Tootfinder

An analysis of 1,000 YouTube Shorts videos recommended to kids: the algorithm pushes bizarre, often nonsensical, AI-generated videos, and many lack disclosures (Arijeta Lajka/New York Times)
https://www.nytimes.com/2026/02/26/us/ai-videos-children-youtube.html…

How A.I.-Generated Videos Are Distorting Your Child’s YouTube Feed
Experts caution that low-quality, A.I.-generated videos on YouTube geared toward children often feature conflicting information, lack plot structure and can be cognitively overwhelming — all of which could affect young children’s development.

Internal documents show Instagram's aggressive campaign to win back teens, including by boosting teen-friendly influencers, adjusting its algorithm, and more (Naomi Nix/Washington Post)
https://www.washingtonpost.com/technology/2025/12/26/meta-instagram…

Inside Instagram’s all-out battle to win the nation’s teens
Instagram has pursued a yearslong strategy to win teens back to the app after its critics said the platform wasn’t safe for them, documents obtained by The Post show.

The "Computational Theory of Mind" has become the modern secular religion, promising that if we just find the right algorithm, we can replicate intelligence.
And it is wrong ...
https://www.ocrampal.com/computation-is-limited-the-hidden-f…

Computation is not the right framework to comprehend the concept of life and intelligence
We live in a computational age. When we want to formalize our thoughts—to express ideas precisely, to build models of intelligence, to create systems that reason—we almost instinctively reach for computation. The Church-Turing thesis tells us that any "effectively computable" function can be computed by a Turing machine,

TikTok US says it is working to restore services after a power outage at a data center and any algorithm changes users noticed were likely due to the outage (Dan Whateley/Business Insider)
https://www.businessinsider.com/tiktok-outage-data-cent…

TikTok says a power outage messed up users' feeds over the weekend
TikTok said a power outage at a US data center on Sunday impacted its services, including its For You feeds.

I won’t lose any sleep over it, but for some reason YouTube’s algorithm is showing me way too many (any is too many) ads for women’s deodorants lately.

Internal documents: Instagram execs have waged an aggressive campaign to win back teens, boosting teen-friendly influencers, adjusting its algorithm, and more (Naomi Nix/Washington Post)
https://www.washingtonpost.com/technology/2025/12/26/meta-instagram-te…

Inside Instagram’s all-out battle to win the nation’s teens
Instagram has pursued a yearslong strategy to win teens back to the app after its critics said the platform wasn’t safe for them, documents obtained by The Post show.

Gizmodo reports on X recently releasing a significant part of the source code for its ordered post display algorithm. Elon Musk characterizes the algorithm as "dumb."
"Elon Musk Makes Part of X Algorithm Open Source, Says It ‘Sucks’"
https://gizmodo.co…

Elon Musk Makes Part of X Algorithm Open Source, Says It ‘Sucks’
He also says it's dumb.

The political effects of X's feed algorithm #politics #SocialMedia

The political effects of X's feed algorithm - PubMed
Feed algorithms are widely suspected to influence political attitudes. However, previous evidence from switching off the algorithm on Meta platforms found no political effects¹. Here we present results from a 2023 field experiment on Elon Musk's platform X shedding light on this puzzle. W …

Disappointing to see that even Electric Sheep has a slop angle now https://electricsheep.org/

AMC Networks' Sundance Now rebrands from offering scripted TV to airing indie film and festival titles, for $7.99/month, as this year's Sundance Festival begins (Mia Galuppo/The Hollywood Reporter)
https://www.hollywoodrepor…

Sundance Now to Rebrand as “Algorithm-Free” Streaming Service for Indie Film
AMC Networks runs the shingle that previously had scripted TV offerings.

"Satellite images indicate that the Doñana Marshland will disappear within 60 years"
#Environment
https://phys.org/news/2026-02-satellite-images-doana-…

Satellite images indicate that the Doñana Marshland will disappear within 60 years
The Doñana National Park, considered one of Europe's most valuable wetlands, is expected to lose its marshland in 61 years, according to calculations from a major water-resource monitoring study carried out by the University of Seville. The study has developed an innovative algorithm, based on machine learning, capable of detecting the presence of surface water with high precision using images from the Sentinel-2 satellite.

The political effects of X’s feed algorithm
https://www.nature.com/articles/s41586-026-10098-2

The political effects of X’s feed algorithm - Nature
Among users initially on a chronological feed, 7 weeks of exposure to X’s algorithmic feed in 2023 shifted political attitudes and account-following behaviour in a more conservative direction compared with those remaining on a chronological feed, whereas switching the feed setting in the opposite direction, from algorithmic to chronological, had no effect.

I went to see the Claude Desktop announcement, and they embed a YouTube Video, and this is what the algorithm thinks I should watch next:

New Study: Evidence shows how X’s algorithmic feed is radicalizing its users toward the Right. https://www.nature.com/articles/s41586-026-10098-2

The political effects of X’s feed algorithm - Nature
Among users initially on a chronological feed, 7 weeks of exposure to X’s algorithmic feed in 2023 shifted political attitudes and account-following behaviour in a more conservative direction compared with those remaining on a chronological feed, whereas switching the feed setting in the opposite direction, from algorithmic to chronological, had no effect.

China's Cyberspace Administration requires firms to file their AI tools in a public algorithm registry, creating a detailed map of the country's AI ecosystem (Yi-Ling Liu/Wired)
https://www.wired.com/story/china-ai-boom-algorithm-registry/

Thousands of Companies Are Driving China’s AI Boom. A Government Registry Tracks Them All
How the Cyberspace Administration of China inadvertently made a guide to the country’s homegrown AI revolution.

An experiment finds X's feed algorithm favored conservative content, and switching to the "For you" feed shifted users' views toward more conservative positions (Ece Yildirim/Gizmodo)
https://gizmodo.com/researchers-find-t

X's Algorithm Pushes Users to Lean More Conservative, Researchers Find
The algorithm also doesn't like posts by traditional news media, the researchers found.

Why does calling out a hallucination feel like calling someone’s kid "ugly"?
We’ve stopped treating AI like software and started treating it like a protégé. Here’s why criticizing an algorithm has suddenly become so personal.
https://www.ocrampal.com/why-criticizing-a

Why Criticizing AI Feels Like Criticizing Someone’s Child
I’m not questioning the usefulness of LLMs and AI in general. What I am questioning is its ability to achieve intelligence. Large Language Models (LLMs) are remarkable tools—tools to augment one’s intelligence—just as a search engine, or Wikipedia have done before them. I use them daily.

The US' TikTok deal is a win for ByteDance: it will keep and license the algorithm instead of selling it, and continue to run TikTok's commercial activities (Jim Secreto/Financial Times)
https://www.ft.com/content/59b91fc8-03a1-48df-9821-e2fdff24bd33
<…

X says it has open sourced its "core recommendation system" on GitHub and that the system "ranks everything using a Grok-based transformer model" (Lucas Ropek/TechCrunch)
https://techcrunch.com/2026/01/20/x-op

X open sources its algorithm while facing a transparency fine and Grok controversies | TechCrunch
In a post to GitHub on Tuesday, the social media giant purported to share its secret sauce.

BRB, rewriting all my graph algorithms https://www.quantamagazine.org/new-method-is-the-fastest-way-to-find-the-best-routes-20250806/

New Method Is the Fastest Way To Find the Best Routes | Quanta Magazine
A canonical problem in computer science is to find the shortest route to every point in a network. A new approach beats the classic algorithm taught in textbooks.

For all newcomers: please use hashtags. (reasonable ones would be cool 😉). There is no algorithm that will push your content.

Here is an updated version of the #NFDI graph: large nodes are consortia, small nodes are institutions. Graph created with data from #Wikidata, visualized with #Gephi, using force atlas algorithm and modularity class c…

This was fun. But I can clearly optimize the algorithm once I learn #Gleam better.
I just completed "Playground" - Day 8 - Advent of Code 2025 #AdventOfCode

Need some data to test a plot idea or algorithm? On #rstats #synthetic…

An experiment finds X's feed algorithm favored conservative content, and switching to the "For you" feed shifted users' views toward more conservative positions (Ece Yildirim/Gizmodo)
https://gizmodo.com/researchers-find-t

X's Algorithm Pushes Users to Lean More Conservative, Researchers Find
The algorithm also doesn't like posts by traditional news media, the researchers found.

This is a really good assessment (why people are complaining BlueSky is “mean”), IMO.
It’s also true of Mastodon, for very similar reasons.
If I may mangle a line from “Grosse Pointe Blank”:
“if people are hot in your mentions, you did something to bring them there”
OTOH, BS and Masto have tools to manage that. (though they need improvement)

Whenever I see yet another #AI "AGENTS" file, trying to write instructions for *machines* in human language, like the #LLM statistical algorithm could actually reason about them, a Butlerian jihad opens in my pocket. And the fact of giving clear instructions like they were talking to an #ActuallyAutistic person is adding insult to the injury.
#NoAI

I discovered Reviewbrah on YouTube a month ago. Not sure why it took the algorithm so long to connect us given that I love non sequitur / absurdist humor.
“My disappointment is immeasurable, and my day is ruined.”
#ReviewBrah #ScottHumor

Deezers 'we think you would like' algorithm is great! Re-found this album #paradiselost

Memo: the TikTok US deal is set to close on Jan. 22; terms include retraining the recommendation algorithm on US user data and Oracle overseeing data protection (Alex Weprin/The Hollywood Reporter)
https://www.hollywoodreporter.com/business

TikTok U.S. Deal to Close Next Month: Oracle and Silver Lake Confirmed in Buyer Consortium
Abu Dhabi's state investment fund will also be a managing investor.

> #X is no longer an open and balanced tool for political #communication or #journalism, since algorithm changes in the wake of Elon #Musk’s takeover. … The @… and national governments should not communicate on a platform where women cannot participate in the debate without risking image-based sexual violence.
…
Our national #security and #democratic health depend on citizens being able to access and participate in open, inclusive debate, and without…
3/4

In times like these I'm just SO grateful that the fediverse doesn't operate through algorithms AND that I never wanted to make a business out of my social media presence.
Just posting whatever I want, whenever I want. And you folks are just there and interact (or not). And for every interaction, I know that it's a genuine one, not "for the algorithm". - And THIS is just SO precious to ME.
It simply feels so honest.
Thank you for being there.

"regardless of the underlying technology, the pursuit of artificial general intelligence is not necessarily the most efficient route to useful applications. Artificial specific intelligence (AI approaches focused on a specific domain, such as the Nobel prize-winning, protein-folding algorithm, AlphaFold2) gives more reliable and transparent results by combining the subtle pattern detection at which GenAI excels with explicitly encoded, domain-specific knowledge."

Memo: the TikTok US deal is set to close on Jan. 22; terms include retraining the recommendation algorithm on US user data and Oracle overseeing data protection (Alex Weprin/The Hollywood Reporter)
https://www.hollywoodreporter.com/business

TikTok U.S. Deal to Close Next Month: Oracle and Silver Lake Confirmed in Buyer Consortium
Abu Dhabi's state investment fund will also be a managing investor.

I need to read it properly, but this looks 🔥 https://arxiv.org/abs/2511.16652

Evolution Strategies at the Hyperscale
We introduce Evolution Guided General Optimization via Low-rank Learning (EGGROLL), an evolution strategies (ES) algorithm designed to scale backprop-free optimization to large population sizes for modern large neural network architectures with billions of parameters. ES is a set of powerful blackbox optimisation methods that can handle non-differentiable or noisy objectives with excellent scaling potential through parallelisation. Na{ï}ve ES becomes prohibitively expensive at scale due to the…

A classic. I've implemented the sweep-merge-interval algorithm in more languages than I can remember. #Gleam
I just completed "Cafeteria" - Day 5 - Advent of Code 2025 #AdventOfCode

They told the poor white man:
"You may be starving,
you may be broke,
but at least you aren't one of Them."
It worked.
The poor whites stopped fighting the rich.
They started guarding the rich.
They accepted their poverty -- because they had been given a false sense of superiority.
The "Divide and Rule" algorithm was born.
350 years later, the campaign is still running.
The Elite are still terrified of Unity…

MikeDunnAuthor (@MikeDunnAuthor@kolektiva.social)
Attached: 1 image Andre Gonsalves The most effective marketing campaign in history was launched in 1676. The product was Racism. Before that year, the Elites in America had a massive problem. In the colony of Virginia, poor white indentured servants and black slaves were friends. They worked the same fields. They lived in the same shacks. They even married each other. They realized they had the same enemy: The rich plantation owners. So, they united. Led by a man named Nathaniel Bacon, this…

An interview with Nikita Bier, who wants to lure journalists back to X with a rebuilt recommendation algorithm with no keyword suppression or manual downranking (Alex Heath/Sources)
https://sources.news/p/x-wants-its-haters-back

X wants its haters back
Product chief Nikita Bier's message to journalists who left: “We have to earn their trust.”

So how do I know which TikTok I’m on when I’m opening the app in Germany— oracle TikTok or bytedance TikTok? Because it feels to me like their algorithm has changed somehow, it’s trying to show me a lot more „controversial“ posts.

Meta launches an AI feature that lets Threads users temporarily personalize their feed by specifying topics in a public post that begins with "Dear Algo" (Jonathan Vanian/CNBC)
https://www.cnbc.com/2026/02/11/meta-threads-dear-algo-ai-algori…

Meta launches AI algorithm personalization feature for Threads
The feature, called "Dear Algo," lets Threads users personalize what content they see by publicly posting an AI prompt.

Instagram launches Your Algorithm, which shows users an AI analysis of their activity to let them customize the topics shaping their Reels recommendations (Reece Rogers/Wired)
https://www.wired.com/story/instagram-lets-you-pick-what-shows-up-in-reels/

Instagram Will Start Letting You Pick What Shows Up in Your Reels
In the battle for your attention, Instagram is betting that more control over the algorithm could keep you scrolling on Reels.

An interview with Nikita Bier, who wants to lure journalists back to X with a rebuilt recommendation algorithm with no keyword suppression or manual downranking (Alex Heath/Sources)
https://sources.news/p/x-wants-its-haters-back

X wants its haters back
Product chief Nikita Bier's message to journalists who left: “We have to earn their trust.”

How Craigslist has stayed relevant for users as a place to find jobs, housing, and personal connections without relying on algorithmic feeds or public profiles (Jennifer Swann/Wired)
https://www.wired.com/story/is-craigslist-the-last-real-place-on-the-internet/…

Is Craigslist the Last Real Place on the Internet?
Millennials are still using Craigslist to find jobs, love, and even to cast creative projects—eschewing other AI- and algorithm-dominated online spaces. “There’s a purity to it.”

Instagram launches Your Algorithm, which shows users an AI analysis of their activity to let them customize the topics shaping their Reels recommendations (Reece Rogers/Wired)
https://www.wired.com/story/instagram-lets-you-pick-what-shows-up-in-reels/

Instagram Will Start Letting You Pick What Shows Up in Your Reels
In the battle for your attention, Instagram is betting that more control over the algorithm could keep you scrolling on Reels.

Meta launches an AI feature that lets Threads users temporarily personalize their feed by specifying topics in a public post that begins with "Dear Algo" (Jonathan Vanian/CNBC)
https://www.cnbc.com/2026/02/11/meta-threads-dear-algo-ai-algori…

How Craigslist has stayed relevant for users as a place to find jobs, housing, and personal connections without relying on algorithmic feeds or public profiles (Jennifer Swann/Wired)
https://www.wired.com/story/is-craigslist-the-last-real-place-on-the-internet/…

Is Craigslist the Last Real Place on the Internet?
Millennials are still using Craigslist to find jobs, love, and even to cast creative projects—eschewing other AI- and algorithm-dominated online spaces. “There’s a purity to it.”

Crosslisted article(s) found for cs.DS. https://arxiv.org/list/cs.DS/new
[1/1]:
- A Fault-Tolerant Version of Safra's Termination Detection Algorithm
Wan Fokkink, Georgios Karlos, Andy Tatman
https://arxiv.org/abs/2602.00272 https://mastoxiv.page/@arXiv_csDC_bot/116005743114928343
- Non-Clashing Teaching in Graphs: Algorithms, Complexity, and Bounds
Sujoy Bhore, Liana Khazaliya, Fionn Mc Inerney
https://arxiv.org/abs/2602.00657 https://mastoxiv.page/@arXiv_csCC_bot/116005533092908045
- Sublinear Time Quantum Algorithm for Attention Approximation
Zhao Song, Jianfei Xue, Jiahao Zhang, Lichen Zhang
https://arxiv.org/abs/2602.00874 https://mastoxiv.page/@arXiv_quantph_bot/116006587128552159
- Hallucination is a Consequence of Space-Optimality: A Rate-Distortion Theorem for Membership Testing
Anxin Guo, Jingwei Li
https://arxiv.org/abs/2602.00906 https://mastoxiv.page/@arXiv_csLG_bot/116006973804501595
- Counting Unit Circular Arc Intersections
Haitao Wang
https://arxiv.org/abs/2602.01074 https://mastoxiv.page/@arXiv_csCG_bot/116005535191670040
- Profit Maximization in Closed Social Networks
Poonam Sharma, Suman Banerjee
https://arxiv.org/abs/2602.01232 https://mastoxiv.page/@arXiv_csSI_bot/116005691056225955
- Totally $\Delta$-Modular Tree Decompositions of Graphic Matrices for Integer Programming
Caleb McFarland
https://arxiv.org/abs/2602.01499 https://mastoxiv.page/@arXiv_mathCO_bot/116006381339684193
- Finite and Corruption-Robust Regret Bounds in Online Inverse Linear Optimization under M-Convex A...
Taihei Oki, Shinsaku Sakaue
https://arxiv.org/abs/2602.01682 https://mastoxiv.page/@arXiv_csLG_bot/116007186801693076
- Stable Matching with Predictions: Robustness and Efficiency under Pruned Preferences
Samuel McCauley, Benjamin Moseley, Helia Niaparast, Shikha Singh
https://arxiv.org/abs/2602.02254 https://mastoxiv.page/@arXiv_csGT_bot/116005915605630934
- Deciding Reachability and the Covering Problem with Diagnostics for Sound Acyclic Free-Choice Wor...
Thomas M. Prinz, Christopher T. Schwanen, Wil M. P. van der Aalst
https://arxiv.org/abs/2602.02447 https://mastoxiv.page/@arXiv_csFL_bot/116005733006023408
toXiv_bot_toot

More than 20% of the videos that YouTube’s algorithm shows to new users are “AI slop”
– low-quality AI-generated content designed to farm views, research has found.
The video-editing company Kapwing surveyed 15,000 of the world’s most popular YouTube channels
– the top 100 in every country
– and found that 278 of them contain only AI slop.
Together, these AI slop channels have amassed more than 63 billion views and 221 million subscribers,
-- generating ab…

The EU says that TikTok's "addictive design" is illegal under the DSA, citing the app's infinite scroll and recommendation algorithm, in preliminary findings (Adam Satariano/New York Times)
https://www.nytimes.com/2026/02/06/business/tiktok-a…

TikTok says the new US entity has not updated its algorithm since the joint venture was announced, after users accused it of suppressing posts related to ICE (David McCabe/New York Times)
https://www.nytimes.com/2026/01/27/us/tiktok-ice-minneapolis-techni…

TikTok Blames Technical Problems After Users Claimed It Blocked ICE Posts
Some users had accused the app of blocking them from posting videos about Immigration and Customs Enforcement. The app said it was a power outage issue.

TikTok says the new US entity has not updated its algorithm since the joint venture was announced, after users accused it of suppressing posts related to ICE (David McCabe/New York Times)
https://www.nytimes.com/2026/01/27/us/tiktok-ice-minneapolis-techni…

TikTok Blames Technical Problems After Users Claimed It Blocked ICE Posts
Some users had accused the app of blocking them from posting videos about Immigration and Customs Enforcement. The app said it was a power outage issue.

Welfarist Formulations for Diverse Similarity Search
Siddharth Barman, Nirjhar Das, Shivam Gupta, Kirankumar Shiragur
https://arxiv.org/abs/2602.08742 https://arxiv.org/pdf/2602.08742 https://arxiv.org/html/2602.08742
arXiv:2602.08742v1 Announce Type: new
Abstract: Nearest Neighbor Search (NNS) is a fundamental problem in data structures with wide-ranging applications, such as web search, recommendation systems, and, more recently, retrieval-augmented generations (RAG). In such recent applications, in addition to the relevance (similarity) of the returned neighbors, diversity among the neighbors is a central requirement. In this paper, we develop principled welfare-based formulations in NNS for realizing diversity across attributes. Our formulations are based on welfare functions -- from mathematical economics -- that satisfy central diversity (fairness) and relevance (economic efficiency) axioms. With a particular focus on Nash social welfare, we note that our welfare-based formulations provide objective functions that adaptively balance relevance and diversity in a query-dependent manner. Notably, such a balance was not present in the prior constraint-based approach, which forced a fixed level of diversity and optimized for relevance. In addition, our formulation provides a parametric way to control the trade-off between relevance and diversity, providing practitioners with flexibility to tailor search results to task-specific requirements. We develop efficient nearest neighbor algorithms with provable guarantees for the welfare-based objectives. Notably, our algorithm can be applied on top of any standard ANN method (i.e., use standard ANN method as a subroutine) to efficiently find neighbors that approximately maximize our welfare-based objectives. Experimental results demonstrate that our approach is practical and substantially improves diversity while maintaining high relevance of the retrieved neighbors.
toXiv_bot_toot