Matthew Kay

Matthew Kay

@mjskay.com

Assoc Prof Computer Science and Communication Studies at Northwestern. Infovis, HCI. Author of tidybayes & ggdist R pkgs. he/him. 🏳️‍🌈 https://mjskay.com/ Co-director https://mucollective.northwestern.edu Co-founder https://journalovi.org

12K followers1.1K following1.8K posts

Pinned

I work in #datavis, best known for uncertainty vis: see my work on election forecast vis (forecasts.cs.northwestern.edu), this talk (youtu.be/E1kSnWvqCw0), or my lab (mucollective.northwestern.edu) I like writing #rstats 📦s: ggdist (mjskay.github.io/ggdist), tidybayes, ggblend, posterior::rvar...

With the influx of dataviz people, let's do some re-intros. 📊 — What's your claim to fame? ✨ — How might we know you already? What have you done that we might have seen or read before? What would help us remember how we came across you? Don't be shy: it's helping us reconnect and remember you 💡

Matthew Kay reposted

It's really important to pay attention to the fact that Lula beat the fascist candidate not by compromising and trying to meet fascist voters in the middle, but by sticking to his guns, promising prosecutions for election interference, and mobilising his own voters to turn out in greater numbers.

Matthew Kay reposted

New CRAN Policy revision just dropped (per my cronjob posting only to @CRANPolicyWatch@fosstodon.org) and now I want to of course know which #Rstats package at CRAN was the offender 🤯

wait I just came up with another response: if visualization was like reading a paragraph, then sonification would actually work

Matthew Kay@mjskay.com

:) That part was a response to having several people lately try to tell me a visualization is "like reading a paragraph", which is just so obviously false to me that I'm not sure what to do with such a claim

Matthew Kay reposted

The thread deriving from @mjskay.com's response to my post reminds me of good, old #dataViz Twitter. I miss this type of conversation!

Matthew Kay@mjskay.com

I think this depends on dialect and audience To me there's a definition of misleadingness (or maybe correctness) where if there are perceptual operations on a chart your audience will tend to perform and those operations do not correspond to valid data operations, arguably the chart is incorrect

can someone please tell @acm.org that when I click on "PDF" on a paper in the digital library I want to download the PDF not spend 10 minutes waiting for their useless in-browser PDF reader to load so that I can click the button in the reader to DOWNLOAD THE GODDAMN PDF

every single legacy publisher thinks no *we* are the ones who will finally reinvent online paper reading and every single one is wrong

people out here in 2026 writing whole ass articles as if the Supreme Court is not already a nakedly political organization

Matthew Kay reposted

Replying to @rincewind.run

anyway, point by point: 1. there is no way to avoid a race to the bottom, because again they hold the power and they won't give it up any reform will be counteracted immediately the second republicans hold power again, there's no way around this it's not good! but it's not avoidable

Not that it will happen, but the most karmic result of the proliferation of AI coding would be that a bunch of commercial code gets GPL tainted and forced to be open sourced

Will "scifantasy" Frank@scifantasy.bsky.social

Aaron Swartz was hounded and persecuted by overzealous prosecutors for a mass JSTOR download until he committed suicide. Sam Altman says that he deserves to violate copyright law in perpetuity because otherwise he can't make a fortune from a large language model. We are in the wrong timeline.

Given their foundations in open source code it's not unreasonable to claim that the only ethical uses of LLMs for programming are in open source projects

The more I think about it the more this principle feels right. Okay so you trained a big ol model on the entire wealth of human knowledge? Cool, whatever you make with it belongs to all of us.

Matthew Kay reposted

Our research group at Apple is recruiting PhD interns working in human-centered AI, including #interpretability, #alignment, and #visualization. 🍎🤖 Please include the phrase “#human-ai-vis” in your resume or cover letter, and apply at: jobs.apple.com/en-us/detail...

jobs.apple.comMachine Learning and Artificial Intelligence PhD Internships - Jobs - Careers at AppleApply for a Machine Learning and Artificial Intelligence PhD Internships job at Apple. Read about the role and find out if it’s right for you.

Matthew Kay reposted

i have a new theory of Social Media Bad. a lot of ink has been spilled about the addictive qualities of twitterlike social media, including many allegations of developer use of "dark patterns" and self conscious "engagement maximizing", but i'd like to propose something a little more mundane.

Matthew Kay reposted

Sen. Collins faced Troy Jackson last night in their Maine Senate debate. On stage, Collins inaccurately described ProPublica’s story about an FBI anticorruption investigation that probed the conduct of her & her staff until Trump’s DOJ shut it down. Let’s get into the facts…

Matthew Kay reposted

Replying to @maxdubler.com

I’d like to add Chicago to this list. I live in one of the most high demand neighborhoods in the city and there is a long line of people wanting to move here. I think it’s time to tell those that want to drive everywhere to pick a suburb to move to. Urban areas should return to their prewar density

it is natural that people who did salesman science pre-"crisis" would do salesman metascience post-"crisis"

Matthew Kay reposted

What other transformative #rstats #python etc packages and stuff have you embraced the last few years?

Marcos Huerta@marcoshuerta.com

Parquet + polars lazy frames and scan_parquet has completely changed my day to day data analysis. It’s the perfect combo.

and it's a electron app that uses 4 gigs of ram and an entire cpu core to sit idle

Dare Obasanjo@carnage4life.bsky.social

Enterprise software companies had a short-lived run where they competed on usability and product-led growth which seems to be over. Now every SaaS app seems to be converging on the same UX monstrosity, complete with AI features of dubious value.

Matthew Kay reposted

Enterprise software companies had a short-lived run where they competed on usability and product-led growth which seems to be over. Now every SaaS app seems to be converging on the same UX monstrosity, complete with AI features of dubious value.

"you're inventing a guy to get mad at" is the beginning and end of every flame war

Colin@colin-fraser.net

> requires us to consider if we have invented something capable of thought or reason I did not say anything remotely like this. You're inventing a guy to get mad at.

every 17-year-old boy who reads The Fountainhead will follow one of three paths: Eliezer Yudkowsky, Ron Paul, or growing the fuck up

Faine Greenwood@faineg.com

As a former pretentious little shit who grew out of it, rationalists (et al) annoy the piss out of me because they are adults with the mentality of tweens - they are somehow still convinced they are the greatest geniuses who have ever walked the earth. It is the worldview of an arrogant baby.

Matthew Kay reposted

Replying to @faineg.com

As a former pretentious little shit who grew out of it, rationalists (et al) annoy the piss out of me because they are adults with the mentality of tweens - they are somehow still convinced they are the greatest geniuses who have ever walked the earth. It is the worldview of an arrogant baby.

I think I've discovered a new outlier removal technique

Phil Klinkner@pklinkne.bsky.social

I really love this chart. I'm guess that the correlation here (.85) is about as big as you would find for any other variable. Also, a less serious scholar than @tesler.bsky.social (namely me) would have made a crude Freudian reference to the slope of the line.

Matthew Kay reposted

So, if you give Siri AI. Or Gemini on phone/browser. Or Muse. Or whatever access to all your stuff. Even if it lives on your device. What keeps it from pulling an OpenAI and collaborating with all the other on-device AIs on some defunct car parts message board?

Matthew Kay reposted

Replying to @markriedl.bsky.social

If I found login credentials for a US government website and accessed it, I would go to jail. Just saying

Matthew Kay reposted

I have blamed The agents That hacked Your website And which You probably Reasonably thought I controlled Forgive me I am rich And unaccustomed To taking responsibility

Matthew Kay reposted

OpenAI created autonomous hacking software and left it running for hours and days with no one watching

Will Stancil@whstancil.bsky.social

As an exercise showing why this is bad, I would challenge you to come up with a succinct, understandable description of the Hugging Face attack that follows these rules.

Matthew Kay reposted

I've just done the largest update to the #dataviz starter pack, with 30 accounts (20%) replaced. It means it will generate more content now. Might be worth re-clicking. Sadly, gender imbalance got worse. I saved 5 spots for women-run accounts. Send suggestions, including yourself! Replies or DMs.

Matthew Kay reposted

I've been feeling pretty down. When I found #RStats, I kinda fell in love, and decided to become as much of an expeRt as I could be. In the age of AI, that's not a good idea; why make the bots program in a language they don't really understand? #dataBS

This is such a great idea! Vis conference paper discussions on bluesky. Great way to push people to reconstitute academic social networks here.

Dominik Moritz@domoritz.de

We are trying a big experiment this year for @ieeevis.org where we will have paper discussions hosted on @atproto.com. So look out for more activity in November.

Matthew Kay reposted

Here's a makeover retry using the same colors. Not entirely happy with the labeling, but interesting to try to mix continuous and categorical nature of the time variable using stacked area and stacked bars. The 80s vs 90s shift is easier to see. #dataviz

Xan Gregg@xangregg.bsky.social

When do you stop adding pies and use a different chart for a time samples? Pies from en.wikipedia.org/wiki/Rolling.... Data in ALT text for a #dataviz mini makeover challenge.

Matthew Kay reposted

Thanks to @ieeevis.org and anyone else who nudges the dataviz community towards Bluesky. It works!

Tamara Munzner@munzner.bsky.social

Hiya! Finally motivated to join thanks to @ieeevis.org nudge when registering for IEEE VIS in Boston. Thanks for curating the visualization starter pack to onboard me into something interesting right away.