Social Media & Big Tech news and discussions

[firestar464]

[wjfox]

‘They’re thinking about the likes’: deadly TikTok motorway trend horrifies Ireland

Social media platforms face public ire after fatal crash involving teenagers filming themselves driving on wrong side

Sat 22 Aug 2026 06.00 BST

Videos often begin with the camera panning across a glowing dashboard as a gloved hand starts the ignition. The camera turns to show figures with masked faces in the rear seat, who are also filming, and swivels back to show the windscreen. Beyond it, a dark cityscape.

The engine throbs, headlights puncture the night and the journey begins, first down a suburban road, then a motorway. The speedometer shows acceleration, the needle juddering as the car gathers velocity.

The destination is irrelevant, for this is a show – an exhibition of audacity, lawlessness and recklessness that draws viewers to TikTok and other social media platforms with the possibility of chases by police and dramatic crashes.

Such videos are part of a macabre trend that has astonished and outraged Ireland in the wake of a collision last week that killed five teenagers – who were driving the wrong way down a motorway – and critically injured a family of four, prompting debate about policing, criminal justice, social services and social media regulation.

“You have to ask: how do five young men gravitate to each other and decide this is a good idea? Where have we gone wrong as a society?” said Graham Kavanagh, a former officer with the police force An Garda Síochána, who used to pursue car thieves.

https://www.theguardian.com/media/2026/ ... es-ireland



Photograph: Niall Carson/PA

[firestar464]

TikTok reaches $400M settlement with DOJ over alleged children’s privacy law violations

https://www.msn.com/en-us/money/general ... r-AA2aGaZT

[firestar464]

Teen drops high-profile social media addiction lawsuit

https://www.msn.com/en-us/money/general ... r-AA2aGO8H

[weatheriscool]

Teen drops high-profile social media addiction lawsuit

https://www.msn.com/en-us/money/general ... r-AA2aGO8H

What's the point? They choose to use it and millions of other people enjoy using it. Including me. I use it to keep in touch with my family and to find different information. So the solution is to take it away from us all? This is bullshit. Anyone that supports this is a fascist f*ck. This reminds me of the detransition movement where assholes bitch about how being trans was a mistake and we shouldn't have the choice to be trans. It is an argument that blocks others from making their own choice. Same shit.

One can simply log off and do something else if they don't like it. Why tell me or millions of other people what we can enjoy in our lives?

[firestar464]

"one can simply stop taking drugs"

also imagine comparing life-saving gender affirming care to freaking Zuckbook. just stop using trans people as a shield for your corporate whiteknighting already

[firestar464]

Cyberbullying may fuel teen mental health problems through loneliness, study finds

https://medicalxpress.com/news/2026-08- ... ealth.html

[firestar464]

https://techcrunch.com/2026/08/25/x-sen ... -scraping/

Lawyers for X said the actions are in violation of “various state and federal laws, including, but not limited to, the Texas Harmful Access by Computer Act (§ 143.001 and § 33.02) and the Lanham Act (15 U.S.C. §§ 1114, 1125).” The letter gave Nitter until 5 p.m. EST on August 25 to shut down.

Not sure how XCancel and Nitter would make people confuse them with X? Makes no sense

[caltrek]

Meta Doesn’t Want to Be a Social Media Company Anymore
By Anna Merlan
August 26, 2026

Introduction:

(Mother Jones) A nationwide social media addiction trial against Meta came to an abrupt end on Wednesday, with the company announcing a sweeping settlement with 52 attorneys general. As part of the proposed settlement, the company promised to pay around $18 billion to participating states over a ten-year period and enact a set of changes to make their apps, like Facebook and Instagram, supposedly safer for teenagers, including by enacting a default two-hour limit for teenage users and setting up “night” and “school” modes to block the apps or reduce notifications during certain times. In their press release, the company also pointedly urged rival companies like YouTube and TikTok to do the same.

“While this is an important step, the fact is that teens move fluidly between dozens of apps a day,” the announcement read. “All platforms should empower parents and support teens by putting the same measures in place, because we know that when teens are restricted on one app, they simply move to another.” (Though it was somewhat buried in the announcement, Meta also stipulated that it would only commit to enforcing its time limits and “night” modes for five years, but said it would extend that commitment to ten years “if industry peers sign on to the agreement.”)

All of this, of course, builds on a series of recent losses for Meta: the company also lost two phases of a similar trial in New Mexico, with the jury finding that the company misled consumers about the safety of its platforms and intentionally built its products to addict minors. The state also claimed that “Meta’s design features enabled pedophiles and predators to engage in child sexual exploitation on Meta’s platforms,” a claim the jury also agreed with.

Conclusion:

Like virtually every other tech company, then, Meta executives have turned from social media as their core product, to a gauzy, pie-eyed future of AI as the future of work, business and everything else. Laws, legislation and jury settlements are often backwards-looking when it comes to tech companies. This particular settlement attempts to mitigate the harm of their current products, while companies like Meta are already focused, for better or much more likely worse, on what’s next.

Read more here: https://www.motherjones.com/politics/2 ... asses/

[firestar464]

Is Twitter back? Ex-employee fights to revive the blue bird

https://www.euronews.com/next/2026/08/2 ... -blue-bird

[firestar464]

Brain scans reveal how personalized social media content may amplify negative moods

https://medicalxpress.com/news/2026-08- ... ocial.html

This goes back to the age-old issue of companies optimizing for engagement while not doing enough to protect vulnerable users, and how they need to be regulated into not doing that. The EU DSA is a good start, though enforcement is nowhere near what it should be just yet, because of legal caution on the part of regulators

[firestar464]

Me: What are some solutions so that algorithms don't worsen the mental health of distressed users?

Claude 5 Opus: Split them by whether they need to know what content is about, because that determines both the failure modes and the legal exposure.

Classifier-free (no content judgment, so no overblocking, no First Amendment content-discrimination problem):

Decay the interest signals. Right now a bad fortnight permanently reshapes a feed — engagement signals accumulate and don't expire. Exponential decay on learned interests means a depressive episode stops being a durable rewrite of what you see. This is close to the mechanism the paper actually describes, it's trivially implementable, and it needs no idea what the content is.

Lengthen the optimization horizon. Predicting next-session watch time is what generates the pathological short loop; predicting 28-day retention or next-week satisfaction rewards different things. Platforms already do versions of this internally.

Cap topical concentration per session and enforce an exploration floor outside the learned cluster. Both attack specialization directly. Neither requires deciding a topic is bad — just that too much of one thing is a bad shape for a feed regardless.

Classifier-dependent, [which has] precision problems:

A floor on positive-valence content rather than a ceiling on negative, since absence of positive is the variable that actually correlated with anything. Gameable, and only as good as the classifier.

Bridging-style ranking, and here's the honest bit: Stray et al. ran it as an arm of a large independent field experiment across Facebook, X and Reddit in 2026, using GPT-4o to predict posts both Democrats and Republicans would find valuable. It produced no significant reduction in affective polarization and significantly lowered engagement. The most-hyped alternative objective returned a null on the outcome and a confirmed cost. Different outcome variable than yours, but it should temper expectations.

On the identification paradox: user-declared state is the partial escape. Let someone say "dial it down for a while" and have that actually change ranking. The platform doesn't infer distress, the user reports it, and auditing happens at population level rather than per-user classification.

The constraint underneath all of it: independent researchers running reranking experiments can only rerank what the platform already prefiltered, which is mostly high-engagement content. The dynamic range is truncated before the experiment starts, so external work systematically underestimates what a different objective would do. You cannot answer this question from outside the platform. That's the strongest argument for mandated data and inventory access, and it's why Article 40 [of the DSA] keeps turning out to be the load-bearing provision.

And every intervention here costs engagement — Stray measured that directly. So which one gets adopted isn't a design question, it's a question about what makes a firm ship a revenue-negative change.

[caltrek]

The Liberal Currents article I have linked below is somewhat longer than the typical citation that I make. It is also full of a lot of interesting points concerning elites and social media and it would not do it justice to quote out of context. So, I will just leave the link to that article:

https://www.liberalcurrents.com/elites- ... ial-media/

[weatheriscool]

People need to start fighting for social media as our public sphere the same as the street corner. Because it fucking is.

People that want to close it off are aiming to rob us of the ability to talk freely online and it is evil and needs stop!!! A counter suite based on the first amendment needs to be filed against all these fascist fuckers. And a constitutional amendment clearly stating social media and the open internet as protected by the first needs to be passed by the people.

an updated first amendment needs to state that social media is also part of the street corner. People should be able to speak to one another and to protest within its confines.

[firestar464]

The Liberal Currents article I have linked below is somewhat longer than the typical citation that I make. It is also full of a lot of interesting points concerning elites and social media and it would not do it justice to quote out of context. So, I will just leave the link to that article:

https://www.liberalcurrents.com/elites- ... ial-media/

TLDR: elites and people with power and influence need to touch grass and stop acting like social media is representative of society or real life

(Also Mamdani based and not mentally fried, though I still believe he should leave X)

[weatheriscool]