’s image everywhere. (Building an algorithm that upscales low-res images to match faces in a police database would be both a horrifying misuse of this technology and not out of character with how law enforcement currently manipulates photos to generate matches.)

However, speaking of what the neural net’s been rewarded to see - shortly after this particular neural net was released, twitter user chicken3gg posted this reconstruction:

Others then did experiments of their own, and many of them, including the authors of the original paper on the algorithm, found that the PULSE algorithm had a noticeable tendency to produce white faces, even if the input image hadn’t been of a white person. As James Vincent wrote in The Verge, “It’s a startling image that illustrates the deep-rooted biases of AI research.”

Biased AIs are a well-documented phenomenon. When its task is to copy human behavior, AI will copy everything it sees, not knowing what parts it would be better not to copy. Or it can learn a skewed version of reality from its training data. Or its task might be set up in a way that rewards - or at the least doesn’t penalize - a biased outcome. Or the very existence of the task itself (like predicting “criminality”) might be the product of bias.

In this case, the AI might have been inadvertently rewarded for reconstructing white faces if its training data (Flickr-Faces-HQ) had a large enough skew toward white faces. Or, as the authors of the PULSE paper pointed out (in response to the conversation around bias), the standard benchmark that AI researchers use for comparing their accuracy at upscaling faces is based on the CelebA HQ dataset, which is 90% white. So even if an AI did a terrible job at upscaling other faces, but an excellent job at upscaling white faces, it could still technically qualify as state-of-the-art. This is definitely a problem.

A related problem is the huge lack of diversity in the field of artificial intelligence. Even an academic project with art as its main application should not have gone all the way to publication before someone noticed that it was hugely biased. Several factors are contributing to the lack of diversity in AI, including anti-Black bias. The repercussions of this striking example of bias, and of the conversations it has sparked, are still being strongly felt in a field that’s long overdue for a reckoning.

Bonus material this week: an ongoing experiment that’s making me question not only what madlibs are, but what even are sentences. Enter your email here for a preview.

writing article: Hmmmmm not as fun for sure but like. The process.  helps a little to put things into words.  5/10 i guess it’s o k

editing article: I crave the sweet release of death.  Endless years I have suffered - tell me when it will end?  Language is meaningless and I have lost sight of my path.  Here alone in the desert I wander, searching the empty expanses of my mind for a synonym.  How to tempt Word into formatting my diagrams without ruinous consequence?  I am the master of my own garden of sorrow.

Newton: *proceeds to write calculus derivation on months-old paper*

“I always remember having this fight with a random dude who claimed that ‘straight white men’ were the only true innovators. His prime example for this was the computer… the computer… THE COMPUTER!!! THE COM-PU-TER!!!

Alan Turing - Gay man and ‘father of computing’ Wren operating Bombe - The code cracking computers of the 2nd world war were entirely run by women Katherine Johnson - African American NASA mathematician and ‘Human computer’ Ada Lovelace - arguably the 1st computer programmer”

- Sacha Coward

Also Margaret Hamilton - NASA computer scientist who put the first man on the moon - an as-yet-unmatched feet of software engineering, here pictured beside the full source of that computer programme. #myhero

Grace Hopper - the woman that coined the term “bug”  

also computing was typically a job for women (many of whom were black women that made incredible contributions) back in the day, so it’s absolutely fucking wild that straight white men think they are the foundation of computer innovation. men PUSHED women out and took the credit.

× multiplication

÷ un-multiplication (sometimes)

√ gives the bottom half of a plant

≈ not equal, but we wish it was

≝ just like, accept it, man

≡ not equal, but like it’s kinda similar

≤ hungry alligator in water

⇒ look this way

↦ look this way, now draw a picture

⇔ look punk these things go together, that’s just how it is

¬ droopy negative sign

∨ ice cream cone

∧ empty ice cream cone :(

⊕ protected plus

∀ the sound of an Australian screaming

∃ the sound of a confused Australian

∈ toucan beak

∉ no toucan beaks

⊂ you can fit this inside the other thing

⊄ you can’t fit this thing in the other thing

⊆ you can fit this thing in the other thing but it might be tight

∪ full bucket

∩ empty bucket

∅ nuttin’

α β γ you’re working with angles

δ ε you’re working on analysis

ζ η you’re working with complex numbers

θ κ π τ circles bitch

λ you’re working with linear algebra

μ ν ξ ????

Δ∇ what direction is this curve going?

∏ ∑ author is too lazy to write out repeated operations

∞ binoculars to see really far

∫ ∮ ∯ time to add up lotsa itty bitty things

ℂ numbers too complicated

ℕ the positive integers, give or take a zero depending on how cool the author is

ℚ integers over and under cute lil lines :3

ℝ nigce smoot h line

ℤ author having a nap