kevin zheng 🍯

How a computer that 'drunk dials' videos is exposing YouTube's secrets YouTube is about to turn 20. An unusual research method is unveiling statistics about the platform that Google would rather keep secret.

The BBC just published an excellent dive into Ryan's (@antisomniac.bsky.social), Kevin's (@ze.vin), and Ethan's (@ethanz.bsky.social) random sampling work on YouTube, with visualizations based on data pulled from from Kevin's fantastic TubeStats project.

www.bbc.com/future/artic...

AI Companies Threaten Independent Social Media Research | TechPolicy.Press UMass Amherst researchers argue that researcher access to social media data should be one of the rare things all political factions in the US can agree on.

Really happy with this piece about the social media scraping crackdown, effects on research, bipartisan importance of ensuring such research is possible, & permissioned vs. unpermissioned research. @techpolicypress.bsky.social (with @ethanz.bsky.social & @ze.vin) www.techpolicy.press/ai-companies...

AI Companies Threaten Independent Social Media Research | TechPolicy.Press UMass Amherst researchers argue that researcher access to social media data should be one of the rare things all political factions in the US can agree on.

Researchers Ryan McGrady, Ethan Zuckerman, and Kevin Zheng argue that if we want responsible policy about social media, it begins with making it possible for American researchers to answer fundamental questions about how social media is shaping our public discourse. https://buff.ly/4hglUdV.

a 5x5 grid of the following album covers: 1. Lizzy McAlpine - Older (and Wiser) 2. Charli xcx - Brat and it’s completely different but also still brat 3. Clairo - Charm 4. Katy Kirby - Blue Raspberry 5. Carol Ades - Late Start 6. Maude Latour - Sugar Water 7. Rachel Chinouriri - What A Devastating Turn of Events 8. Abby Holliday - CRACK A SMILE COME ON STAY A WHILE 9. Madi Diaz - Weird Faith 10. Jordana - Lively Premonition 11. Katie Gavin - What a Relief 12. Michelle - Songs About You Specifically 13. Sabrina Carpenter - Short n’ Sweet 14. Remi Wolf - Big Ideas 15. Trella - expiration date 16. Grace Enger - The Alchemist 17. The Last Dinner Party - Prelude to Ecstasy 18. Annika Bennett - Live From Mother Earth 19. Melt - If There’s a Heaven 20. Orla Gartland - Everybody Needs a Hero 21. Odie Leigh - Carrier Pigeon 22. Sammy Rae & The Friends - Something for Everybody 23. Bonny Light Horseman - Keep Me on Your Mind/See You Free 24. Caity Krone - Nose Job 25. Jo Hill - girlhood.

my favorite albums of 2024 — featuring lizzy mcalpine, charli xcx, clairo, katy kirby, carol ades, maude latour, and more! (full list in alt text)

screenshot of an excerpt of the linked article on The Washington Post's website Kevin Zheng, a researcher at the University of Michigan, co-founded TubeStats.org, a website that tracks YouTube content statistics. With his collaborators, he published an academic article that tried to quantify the percentage of YouTube videos with small audiences. The paper found that there is "another" YouTube apart from the most popular accounts and videos — one where people "store video or share content with a small audience (from classroom assignments to video greetings for friends to simulcast religious services)" — and that this part of YouTube is more common than the ones that aim for virality.

TubeStats.org, our platform transparency dashboard for YouTube developed at @idpiumass.bsky.social, was referenced in @washingtonpost.com!
www.washingtonpost.com/technology/2...

happy to have helped with organizing this workers’ inquiry into “ai” at umich (graphic design is my passion 😌)

I’m excited to see what workers at other universities/workplaces can come up with when they work together to counter narratives of “ai” inevitability and address algorithmic harms at work!

AI Scams Are the Point Propaganda and deceit are a feature of AI, not its downfall.

For the December issue of The New Republic, I reviewed AI SNAKE OIL. For those learning about AI, it offers a lot: exegesis of technical details, a taxonomy to sift through different algorithmic systems, and a depoliticized history that taints the book’s analysis. newrepublic.com/article/1883...

i love tea time with max/well(s)

school of information, university of michigan Join the conversation

i'm back on bluesky...with more friends this time! check out this starter pack of all my awesome information phd colleagues at umich!! go.bsky.app/QcSqJrU

And check out the full paper in the open access @journalqd.bsky.social with @antisomniac.bsky.social, me, Rebecca Curran, Jason Baumgartner, and @ethanz.bsky.social for more on our sampling methods and analysis
journalqd.org/article/view...

Notes from Random YouTube Coding: 5 Main Takeaways - Initiative for Digital Public Infrastructure We’re excited to announce the publication of our paper, Dialing for Videos: A Random Sample of YouTube, in the Journal of Quantitative Description: Digital Media. The article is the culmination of a...

Check out @antisomniac.bsky.social's summary on the @idpiumass.bsky.social blog! publicinfrastructure.org/2023/12/21/n...

5. There are an awful lot of video games
Our hand-coding task found that nearly 30% of videos were about video games. Other topics represented a smaller part of our sample but were nonetheless surprising, like religious content, which made up about 3% of the videos we hand-coded!

4. Not everyone is participating in the “creator economy”
Low-effort videos of still photos, homework assignments, inaudible three-second clips of ceremonies, and Zoom meeting recordings — YouTube's vast "dark matter" is more variable and strange than the YouTube we're used to.

chart of view count ranges on YouTube and their relative frequencies

3. Most of YouTube doesn’t get many views
Odds are, the videos you watch have >1,000 views, but those videos were just 13% of our sample. 4.9% don’t have any views at all. Even more stark is when it comes to comments and likes — 72.6% have no comments and 88.7% have no likes.

chart of spoken languages detected on YouTube and their relative frequencies

2. YouTube is mostly not in English
Our current best estimate is that 32% of videos where we can detect spoken language are in English, with 10.5% in Hindi, 8% in Spanish, slightly fewer in Portuguese, and just over 6% in Arabic.

TubeStats TubeStats is an open-source tool that lets you navigate YouTube data and compare videos across the platform.

1. YouTube hosts over 13 billion public videos (as of this month)
Our sampling method allows us to estimate YouTube's total size and growth. This current estimate is up from our paper's year-old estimate of ~10 billion. Check out tubestats.org for our latest data!

Our team's paper, "Dialing for Videos: A Random Sample of YouTube," is out now in @journalqd.bsky.social. We analyzed 10,000 random YouTube videos through metadata analysis, a spoken language identification model, and hand-coding. Here are our 5 main takeaways:

VERY excited to delve into:

loud men talking loudly report on exclusionary cultures of internet governance – corinne cath.

https://www.criticalinfralab.net/wp-content/uploads/2023/04/LoudMen-CorinneCath-CriticalInfraLab.pdf

waking up and checking bluesky immediately

muna performing at the roadrunner in boston

went to gay church (muna show) last night 🤠

view of the smithsonian castle from the national mall on a partly cloudy day

this past weekend in dc, i learned that the smithsonian was named after some british guy named smithson??? and he’s entombed in this castle???

This morning, I ground some roasted beans then irrigated them with water.

I drank the resulting liquid. And you won’t guess what happened next.

I experienced higher levels of energy and productivity.

And I want to share this experience with you.

Are you ready to join me on this journey?

boygenius - the record lizzy mcalpine - five seconds flat laura elliott - people pleaser samia - honey caroline polachek - desire, i want to turn into you caroline polachek - pang sarah kinsley - the king reneé rapp - everything to everyone boygenius - boygenius olivia barton - this is a good sign phoebe bridgers - punisher maude latour - 001 ella jane - marginalia lexi jande - closer to closure maude latour - strangers forever trousdale - what happiness is lawrence - living room samia - the baby charli xcx - crash laufey - a night at the symphony alix page - goose julien baker - turn out the lights muna - muna sammy rae & the friends - the good life searows - guard dog

april 5x5!

me and my best friend (my roommate’s dog Bourbon) love to look wistfully out the window

oops @ethanz.bsky.social 🫣

saw caroline polachek in boston and her voice is incredible 😍

i’ll have to try these next time!! also didn’t know that costco had frozen soup dumplings, that might be a game changer…

what should i get from costco today

wearing blue pants in honor of joining blue sky (not planned)