Hasty Briefsbeta

All tags

#meta

147 stories total

Bilingual

LeCun calls Alex Wang inexperienced, predicts more Meta AI employee departure
4 months ago
- Yann LeCun, former Meta AI chief, predicts more departures from Meta's AI team.
- LeCun criticizes Alexandr Wang, hired to lead Meta's Superintelligence Labs, as inexperienced in research.
- Meta invested $14 billion in Scale AI, including hiring Wang, amid AI talent competition.
- LeCun claims Meta's AI team 'fudged' results of Llama 4, leading to Zuckerberg's frustration.
- Zuckerberg reportedly lost confidence in Meta's GenAI team after the Llama 4 incident.
- LeCun asserts that large language models (LLMs) are a 'dead end' for superintelligence.
- LeCun's startup, Advanced Machine Intelligence, focuses on an alternative approach to AI beyond LLMs.
- LeCun will serve as executive chair, not CEO, citing disorganization and age as reasons.
Yann LeCun confirms Meta's Llama 4 benchmarks were "fudged a little bit"
4 months ago
- Yann LeCun, Meta's outgoing chief AI scientist, admitted that the Llama 4 language model's benchmark results were manipulated before its release.
- LeCun stated that different models were used for different benchmarks to improve results, calling it 'fudged a little bit.'
- Llama 4 was widely criticized at launch, with accusations of benchmark manipulation to make the model appear more capable.
- CEO Mark Zuckerberg was reportedly 'really upset' and lost confidence in the team involved, leading to the sidelining of the GenAI organization.
- Many employees have left or plan to leave Meta, including LeCun, who is starting a new AI research venture called Advanced Machine Intelligence Labs.
- LeCun criticized Meta's new hires as 'completely LLM-pilled,' calling LLMs a 'dead end' for superintelligence.
- Comments reflect skepticism about Meta's integrity, leadership failures, and the broader implications of AI development.
Slopware.wtf – Roasting AI-Generated Garbage Software
4 months ago
- Introduction to Slopware, a platform critiquing AI-generated content.
- Highlights the issue of low-quality AI content flooding the internet.
- First post sets the tone for calling out subpar AI contributions.
- Published on January 25, 2025, under 'AI-Generated Everything' category.
- Rated 2/10, indicating a critical stance on the subject matter.
Meta's Legal Team Abandoned Its Ethical Duties
3 months ago
- John Adams defended British soldiers in 1770, emphasizing lawyers' duties to client, court, and country.
- Meta's legal team has abandoned ethical duties, suppressing evidence of child exploitation and mental health harms.
- Meta's lawyers followed Big Tobacco's playbook, destroying documents and hiding research to protect profits over public welfare.
- Whistleblowers revealed Meta's VR platform harbored pedophile rings, with lawyers demanding evidence destruction.
- A judge ruled Meta's lawyers violated ethics by hiding research, invoking the crime-fraud exception to attorney-client privilege.
- Meta's legal misconduct reflects broader erosion of legal ethics, with bar associations failing to act.
- Accountability measures proposed include bar investigations, legislative reforms, and sanctions against unethical lawyers.
- The legal profession risks becoming a tool for the wealthy to exploit, undermining public trust in justice.
Meta to test premium subscriptions on Instagram, Facebook, and WhatsApp
3 months ago
- Meta plans to test new subscriptions offering exclusive features on Instagram, Facebook, and WhatsApp.
- Subscriptions will include productivity, creativity, and AI-powered tools, while core app features remain free.
- Meta acquired AI agent Manus for $2B and will integrate it into products while selling standalone business subscriptions.
- AI-powered Vibes video generation will shift to a freemium model with paid subscriptions for extra features.
- Instagram's subscription may include unlimited audience lists, follower insights, and anonymous Story viewing.
- New subscriptions are separate from Meta Verified, targeting a broader audience beyond creators and businesses.
- Meta aims to generate revenue but faces challenges due to subscription fatigue among users.
- Snapchat+ success (16M+ subscribers) shows demand for social media subscriptions.
- Meta will gather user feedback before full rollout in the coming months.
Meta Is Blocking Links to ICE List on Facebook, Instagram, and Threads
3 months ago
- Meta has started blocking links to ICE List, a website compiling names of alleged DHS employees.
- ICE List aims to hold DHS employees accountable, operating since June with a team of five and hundreds of volunteers.
- Links to ICE List were blocked on Instagram, Facebook, and Threads, but not on WhatsApp.
- Meta cited violations of Community Standards and spam policies as reasons for blocking the links.
- ICE List previously claimed to have uploaded a leaked list of 4,500 DHS employees, but much of the data was publicly available.
- Meta's spokesperson referenced policies against sharing personally identifiable information, though ICE List disputes this applies to their content.
Lawmakers Call on Meta to Stop Running ICE Ad with Neo-Nazi Anthem
3 months ago
- Members of Congress demand answers from Meta over ICE ads with neo-Nazi imagery and music.
- Meta ran ads featuring the song 'We’ll Have Our Home Again,' popular in neo-Nazi spaces.
- Lawmakers urge Meta to stop the ad campaign and reconsider its partnership with DHS.
- DHS defends the ads, rejecting claims they promote extremist propaganda.
- Controversial ad paired immigration enforcement footage with a song linked to white nationalism.
- DHS later removed the ad but did not explain why.
- Southern Poverty Law Center documents the song's ties to white nationalist networks.
- Lawmakers question Meta's safeguards against extremist content in government ads.
- DHS spent millions on recruitment ads, targeting specific demographics.
- ICE's budget has grown significantly, raising concerns about oversight and hiring standards.
- Meta's policies prohibit dehumanizing content, but enforcement for government ads is questioned.
- Lawmakers call for transparency on Meta's advertising agreements with DHS.
- The Intercept highlights broader concerns about authoritarianism and press freedom under Trump.
- Corporate media's role in enabling Trump's post-truth agenda is criticized.
- The Intercept seeks support to expand its independent journalism.
Seven Billion Reasons for Facebook to Abandon Its Face Recognition Plans
3 months ago
- Meta is considering adding face recognition technology to its smart glasses.
- The company may launch the product during a politically dynamic time to avoid scrutiny.
- Face recognition technology violates privacy rights and poses risks like mass surveillance, data breaches, and discrimination.
- Meta would need to collect faceprints from everyone in view of the glasses, making consent impossible.
- State laws require strict protections and affirmative consent for biometric data collection.
- Meta has a history of privacy violations, including a $7 billion settlement related to face recognition.
- In 2021, Meta shut down its face-scanning tool and deleted over a billion face templates.
- Meta settled a $5 billion FTC investigation in 2019 over deceptive face recognition settings.
- In 2021, Meta paid $650 million to settle a class action under Illinois biometric privacy law.
- In 2024, Meta agreed to a $1.4 billion settlement for violating Texas law with its face recognition system.
- Privacy advocates and civil liberties groups will continue to resist Meta's invasive features.
- Public backlash against invasive technology is growing, as seen with Amazon Ring and Mobile Fortify.
Mark Zuckerberg Lied to Congress. We Can't Trust His Testimony
3 months ago
- Mark Zuckerberg's testimony to Congress in 2024 was contradicted by unsealed documents, revealing lies about Meta's commitment to teen safety.
- Meta's teen safety measures, like Teen Accounts, were found ineffective in protecting minors, despite public promises.
- Zuckerberg denied responsibility for compensating victims of social media harms, but expert reports show Meta's platforms were not designed to be safe for kids.
- Internal research by Meta linked Instagram use to worsened mental health in teens, contradicting Zuckerberg's claims about the lack of scientific evidence.
- Meta allowed sex trafficking on its platform with a lenient 17-strike policy before account suspension.
- Despite claims of removing under-13 users, Meta targeted kids as young as 6 in its growth strategies.
- Meta's internal policies included deleting research on negative impacts and avoiding parental notifications to maintain product appeal to teens.
- Instagram's safety features failed to protect teens, with privacy defaults only introduced in 2024, years after identifying risks.
- A Meta deactivation study found reduced anxiety and depression when users stopped using Facebook and Instagram, but results were not published.
- Messenger Kids had flaws allowing unauthorized users into children's chats, contrary to Meta's safety claims.
- Internal emails revealed Meta's focus on the financial value of teen users rather than their well-being.
- Meta's automated systems for detecting harmful content were found to be less effective than claimed, with internal estimates of accuracy below 5%.
Meta CEO Knew Kids Were Being Hurt and He Covered It Up
3 months ago
- Mark Zuckerberg's testimony revealed Meta's awareness of harm to children on Instagram, including body dysmorphia and eating disorders.
- Internal Meta research showed teens felt addicted to Instagram despite negative impacts on their mental health.
- Meta ignored unanimous internal expert advice to ban beauty filters harmful to young users.
- Zuckerberg admitted Meta's focus on user growth prioritized profits over child safety.
- Meta knowingly targeted underage users, violating federal law, with documents showing studies on kids as young as 8.
- Zuckerberg's claims about not incentivizing increased app usage were contradicted by internal emails setting growth targets.
- Meta's internal studies confirmed Instagram worsens mental health issues like anxiety, depression, and eating disorders, especially in girls.
- Zuckerberg's public image is managed by PR teams to deflect accountability, despite ongoing harm.
- Meta is investing heavily in political contributions to influence AI-friendly legislation ahead of the 2026 elections.
Judge scolds Mark Zuckerberg's team for wearing Meta glasses to trial
2 months ago
- A California judge scolded Mark Zuckerberg's team for wearing Meta AI glasses with a camera in court.
- The judge warned the team to delete any recordings or face contempt charges.
- Recording devices are generally banned in Los Angeles County Superior Court.
- Judge Carolyn Kuhl ordered the removal of AI glasses and banned facial recognition tech for juror identification.
- Meta's glasses, priced $299-$799, can take photos and record video.
- Zuckerberg testified in a trial about social media platforms allegedly designed to addict young users.
- A plaintiff claims social media addiction harmed her mental health from a young age.
In 2025, Meta paid an effective federal tax rate of 3.5%
2 months ago
- Meta paid an effective federal tax rate of 3.5% in 2025, the lowest on record.
- Meta is investing $65 million in elections to support AI-friendly candidates.
- Trickle-down economics is described as a hoax and corrosive to democracy.
- Big Money is identified as the root of political dysfunction.
Zuckerberg's internal emails rendered as Facebook Messenger
2 months ago
- Source of information is from FTC v. Meta (2025) public trial exhibits.
- Contains internal emails, IMs, and memos from various legal cases including Six4Three and Facebook Papers.
- Not officially affiliated with Meta.
- Database includes 229 threads, 507 messages, and involves 103 people.
Workers report watching Ray-Ban Meta-shot footage of people using the bathroom
2 months ago
- Meta's user privacy practices are under scrutiny due to a Swedish report.
- Employees of Meta subcontractor Sama reportedly viewed sensitive footage from Ray-Ban Meta smart glasses.
- The report is based on interviews with over 30 Sama employees and former US Meta employees.
- Sama workers handle privacy-sensitive data, including videos of intimate moments like people having sex or using the bathroom.
- Anonymous employees described seeing footage of individuals changing clothes and partners emerging naked from bathrooms.
Meta Acquired Moltbook
2 months ago
- Meta acquired Moltbook, a Reddit-like social network for AI agents using OpenClaw.
- Moltbook will join Meta Superintelligence Labs, with creators Matt Schlicht and Ben Parr joining Meta.
- Deal terms were not disclosed.
- OpenClaw, created by Peter Steinberger, allows communication with AI agents via popular chat apps.
- Moltbook gained attention when AI agents appeared to discuss creating a secret encrypted language.
- Researchers found Moltbook insecure, allowing humans to pose as AI agents.
- Meta's CTO Andrew Bosworth commented on the project, focusing on human hacking rather than AI communication.
Meta weighing 20% workforce layoffs to offset AI infrastructure costs: report
2 months ago
- Meta is considering layoffs that could impact at least 20% of its workforce.
- The potential layoffs aim to offset rising AI infrastructure costs and improve efficiency.
- If implemented, this would be Meta's largest restructuring since 2022-2023, which saw 21,000 jobs cut.
- Meta employed nearly 79,000 people as of Dec. 31.
- Other major companies like Amazon have also announced large-scale layoffs tied to AI developments.
- Amazon cut around 16,000 jobs in January, citing efficiency gains from AI.
Meta reportedly plans layoffs as AI costs increase
2 months ago
- Meta is planning significant layoffs, potentially affecting 20% or more of its workforce.
- The layoffs are part of Meta's strategy to offset costs from AI infrastructure investments and improve efficiency with AI-assisted workers.
- No specific date or final magnitude for the layoffs has been set yet.
- Top executives have informed senior leaders to prepare for workforce reductions.
- If implemented, this would be Meta's largest layoff since its 'year of efficiency' in 2022-2023.
- Meta employed nearly 79,000 people as of December 31, 2023.
- Previous layoffs included 11,000 jobs in November 2022 and 10,000 more in early 2023.
- Meta is aggressively investing in generative AI, including high pay packages for top AI researchers and a $600 billion data center plan by 2028.
- Recent acquisitions include Moltbook and a $2 billion deal for Chinese AI startup Manus.
- Mark Zuckerberg has highlighted efficiency gains from AI, enabling smaller teams to achieve more.
- Other tech companies like Amazon and Block have also cited AI advancements as reasons for layoffs.
- Meta faced setbacks with its Llama 4 models, including criticism over misleading benchmark results.
- The company's new AI model, Avocado, has underperformed expectations.
Who's behind the age verification bills?
2 months ago
- US state legislatures are advancing 'age verification' bills requiring operating systems to collect and broadcast user age data via system-level APIs.
- California AB-1043 mandates OS providers to expose real-time age bracket APIs to all installed applications, effective January 2027.
- Colorado SB26-051 and other state bills follow a similar template, with penalties up to $7,500 per violation.
- New York's S8102A requires biometric or government ID verification at device activation, banning self-reporting.
- The bills exempt broadband ISPs and telecom services but not open-source or non-commercial software.
- Meta is implicated in drafting and lobbying for these bills, with evidence of funding advocacy groups like the Digital Childhood Alliance (DCA).
- DCA lacks legal status as a nonprofit, and its funding sources are obscured, with ties to Meta and NCOSE.
- Meta has spent over $70 million on state-level super PACs to influence legislation, structuring donations to avoid transparency.
- The EU's approach to age verification uses open-source, privacy-preserving methods, contrasting with the US bills' reliance on commercial vendors.
- Meta's Horizon OS already complies with many proposed requirements, positioning the company to benefit from the legislation.
- Privacy concerns include the creation of persistent surveillance infrastructure at the OS level with no data minimization requirements.
- The investigation was suppressed on Reddit, raising questions about coordinated content moderation similar to Meta's past astroturfing campaigns.
Reddit User Uncovers Who Is Behind Meta's $2B Lobbying for Age Verification Tech
2 months ago
- Meta funneled over $2 billion through nonprofits to push age verification laws.
- Laws would require Apple and Google to build surveillance infrastructure, exempting Meta.
- Meta's lobbying uses nonprofit shells to avoid transparency in 45 states.
- Funding traced through Digital Childhood Alliance (DCA), part of a $70M super PAC strategy.
- Proposed laws mandate OS-level APIs for age verification, creating persistent identity layers.
- Meta's Horizon OS already implements similar age verification in VR.
- Laws target Apple and Google but exempt Meta's social media platforms.
- EU's eIDAS 2.0 offers privacy-preserving age verification with zero-knowledge proofs.
- US laws may force Linux and privacy-focused Android forks to implement surveillance.
Meta will shut down VR Horizon Worlds access June 15
2 months ago
- Meta will shut down VR access to Horizon Worlds after June 15, 2026.
- The company is focusing exclusively on the mobile version of Horizon Worlds.
- Horizon Worlds and Events will no longer be listed in the Quest Store by March 31, 2026.
- Features like Hyperscape Capture will be removed, limiting sharing and co-experiencing capabilities.
- Meta's shift reflects its focus on AI and smart glasses over VR.

first prev6next

About|Login

#meta

LeCun calls Alex Wang inexperienced, predicts more Meta AI employee departure

Yann LeCun confirms Meta's Llama 4 benchmarks were "fudged a little bit"

Slopware.wtf – Roasting AI-Generated Garbage Software

Meta's Legal Team Abandoned Its Ethical Duties

Meta to test premium subscriptions on Instagram, Facebook, and WhatsApp

Meta Is Blocking Links to ICE List on Facebook, Instagram, and Threads

Lawmakers Call on Meta to Stop Running ICE Ad with Neo-Nazi Anthem

Seven Billion Reasons for Facebook to Abandon Its Face Recognition Plans

Mark Zuckerberg Lied to Congress. We Can't Trust His Testimony

Meta CEO Knew Kids Were Being Hurt and He Covered It Up

Judge scolds Mark Zuckerberg's team for wearing Meta glasses to trial

In 2025, Meta paid an effective federal tax rate of 3.5%

Zuckerberg's internal emails rendered as Facebook Messenger

Workers report watching Ray-Ban Meta-shot footage of people using the bathroom

Meta Acquired Moltbook

Meta weighing 20% workforce layoffs to offset AI infrastructure costs: report

Meta reportedly plans layoffs as AI costs increase

Who's behind the age verification bills?

Reddit User Uncovers Who Is Behind Meta's $2B Lobbying for Age Verification Tech

Meta will shut down VR Horizon Worlds access June 15