Eyerys - Articles

OpenAI Is Hiring For The AI Singularity, Paying $445,000 To Solve Problems That Might Not Yet Exist

Danny Krisnanto — Sun, 24 May 2026 04:21:41 +0000

24/05/2026

OpenAI has posted a striking job opening in its Preparedness safety team in San Francisco area focused on the emerging risks of recursive self-improvement in artificial intelligence systems.

The researcher role titled "Recursive Self-Improvement Preparedness" involves preparing for scenarios in which an AI could autonomously research, design, and train improved versions of itself with minimal human involvement, a concept that has shifted from distant theory to a pressing industry priority amid rapid advances in coding and reasoning tools.

Compensation for the position reaches as high as US$445,000, underscoring the high stakes and specialized judgment required.

Read: Paving The Roads To Artificial Intelligence: It's Either Us, Or Them

Candidates will tackle tasks such as protecting models from data poisoning, building tools to interpret complex reasoning processes inside AI systems, running experiments to assess safety implications of self-improving architectures, and tracking how quickly AI tools are automating technical work inside the company itself.

The job description openly acknowledges that much of the work centers on problems that might not yet exist but could arise soon, which is why it stresses the need for candidates who are especially tasteful and strategic in their thinking.

OpenAI CEO Sam Altman has shared ambitious internal targets, including operating an automated AI research intern across hundreds of thousands of chips by September 2026 and achieving a fully capable automated AI researcher by March 2028.

While he notes these goals could fail, the transparency reflects a recognition of the extraordinary societal impacts that could follow if such systems become reality. This recruitment aligns with broader efforts across the Preparedness team to evaluate frontier risks, including cybersecurity threats, biological concerns, and the behavior of increasingly agentic AI models.

The timing of this hiring highlights how quickly AI technology is evolving, often in ways that surprise even the researchers and engineers who build it.

Tools that started as helpful assistants for writing code or summarizing text have developed unexpected emergent abilities, demonstrating creativity, long-term planning, and problem-solving skills that outpace initial expectations.

Progress in the field has accelerated dramatically, with the length of complex tasks that frontier models can complete reliably doubling roughly every seven months according to independent evaluations.

These surprises stem from the sheer scale of training data, computational power, and architectural refinements that allow models to generalize in novel situations, sometimes revealing capabilities their creators had not explicitly programmed or anticipated.

OpenAI headquarters located in the Mission Bay neighborhood of San Francisco.

This fast-moving landscape connects directly to deeper concepts such as artificial general intelligence, commonly known as AGI.

AGI describes systems capable of understanding, learning, and performing any intellectual task that a human can, across diverse domains rather than being limited to narrow applications. Unlike today's specialized AI, an AGI would adapt fluidly to new challenges, potentially accelerating scientific discovery, engineering, and innovation at unprecedented rates.

Many in the field view the path toward AGI as intertwined with the idea of the technological singularity, a hypothetical future point where AI systems begin recursively self-improving. In such a scenario, each generation of AI would rapidly enhance its own intelligence, leading to an intelligence explosion that could produce technological change so swift and profound it becomes difficult for humans to comprehend or direct.

OpenAI's new role is explicitly tied to managing the transition risks around these possibilities.

By focusing on defenses against uncontrolled self-improvement and by studying how models might automate further AI development, the company seeks to maintain oversight even as capabilities advance.

Industry voices echo this urgency. Google DeepMind CEO Demis Hassabis has referred to the current era as the foothills of the singularity, while estimates from researchers suggest a substantial chance that AI could handle most research and development tasks without human involvement by the end of 2028.

Thoughtful preparation today, through roles like this one, aims to ensure that the immense benefits of advanced AI arrive without catastrophic downsides, allowing humanity to guide progress responsibly as the boundary between current tools and future superintelligent systems continues to blur.

In an age where AI already surprises its creators on a regular basis, investing in strategic foresight feels less like speculation and more like essential prudence.

Further reading: 'I Think We've Achieved AGI,' And Why The Benchmark Keeps Shifting

Sam Altman, CEO of OpenAI, during BlackRock's 2026 Infrastructure Summit in Washington, D.C., on March 11, 2026.

It's worth noting that the OpenAI's Preparedness team (part of the broader Safety Systems organization) currently has multiple active job listings focused on frontier AI risks. The exact category that includes AGI-level threats, recursive self-improvement, agentic behavior, and other catastrophic potential harms. They include:

Researcher, Recursive Self-Improvement Preparedness.
Data Scientist, Preparedness to focus on building and evaluating mitigations to prevent extreme harms from advanced AI systems.
Threat Modeler, Preparedness, owns holistic threat modeling and forecasting for frontier risks across misuse areas (bio, cyber, attack planning, etc.).
Researcher, Automated Red Teaming who– builds scalable systems to automatically discover failure modes and jailbreaks in models, with emphasis on catastrophic risks.
Researcher, Frontier Cybersecurity Risks and similar roles covering biological/chemical risks.

These positions all sit under the same Preparedness framework, which is OpenAI's official system for tracking and preparing for dangerous capabilities in frontier models, the kind that could lead toward AGI and self-improving systems.

The recursive self-improvement researcher role stands out because it's the most directly focused on the "AI building better AI" scenario, but it's part of a larger hiring push across the team to get ahead of these exact risks.

With 'Aleph 2.0,' Runway Pushes AI Video Editing Beyond Generation Into Precision Control

Aditya — Sat, 23 May 2026 10:05:31 +0000

The AI industry has transformed the advancement of the tech industry at an extraordinary pace, and the momentum is only accelerating.

Everything changed after OpenAI introduced ChatGPT in 2022. What began as a breakthrough product quickly became a worldwide phenomenon, setting off an intense race among tech companies to develop increasingly advanced large language models.

The competition rapidly expanded beyond text, ushering in a new era of AI systems capable of generating highly realistic images and multimodal experiences.

Yet despite these advances, video generation remained the most difficult challenge of all, since it's a format far more complex, dynamic, and demanding than text or static imagery.

Runway positioned itself early as a specialist in this space, building tools that let creators generate and especially edit footage with AI rather than starting from scratch each time.

The original Aleph model was representing the company's focus on practical editing capabilities that bridge traditional video production and generative technology. Now, with 'Aleph 2.0,' Runway brings a targeted improvement to that approach.

Aleph 2.0 is here. Now you can edit a single frame in your video, preview the change and then Aleph 2.0 carries that edit across the rest of your video.

Try it now in the new Edit Studio on web at the link below. pic.twitter.com/LA8NgAMnZA

— Runway (@runwayml) May 21, 2026

Aleph 2.0 introduces a new Edit Studio interface, where users can select any single frame from a video clip, make changes to it using prompts or direct adjustments, preview the result immediately, and then allow the model to propagate the edit consistently across the entire sequence.

The system preserves the original motion, lighting, camera movement, and details that were not intended to change.

Clips up to 30 seconds long at 1080p resolution are supported, making the feature usable for short scenes or segments within longer projects.

The demonstration video highlights how this works in practice.

One sequence begins with a close-up of a woman holding a small octopus in her hand on a sunlit street; subsequent examples shift to stylized cyberpunk characters wielding colorful water guns in neon-lit alleys, dramatic landscapes featuring a mother and child walking through a stormy field near an isolated house, and a tactical scene with a woman standing beside a vehicle.

Another shows a pirate ship tossed on rough seas beneath a skull-shaped mountain under lightning.

In the interface itself, a prompt modifies the ground around a small white truck in a parking lot, creating a caved-in appearance everywhere except for a small circular patch supporting the vehicle, with the model applying the change smoothly throughout the camera movement.

These cases illustrate the flexibility for altering objects, characters, environments, or entire visual styles without manual frame-by-frame fixes.

For creators working with a mix of real and generated footage, this level of control addresses a longstanding pain point.

Earlier AI video tools often required regenerating entire clips to fix small issues or risked breaking continuity when making targeted adjustments. Aleph 2.0 shifts the workflow toward surgical precision: edit once, preview, generate.

Aleph 2.0 builds on Runway’s prior Aleph release, which already emphasized in-context editing of existing material, and aligns with the company’s broader lineup of generation models that prioritize consistency across subjects, styles, and worlds.

The wider industry has moved beyond the initial excitement of raw video generation from text or images.

Attention now centers on reliable temporal consistency, precise user control, and seamless integration of AI into established production pipelines.

Professionals expect tools that handle longer coherent sequences, support fine-tuning to specific brands or aesthetics, incorporate synchronized audio generation, and deliver faster iteration cycles approaching real-time previews.

As competition intensifies among platforms, the emphasis is shifting from sheer creative novelty toward efficiency, predictability, and hybrid workflows that combine the strengths of human-directed footage with generative augmentation.

In this environment, advances like single-frame propagation in Aleph 2.0 signal a maturing phase where AI video becomes a practical extension of the editor's toolkit rather than a replacement for it.

Yesterday we released Aleph 2.0, our upgraded video editing model that lets you change exactly what you want while keeping everything else the same. Available inside our new Edit Studio, you can work with multishot sequences up to 30 seconds long at 1080p.

Learn how to get… pic.twitter.com/shxAKJgUwb

— Runway (@runwayml) May 22, 2026

Around this time, some of the most capable models include Alibaba's HappyHorse 1.0, Kuaishou's Kling 3.0, ByteDance's Seedance 2.0, and the more recently introduced Google's Gemini Omni.

HappyHorse 1.0 leads in overall generation quality with native joint audio-video output and strong motion synthesis. Kling 3.0 excels in photorealism and fluid motion in dynamic scenes. Seedance 2.0 stands out for multimodal references and high character consistency across multi-shot sequences. Google’s Gemini Omni focuses on conversational editing through natural language prompts for iterative, coherent changes.

In comparison, Aleph 2.0 may lack what its rivals excel. But in return, it emphasizes on precise single-frame editing with immediate preview and reliable propagation across clips, preserving original motion, lighting, and camera work with minimal unintended shifts.

For example, HappyHorse 1.0 and Kling 3.0 often require additional iterations for the same targeted continuity, while Seedance 2.0 favors full-scene generation and retexturing, and Gemini Omni can introduce broader variations in lighting or backgrounds that still demand refinement.

Published:

23/05/2026

News

Runway

Aleph

Trends

Review

Google Doubles Down On AI As YouTube Launches 'Ask YouTube' Conversational Search

Linda Sutisna — Sat, 23 May 2026 08:44:49 +0000

AI here, there, everywhere. At least on Google, there shall be no corners without it.

Google has moved with remarkable speed in recent months to weave artificial intelligence into nearly every corner of its product lineup. From major model upgrades such as Gemini 3 and 3.5 Flash to new agentic capabilities rolled out across Search, Workspace, Android, Cloud, and beyond, the company has delivered frequent and wide-reaching updates that embed advanced AI directly into tools people use daily.

This accelerated push reflects a clear strategy of integration rather than isolated experiments, with AI now powering everything from smarter search responses and productivity features to creative and discovery experiences.

YouTube CEO Neal Mohan recently highlighted several new features rolled out by the platform in a busy week of updates that exemplify this broader momentum.

Among the most notable is 'Ask YouTube,' a conversational search tool that allows users to ask detailed questions and receive structured answers drawn from videos across the entire catalog.

Another big week of launches at @youtube. Our teams shipped:

Ask YouTube, a new conversational, interactive search experience that lets you go deeper with your searches
Our Premium Lite individual plan bundled with @Google’s AI Pro plan
Gifts & Jewels in more… pic.twitter.com/MTyopwttGX

— Neal Mohan (@nealmohan) May 22, 2026

The feature compiles relevant clips from both long-form content and Shorts, provides summarized insights, and generates step-by-step guides where appropriate.

Thanks to Gemini that powers it, users can follow up with additional questions to refine their exploration, making it easier to dive deeper into topics such as tutorials, explanations, or recommendations.

The tool is currently available to Premium members in the United States aged 18 and older, with plans for a wider rollout later this summer.

Alongside this search enhancement, YouTube has integrated its Premium Lite individual plan into Google’s AI Pro subscription. This bundling offers subscribers a more affordable way to access ad-free viewing on most non-music videos, along with background playback and downloads, combined with advanced AI capabilities from Google.

The move aims to give users greater flexibility in how they engage with both video content and emerging AI tools.

YouTube has also broadened access to its fan-funding options by expanding Gifts and Jewels to more countries, including Brazil and Mexico. These features provide creators with additional ways to earn directly from their audiences through virtual gifts during streams and other interactive elements.

In parallel, YouTube Music introduced Samples, personalized previews that let listeners discover snippets tailored to specific artists, albums, playlists, and mixes, helping users preview content more effectively before committing to full listens.

These launches demonstrate YouTube's continued focus on improving discovery, monetization, and music experiences for both viewers and creators.

The updates arrive as the platform seeks to balance new interactive tools with ongoing platform operations.

Many in the creator community responded to the announcement by noting the value of the features while also calling attention to existing challenges around content moderation and review processes.

Overall, the developments reflect incremental steps in refining how people search, subscribe, support creators, and discover music on YouTube, all within Google's fast-paced effort to place capable AI across its ecosystem.

Published:

23/05/2026

Google Introduces 'Neural Expressive, A New Design Language For Gemini

Aditya — Sat, 23 May 2026 06:05:31 +0000

Google has introduced 'Neural Expressive' as the new design language for its Gemini app.

Marking one of the company's biggest interface overhauls since launching the AI assistant, the design that was announced during Google I/O 2026 brings fluid animations, vibrant colors, refreshed typography, haptic feedback, and dynamic visual layouts aimed at making AI interactions feel more natural and responsive.

Google says the redesign is intended to move Gemini beyond the traditional "wall of text" chatbot experience that has become common across AI platforms.

Instead of static replies, Gemini can now generate more interactive responses that include images, timelines, graphics, narrated videos, and adaptive layouts tailored to different types of queries.

The company has also integrated Gemini Live directly into the main chat interface, allowing users to switch more seamlessly between typing and real-time voice conversations without opening a separate screen.

The update is rolling out across Android, iOS, and the web alongside the company's newer Gemini 3.5 Flash model.

Read: With 'Gemini 3.5 Flash' And Ads In Mind, Google Redesigns Its Ubiquitous Search For The First Time In 25 Years

Built for the AI era. Meet Neural Expressive, Gemini’s new design language featuring fluid animations, vibrant colors, updated typography, haptic feedback, and visual responses.

— Google Gemini (@GeminiApp) May 22, 2026

The Neural Expressive approach reflects a broader shift in the AI industry, where companies are increasingly focusing not only on model performance but also on how AI systems feel during everyday use.

Smooth transitions, tactile feedback, and visually adaptive interfaces are becoming part of efforts to make AI interactions appear more conversational and immersive.

Google executives and designers have described the redesign as a way to make the interface reorganize itself around generated information instead of presenting long streams of plain text responses.

The update arrives as competition between major AI platforms continues to intensify.

Companies including OpenAI, Microsoft, and Anthropic have all expanded the capabilities and presentation styles of their AI assistants over the past year.

Google appears to be positioning Gemini as a more visually expressive and proactive assistant, particularly with the introduction of new tools such as Gemini Spark and enhanced multimodal responses capable of handling text, voice, images, and video together.

User reactions to Neural Expressive have been mixed.

Many users have praised the modernized appearance and the more energetic feel of the interface, particularly the richer visuals and conversational flow. Others, however, argue that cosmetic improvements do not fully address ongoing concerns around Gemini's practical limitations. Discussions across social platforms and community forums frequently mention rate limits for paid users, occasional reliability issues, and bugs affecting features such as screen sharing and language handling.

Some users say they would prefer stronger improvements in consistency and functionality before major visual redesigns.

Despite the criticism, Neural Expressive represents an important step in Google’s broader AI strategy.

The company is increasingly integrating Gemini across its ecosystem, including Search, Android, Workspace, Chrome, and other services, while simultaneously refining how users interact with AI on a visual and emotional level.

Published:

22/05/2026

Elon Musk's $800 Billion Fortune Makes Other Billionaires Look Poor: Redefining The Scale Of Modern Wealth

Danny Krisnanto — Sat, 23 May 2026 04:13:11 +0000

23/05/2026

Elon Musk's fortune has surged past the $810 billion mark, according to recent estimates from Forbes, further cementing his position as the richest person in modern history.

The staggering rise has been fueled largely by the soaring valuations of companies tied to Musk, particularly Tesla, SpaceX, and xAI, as investors continue betting heavily on electric vehicles, private spaceflight, artificial intelligence, and robotics.

The scale of Musk's wealth now places him in a category that barely resembles traditional billionaire rankings.

His estimated net worth sits hundreds of billions above the world's second richest individuals, with figures such as Larry Page, Sergey Brin, and Jeff Bezos trailing far behind. In other words, Musk's net worth is so far above others that his wealth could make centibillionaires look poor.

The gap highlights how dominant Musk's business empire has become across multiple industries at once, from automotive manufacturing and satellite internet to AI infrastructure and reusable rockets.

Musk's rise did not happen overnight.

Born in South Africa in 1971, he first gained major attention during the dot-com era after co-founding Zip2 and later X.com, which eventually became PayPal. After selling PayPal to eBay in 2002, Musk reinvested much of his fortune into ambitious ventures that many considered unrealistic at the time. Tesla was dismissed by critics who doubted electric vehicles could replace gasoline cars, while SpaceX faced repeated rocket failures before becoming the first private company to reliably transport astronauts and cargo into orbit.

Today, those once risky ventures form the backbone of one of the largest personal fortunes ever recorded.

To put the numbers into perspective, Musk could theoretically purchase every residential property in high value regions such as San Diego County or large portions of Hawaii without exhausting his wealth.

He could buy major luxury automakers many times over, fund massive infrastructure programs comparable to government budgets, or acquire entire industries outright. His fortune also exceeds the annual gross domestic product of most countries on Earth, illustrating just how concentrated modern technological wealth has become.

In fact, he could theoritically buy entire regions of America, or even buy his own country.

Still, much of Musk's net worth exists on paper rather than in cash.

The overwhelming majority is tied to ownership stakes in companies whose valuations can rise or fall dramatically. Selling large portions of those shares would not only reduce his control over the companies but could also impact market prices significantly. That distinction has become central to debates surrounding billionaire wealth, taxation, and economic inequality.

Everything SpaceX And Tesla Say Elon Musk Must Do—Mars Colony, Robots, More—For More Billionshttps://t.co/cjDrA0ul2K pic.twitter.com/7Eyh9WPXXE

— Forbes (@Forbes) May 22, 2026

Online reactions to Musk's new milestone have ranged from admiration to criticism.

Supporters praise his role in accelerating electric transportation, expanding access to space technology, and pushing AI development forward at an unprecedented pace. Critics argue that no individual should command resources larger than the economies of entire nations, pointing to widening wealth inequality and the growing influence of tech billionaires over society and politics.

Social media users have also turned the moment into a flood of memes and mathematical thought experiments.

One of the most widely shared examples involves dividing Musk's wealth among the global population. With the world population now approaching 8.3 billion people, giving every person $100 would cost roughly $830 billion, effectively consuming nearly all of his estimated net worth. Even smaller amounts distributed globally would still total astonishing sums, offering a striking reminder of how enormous these figures really are.

Whether viewed as a symbol of innovation or a warning sign of extreme wealth concentration, Musk's financial ascent continues to captivate the public.

His fortune represents not just personal success, but the immense power and value now concentrated in the worlds of technology, artificial intelligence, and space exploration.

My “net worth” will just track SpaceX+Tesla market cap, so will be whatever that is as a percentage of GDP

— Elon Musk (@elonmusk) March 14, 2026

Sandra Bullock Becomes The First Person To Publicly Purchase A Ticket Online

Linda Sutisna — Fri, 22 May 2026 13:34:22 +0000

28/07/1995

Back in 1995, the internet was still a strange and unfamiliar place to most people. Websites were slow, online shopping barely existed, and many people around the world had never even touched a computer connected to the web.

The young internet was just gaining traction. Dial up models screamed their familiar tune, and the idea of handling over credit card details through those telephone cables to the intenet seemed both thrilling and risky.

Yet during that summer, Sandra Bullock took part in what would become one of the earliest and most memorable moments in online entertainment history.

While promoting her cyber-thriller The Net, she sat down at a desktop computer and purchased two movie tickets online for a showing of her own film on July 28 at a Hollywood theater.

The tickets cost about seventeen dollars, and the transaction was completed through an experimental online ticketing website.

The interface represented the absolute cutting edge of technology: the idea of buying something from a computer without ever leaving the house. It was unbelievable at the time.

Sandra Bullock purchasing a ticket online, considered the first done publicly.

The Net is a 1995 techno-thriller starring Sandra Bullock as Angela Bennett, a computer systems analyst whose life is destroyed after she accidentally uncovers a dangerous digital conspiracy.

The film begins with Angela living a highly isolated life. She works remotely as a software expert, spends most of her time online, orders food and products through her computer, and has very little real-world social interaction. During the mid-1990s, these ideas seemed futuristic.

Her life changes after a coworker sends her a mysterious computer disk containing hidden access to a powerful government-related cybercrime network called the Praetorians. Before he can explain further, he dies under suspicious circumstances.

Soon afterward, Angela travels to Mexico, where she meets a charming man who seduces her before stealing her identification and passport. When she returns to the U.S., she discovers that her entire identity has been erased and replaced. Her name, social records, career history, and personal information have all been altered in digital systems.

To the government and the public, Angela Bennett no longer exists.

Worse, she is now identified as a wanted criminal under a completely different identity.

The rest of the movie follows Angela as she tries to prove who she really is while being hunted by assassins and corrupt operatives who control sensitive computer networks. Because the conspirators can manipulate databases, medical files, criminal records, and financial systems, she finds herself unable to trust institutions that normally protect ordinary people.

At the time of release, many critics viewed the premise as exaggerated science fiction.

However, the movie eventually developed a reputation for being unexpectedly prophetic.

In 1995, only a small percentage of households had internet access, and online shopping barely existed. Many people still viewed the internet with suspicion or outright fear.

The Net tapped directly into those anxieties. The film explored themes like stolen identities, cybercrime, mass data collection, internet dependency, and the vulnerability of digital records. These were ideas that sounded unsettlingly futuristic at the time but would gradually become real-world concerns in the decades that followed.

Now, as the world settles into the new millennium and the internet continues expanding at a remarkable pace, e-commerce and online shopping are beginning to flourish.

More consumers are discovering the convenience of buying goods without ever leaving home. Services that once required a trip to the store or a phone call are steadily moving online, and movie tickets are among the earliest entertainment purchases making that transition.

Companies are racing to experiment with digital storefronts while consumers slowly grow more comfortable with the concept of virtual transactions.

Against that backdrop, Sandra Bullock's now-famous online ticket purchase feels oddly symbolic.

Surrounded by cautious executives during a staged demonstration, Bullock showed how a technology that once seemed almost magical was beginning to evolve into something practical, even if the internet still felt like unfamiliar territory for much of the public.

Sandra Bullock stars in The Net, a thriller that explored some of society's earliest fears about the internet..

Cybersecurity Researcher Found 86,000 Private Screenshots Exposing Alleged Celebrity-Boxer Extramarital Affair Before Internet Frenzy Begins

Bess — Fri, 22 May 2026 10:37:32 +0000

10/05/2026

A major scandal has unfolded at the crossroads of celebrity privacy and digital security after cybersecurity researcher Jeremiah Fowler discovered a vast collection of stolen personal data from a prominent British female celebrity.

The exposed database contained exactly 86,859 screenshots captured from a single mobile phone belonging to the woman who is described across reports as a high-profile entrepreneur and media personality in her thirties. These files included private conversations from messaging apps such as WhatsApp, Instagram, Facebook, and TikTok along with selfies, business invoices, receipts, and thousands of intimate images and messages that pointed to a two-year extramarital affair with a married world-famous boxer who has won multiple titles.

Both parties were married at the time the communications took place though the celebrity has since divorced her businessman husband and the relationship with the boxer is reported to have ended as well.

The breach originated from stalkerware spyware that had been installed on the woman's device likely by someone monitoring her activities.

The perpetrator stored the harvested screenshots on a publicly accessible cloud server without any password protection allowing the data to remain exposed online for at least two weeks before Fowler found it.

Rather than exploiting the material Fowler acted responsibly by notifying law enforcement attempting direct contact with the affected celebrity and her then-husband as well as reaching out to the boxer involved.

The celebrity in question showed little interest in engaging with the researcher who tried to alert her to the exposure choosing instead to block him immediately. Her then-husband did the same. The boxer also did not respond according to available accounts. As of now neither the woman nor the sports star has issued any public statement confirming or denying the affair and no full dataset has surfaced in the wider media or on leak sites.

Because of this, Fowler then published an anonymized summary of his findings through ExpressVPN to draw public attention to the dangers of stalkerware without naming the individuals or releasing any of the actual files.

Media outlets quickly picked up the story in early May 2026 framing it as the exposure of a secret affair between the unnamed British celebrity and the unnamed top sports star.

Sensionally, headlines have leaned heavily into the aspects promising details of a huge British female celebrity's fling with a top sports star yet the core reporting sticks closely to Fowler's verified discovery of the screenshots and the clear evidence of an ongoing romantic and sexual relationship between the two married public figures.

The coverage highlighted the highly sensitive nature of the leaked content which featured not only romantic exchanges but also explicit photos and sexually charged text messages detailing the pair's intimate encounters.

Reports noted that the material would have been deeply damaging if fully released given the public profiles of those involved yet Fowler has consistently refused to identify either person protecting their privacy while emphasizing the broader risks.

Some follow-up articles suggested the boxer himself may have installed the spyware out of suspicion that the woman was cheating on her husband while carrying on with him though this remains an inference drawn from the data rather than confirmed evidence.

This incident underscores the vulnerability of even high-profile individuals to invasive surveillance tools that are readily available on the market.

Stalkerware applications marketed for parental control or employee monitoring can easily be abused to capture every notification screenshot every private photo and every message thread turning a phone into a nonstop recording device.

In this case the cyberstalker's sloppy storage of the files turned a targeted hack into an accidental public leak exposing business details personal contacts and the full scope of the alleged affair including intimate images that captured the couple in compromising sexual situations.

Fowler has stressed that while this story involves celebrities the same threat applies to anyone whose partner ex-partner or acquaintance might install similar software.

Celebrities’ and influencers’ private communications exposed in stalkerware data breach. Read my full report here: https://t.co/GFWrII0rIm

— Jeremiah Fowler Security Researcher (@yoda69) May 1, 2026

Beyond the gossip value the episode serves as a stark reminder of how easily private lives can spill into public view in the digital age.

Everyday users often overlook the permissions granted to apps or the possibility that someone close to them could deploy monitoring tools without consent.

Fowler's decision to publicize the breach anonymously while withholding names highlights a responsible approach to ethical disclosure that prioritizes awareness over exploitation.

Bill & Melinda Gates Foundation Trust Fully Exits Microsoft After 25 Years, Sells Final $3.2 Billion Stake

Linda Sutisna — Fri, 22 May 2026 09:52:06 +0000

15/05/2026

In a move that has captured significant attention in both philanthropic and investment circles, the Bill & Melinda Gates Foundation Trust has fully divested its remaining stake in Microsoft Corporation completing a multi-year wind-down that began in earnest around two years earlier.

According to recent SEC 13F filings disclosed in mid-May 2026, the trust sold its final 7.7 million shares of MSFT during the first quarter of 2026 with the stake valued at approximately $3.2 billion at the time of the transaction.

This brings the foundation's holdings in the company it was originally built upon to zero marking the first time since the foundation's establishment in 2000 that it no longer owns any Microsoft stock.

The news quickly spread through financial media outlets prompting analysts and investors to examine what the sale means for both the foundation’s future and the long-term outlook for Microsoft.

The Bill & Melinda Gates Foundation Trust (often simply called the Gates Foundation Trust) is the investment and endowment arm of the Gates Foundation.

The Bill & Melinda Gates Foundation Trust which manages the endowment supporting one of the world's largest charitable organizations had gradually reduced its Microsoft position over recent quarters starting from roughly 28.5 million shares at the end of the first quarter of 2025 before trimming further and ultimately exiting entirely this year.

While headlines sometimes sensationalize the development as Bill Gates personally selling every last share of his co-founded company the reality centers on the foundation's trust making strategic portfolio decisions separate from Gates’ personal holdings.

The sale aligns with the foundation's broader plan to wind down operations over the next approximately twenty years directing all assets toward immediate charitable spending rather than maintaining a perpetual endowment. By liquidating concentrated positions like Microsoft the trust gains greater liquidity to fund tens of billions in annual grants focused on global health education poverty reduction and other high-impact causes.

Analysts emphasize that the divestment does not reflect any loss of confidence in Microsoft's business fundamentals.

The company continues to demonstrate exceptional strength with trailing revenue exceeding $281 billion operating income around $149 billion and free cash flow surpassing $73 billion.

Its leadership in cloud computing through Azure artificial intelligence infrastructure and enterprise software remains unchallenged with high switching costs protecting its dominant market position.

The foundation's decision appears driven primarily by prudent risk management including portfolio diversification to reduce exposure to a single stock that had performed exceptionally well and the practical need to generate cash for ongoing philanthropic commitments.

Notably billionaire investor Bill Ackman disclosed a substantial new purchase of Microsoft shares around the same period illustrating that other major investors continue to view the company favorably even as the Gates Foundation steps away.

Bill Gates co-founded Microsoft in 1975, the company that made his a billionaire. After retiring as its chief, he shifted his focus to philanthrophy.

Microsoft's stock experienced only modest movement in the immediate aftermath of the filing with reports noting a slight dip of about 0.42% on the day of disclosure before stabilizing. Market observers point out that large institutional sales of this nature are often viewed as routine fiduciary actions rather than bearish signals especially when executed gradually over multiple quarters.

The Gates Foundation Trust had already been trimming its position steadily which helped minimize any sudden market disruption. For the foundation itself the proceeds from the sale further bolster its ability to accelerate grantmaking as it prepares for an eventual full transition away from holding large equity stakes.

This development closes a symbolic chapter in the intertwined history of Microsoft and the Gates Foundation.

What began as the personal fortune of the company's co-founder evolved into one of the most powerful engines for global philanthropy with Microsoft stock once forming a major portion of the endowment. Yet the complete exit underscores a deliberate shift toward greater flexibility in how charitable resources are deployed.

Supporters of the foundation see the move as responsible stewardship ensuring that funds are actively used to address pressing global challenges rather than remaining tied up in long-term investments. As the foundation continues its planned sunset by around 2045 such portfolio adjustments are expected to become more common to maximize near-term impact.

The news don't make a significant change in the market.

Overall the sale represents a significant but carefully managed transition rather than a dramatic rupture.

Microsoft remains a powerhouse in technology with robust growth prospects particularly in AI and cloud services while the Gates Foundation Trust gains enhanced capacity to pursue its mission of improving health and opportunity worldwide.

Investors monitoring the situation are largely encouraged to view the event through the lens of standard asset allocation strategy instead of interpreting it as a red flag for the company's future.

The story highlights how even the most storied connections between a founder a company and its associated philanthropy can evolve over time as priorities shift toward maximizing real-world impact in an ever-changing global landscape.

Long story short, Bill Gates did not sell his personal Microsoft shares.

At this time, he still owns a very large personal stake, which is roughly 103 million shares (worth around $43 billion at recent prices).

Tor Project Launches Crowdfunding Campaign To Support Internet Freedom Projects

Danny Krisnanto — Fri, 22 May 2026 08:46:08 +0000

19/05/2026

The internet is an open space. It's crowded, noisy, and packed with more information than what any person can handle in their lifetime. However, privacy is scarce.

Trackers are everywhere, and almost all websites and platforms out there thrive by knowing their users in one way or another, by knowing too much.

And here, in a bold new effort to safeguard digital privacy and support essential tools for online freedom, the Tor Project has teamed up with Funding the Commons to launch a groundbreaking web3-native crowdfunding campaign that benefits ten vital internet freedom projects.

According to the announcement, the initiative aims to create a more sustainable transparent and community-driven way to fund critical digital infrastructure at a time when traditional funding sources face increasing instability and global internet freedom has been declining for fifteen consecutive years.

The campaign which can be found at internetfreedom.torproject.org and its corresponding Onion Service at http://swvbwbtmajvfrnz4wztx6ovshilm23ntigi73fz5wczj3aqdquq5icad.onion invites donations in Bitcoin Ethereum Zcash Monero and Golem and runs through June 18 2026 with an initial matching pool of one hundred fifteen thousand dollars backed by generous match funders including Cake Wallet Zcash Community Grants Logos and Octant with additional ecosystem participation expected as the drive progresses.

Tor is strongest when the broader internet freedom ecosystem is healthy. In partnership with @FundingCommons we are excited to launch a web3-native crowdfunding initiative benefitting 10 internet freedom projects at https://t.co/hbFUyPMYV5.

Thank you to or generous matchfunders… pic.twitter.com/9yDIs3vBlT

— The Tor Project (@torproject) May 19, 2026

Internet freedom has been under growing pressure as governments and powerful actors deploy increasingly sophisticated tools for monitoring restricting access and censoring information.

Millions of people around the world rely every day on nonprofit projects that provide secure communication channels whistleblower protection uncensored news access and circumvention of censorship yet these organizations often struggle with unstable grant funding leading to staffing cuts delayed software updates and reduced support for the most at-risk communities.

Recognizing that the Tor network itself is strongest when the broader ecosystem of privacy and anti-censorship tools remains healthy and vibrant the Tor Project decided to take a leading role in exploring new participatory funding models that can reduce dependence on potentially unreliable institutional sources and instead empower everyday users and supporters to have a direct say in sustaining the tools they use.

At the heart of this campaign is quadratic funding a sophisticated participatory matching system designed to amplify the impact of many small donations rather than letting the process be dominated solely by a few large contributors. In quadratic funding the amount a project receives from the matching pool grows with the square root of the number of unique donors it attracts which mathematically rewards broad community support and gives even modest contributions real weight in determining final allocations.

This approach has gained traction in web3 circles as an innovative way to fund public goods because it aligns incentives toward collective participation and democratic decision-making rather than top-down grantmaking.

The one hundred fifteen thousand dollar matching pool will be distributed across the ten selected nonprofit projects based on how the community votes with its donations during the campaign period creating a transparent and merit-based distribution that reflects real user priorities.

The ten beneficiary projects represent a carefully chosen cross-section of the internet freedom ecosystem each addressing different but interconnected challenges in privacy secure communication and resistance to censorship.

This initiative arrives at a pivotal moment when many regular users of Tor Browser and related privacy tools are increasingly aware of the importance of contributing directly to the infrastructure they depend on.

By making donations crypto-native and accessible through both a clearnet site and a dedicated Onion Service the campaign lowers barriers for privacy-conscious supporters who prefer to contribute without compromising their own anonymity. Early momentum has already generated discussion across privacy cryptocurrency and open-source communities underscoring a shared recognition that protecting internet freedom requires ongoing collective action from users builders and funders rather than relying solely on sporadic large grants.

For those interested in participating the full details including real-time donation progress the complete list of beneficiary projects and step-by-step instructions are available directly on the campaign site at internetfreedom.torproject.org or via its Onion Service.

The Tor Project has stressed that this is not only about raising immediate funds but also about testing and refining models that could be scaled in the future to support even more projects that often struggle to reach traditional donors.

This effort stands as a powerful example of how the communities that benefit most from these tools can come together to secure their future in innovative and transparent ways.

A Social Media Influencer, Her Father, And Her Boyfriend Arrested In An Alleged Murder-For-Hire Scheme

Bess — Fri, 22 May 2026 07:47:34 +0000

19/05/2026

In a case that blends social media fame, family ties, and a bitter custody battle, a popular TikTok influencer, her father, and her boyfriend now face serious criminal charges in an alleged murder-for-hire scheme.

Gabriela Lauren Gonzalez, better known online as Gabbie Gonzalez, has built a sizable online audience with nearly half a million TikTok followers through videos centered on cooking, dancing, lifestyle content, and glamorous travel moments. Her social media persona often portrayed a carefree image, with posts featuring tropical scenery, colorful bikinis, and captions describing herself as a "lover of life" and a "mermaid at heart."

Prosecutors now allege that behind the scenes, she became involved in a years-long conspiracy tied to a bitter custody battle over her seven-year-old daughter.

And now, the -24-year-old woman has been charged alongside her father, 59-year-old Florida civil trial lawyer Francisco Javier Gonzalez and her ex-boyfriend Kai Faron Cordrey in an alleged murder-for-hire plot targeting singer Jack Avery, a former member of the boy band Why Don't We and the father of her child.

Gabbie Gonzalez.

The roots of the alleged conspiracy trace back to 2020 and 2021 when Gonzalez and Avery became locked in a heated and tense custody battle over their young daughter.

According to prosecutors the conflict grew increasingly acrimonious with Gonzalez turning to her father for support that went far beyond typical parental guidance.

Francisco Gonzalez, a personal injury attorney based in Central Florida, reportedly became deeply involved in the dispute. Authorities claim he even hired a private investigator to surveil Avery in hopes of uncovering evidence such as drinking or smoking that could be used against him in family court.

Witnesses told investigators that Francisco allegedly threatened to have Avery arrested if he showed up for visitation and attempted to leverage law enforcement contacts. Avery later described living in constant fear during that period saying he stayed inside his home for an entire month straight and found himself anxiously looking out the window every night.

Prosecutors allege that during this time Gabriela Gonzalez sought help from her then-boyfriend Kai Faron Cordrey to find someone on the dark web who could carry out the killing of Avery, with payment using Bitcoin.

Why Don't We boyband, with Jack Avery at the far right of the photo.

The plan reportedly involved paying a hitman to eliminate Avery by whatever method proved easiest including a shooting or staging an accident.

In April 2021 Francisco Gonzalez allegedly wired ten thousand dollars to Cordrey under the pretense of payment for web development work with the funds intended as front money to locate hire and compensate the supposed assassin from the dark web.

Two months later in June 2021 Cordrey requested and received an additional four thousand dollars after claiming the hitman needed more money to proceed. Several days after that second payment Cordrey pushed for the murder to happen within just a couple of days.

The scheme unraveled when Cordrey began communicating with an undercover law enforcement officer who was posing as the hired killer.

In those exchanges Cordrey identified Jack Avery as the target discussed payment arrangements and specifically requested proof of death once the act was completed. He also told the undercover agent that Gabriela Gonzalez wanted the murder to happen and that her father would cover the costs. The FBI launched the investigation into the plot which was eventually turned over to the Los Angeles County District Attorney’s Office.

View this post on Instagram

A post shared by Gabriela Gonzalez (@gabbieegonzalez)

According to the Los Angeles County District Attorney’s Office in a website post, prosecutors formally charged all three individuals with one count each of attempted murder conspiracy to commit murder and solicitation of murder. If convicted they could each face up to twenty-five years to life in state prison.

Gabriela Gonzalez was arrested last week in Northern California specifically in Humboldt County where she was taken into custody at or near the California Redwood Coast-Humboldt County Airport.

She is currently being held in Los Angeles County either without bail or on a high bail amount depending on the latest court records.

Her father Francisco Javier Gonzalez was arrested on Monday May 18 2026 in Lake Mary Florida on a California warrant. During his initial court appearance in Seminole County his attorney indicated that he plans to waive extradition and return to California to face the charges.

Kai Faron Cordrey who is twenty-six years old has also been charged though reports indicate he has yet to be detained and his current status remains unclear.

Gabbie Gonzalez and her father, Francisco Javier Gonzalez.

Los Angeles County District Attorney Nathan J. Hochman commented on the unusual nature of the case, stating that while most fathers teach their children to obey the law, this case alleges a father helped orchestrate an extraordinarily violent crime involving his daughter and her boyfriend.

Following the arrests, Avery released a public statement saying he and his family had endured years of "false accusations" and "painful attacks" on their reputation. He explained that he largely stayed silent throughout the legal battle to protect his daughter and respect the judicial process. Avery added that his primary focus remains raising his daughter, who he said is "safe, healthy, and deeply loved."

He has since reportedly filed for a restraining order against Gonzalez.

The case has drawn widespread attention not only because of the serious allegations, but because it involves a social media influencer whose online image centered around positivity, family moments, and lifestyle content. Prosecutors allege the custody battle between Gonzalez and Avery deteriorated so severely that it eventually crossed into criminal territory, culminating in a multi-year federal investigation.

Authorities emphasize that all three defendants are presumed innocent unless proven guilty in court.

James Bond Breaks 40 Years Of Tradition By Having A New Watch Designed For A Video Game

Danny Krisnanto — Fri, 22 May 2026 03:58:30 +0000

22/05/2026

James Bond has always been synonymous with iconic watches that define his image on screen, but a new chapter unfolds with the Omega Seamaster Diver 300M Chronograph 007 First Light.

For the first time in the modern era of the franchise, Omega has created a signature timepiece not for a major motion picture but for a video game titled 007 First Light.

This bold decision shatters a roughly 40 year tradition and marks the first time the brand has tied a dedicated watch to a game rather than a film.

The move also reflects how the James Bond franchise is evolving alongside the entertainment industry itself. While Bond watches have traditionally debuted on the wrists of actors in blockbuster films, 007 First Light positions the character within an interactive medium aimed at a new generation of fans. By attaching a bespoke Seamaster to the game, Omega signals that video games are no longer secondary tie-ins but major cultural events capable of carrying the same prestige and marketing power as a theatrical release.

Beyond its symbolism, the Omega Seamaster Diver 300M Chronograph 007 First Light continues the long-standing partnership between Omega and the Bond universe with a design that blends tactical functionality and cinematic flair.

View this post on Instagram

A post shared by OMEGA (@omega)

The history of James Bond watches stretches back to the very beginning of the film series.

In 1962s Dr. No, Sean Connery wore a Rolex Submariner that immediately became part of the spy's rugged, reliable persona. Rolex models continued to dominate through the Connery and George Lazenby years, appearing as the perfect tool for underwater action and high stakes missions. When Roger Moore took over in the 1970s, the watch choices shifted toward Seiko, with several memorable digital and quartz models.

The most notable among them came in 1985s A View to a Kill, when Moore strapped on a Seiko chronograph, the last time audiences saw Bond wearing any chronograph until now.

Timothy Dalton briefly returned to Rolex in the late 1980s, but everything changed in 1995 with Pierce Brosnan in GoldenEye. That film introduced the Omega Seamaster Professional on Bond’s wrist and launched a long term partnership that has lasted through the entire Daniel Craig era.

Every Omega Seamaster worn by Bond since then has been a clean three hand diver or similar complication free model, reinforcing a consistent tool watch identity focused on simplicity and durability.

Now, the new Seamaster Diver 300M Chronograph 007 First Light breaks that long standing pattern in dramatic fashion.

It is a robust 44 millimeter stainless steel diver with a striking black ceramic dial featuring a laser engraved wave pattern that echoes the classic Seamaster aesthetic.

Its unidirectional black ceramic bezel carries a white enamel diving scale, while the chronograph pushers match in polished ceramic. Bronze gold accents bring subtle warmth through a PVD ring around the three oclock subdial and a matching central chronograph seconds hand, adding a distinctive touch that feels both tactical and luxurious.

Powered by the Co Axial Master Chronometer Calibre 9900 with a 60 hour power reserve, the movement delivers exceptional precision and includes a column wheel chronograph capable of timing up to 12 hours.

Water resistance sits at a solid 300 meters, and the piece ships on a black, grey, and beige NATO strap with engraved 007 and First Light markings on the titanium keepers.

Additional game inspired NATO straps are available separately for collectors who want to mix and match.

What makes this release truly groundbreaking is its deep integration with the game.

Set for launch on May 27, 2026, from IO Interactive and Amazon MGM Studios, 007 First Light serves as an origin story following a 26 year old James Bond as he earns his double O status. In the game the watch functions as more than a stylish accessory. It acts as a genuine gadget with subdials that support hacking mechanics and other tactical tools, complete with in game features like a laser equipped strap that the real world version understandably omits.

Omega worked closely with the developers to translate the digital design into a fully functional production watch, right down to the special 007 First Light logo metallized on the sapphire caseback and the Pelican style presentation box modeled after the in game Omega carrying suitcase.

This move represents fresh territory for both Omega and the Bond franchise.

By introducing this Seamaster chronograph now, Omega revives a gadget heavy spirit that feels perfectly suited to a younger, more reckless version of the spy while bridging the gap until the next film arrives. At 9400 dollars it sits firmly in the premium segment, yet it delivers genuine horological substance alongside its pop culture cachet.

For longtime fans of both Bond and fine watches, the 007 First Light chronograph offers something genuinely new.

It honors the heritage of the Seamaster Diver 300M line while pushing it forward into the world of interactive entertainment, proving that the spy who defined so many cinematic moments can evolve right alongside his audience.

With 'Gemini 3.5 Flash' And Ads In Mind, Google Redesigns Its Ubiquitous Search For The First Time In 25 Years

Aditya — Fri, 22 May 2026 03:27:39 +0000

The competition in the large language model (LLM) space has intensified sharply over the past few years.

OpenAI, Anthropic, and a growing number of startups have rolled out conversational tools that many people now use as their primary way to find information, answer questions, or complete tasks. Google Search nevertheless remains the undisputed leader in online queries, processing billions of searches daily and retaining its position as the default starting point for most internet users worldwide.

Even so, a noticeable shift is underway.

Users who once typed short keywords into the classic search box are increasingly experimenting with full-sentence prompts and autonomous agents elsewhere, treating traditional search as something to be supplemented or even replaced when a more proactive AI feels more natural.

Google has responded directly to this evolution.

At its I/O 2026 event, the company announced the most substantial redesign of its search interface in more than 25 years.

The familiar rectangular box that has defined the Google homepage since 2001 is now being replaced by an intelligent, dynamically expanding input field.

In the announcement, under the AI Search section, Google said that:

"AI Mode is our most powerful AI Search, and it has surpassed more than 1 billion monthly users. And today we're upgrading the experience with Gemini 3.5 Flash as the new default model, globally. We're seeing incredible momentum, with AI Mode queries more than doubling every quarter since launch. And last quarter, we saw Search queries reach an all-time high."

"Today we're launching the biggest upgrade to our Search box in over 25 years — a new, intelligent Search box, now completely reimagined with AI. You can search using text, images, files, videos and Chrome tabs and Search reasons across them all. You’ll continue to get a range of results from Search, just like you do today."

"We're also making it even easier to continue the conversation with Search, bringing AI Overviews and AI Mode into one, seamless AI Search experience. You can flow effortlessly from your question, to a search results page with an AI Overview, to a follow-up in AI Mode, all with links to learn more."

In other words, Google Search's search bar can now comfortably handles longer, conversational queries and accepts a wide range of inputs including images, videos, files, and even open Chrome tabs.

Powered by the new Gemini 3.5 Flash model as the default for AI Mode, the updated search experience merges previous features such as AI Overviews and AI Mode into a single seamless flow. Traditional blue-link results remain visible, but the emphasis moves toward richer, more interactive responses that better match how people already talk to modern AI systems.

Beyond the visual and input changes, the update introduces persistent information agents that users can create simply by asking.

These agents operate in the background, continuously monitoring topics and delivering updates without repeated manual searches. Someone researching apartment rentals, for example, can set an agent to track new listings and notify them in real time. Similar capabilities extend to tracking product releases, sports news, stock movements, or any other ongoing interest.

The system can also generate custom visuals, interactive dashboards, trackers, or small applications on the fly, pulling in real-time data such as location or weather to create tailored mini-experiences directly within search.

Complementing these search-side agents is Gemini Spark, a new mode within the Gemini app designed for truly autonomous background work.

Gemini Spark is your new 24/7 personal AI agent.

Give it a task and it works autonomously in the background, even if your phone and laptop are turned off. You choose to turn it on and it's designed to check with you before taking major actions. #GoogleIO

— Google Gemini (@GeminiApp) May 19, 2026

Spark can scan emails, documents, credit card statements, or other connected sources to surface important updates, compile summaries, or maintain to-do lists even when devices are offline or locked. It runs on cloud infrastructure, continuing its tasks independently and prompting users only when confirmation is needed for actions like spending money or sending messages.

This capability reflects a broader push toward agentic AI that does not require constant user prompting, addressing one of the limitations that has kept many people toggling between traditional search and dedicated AI assistants.

On the commercial front, Google is simultaneously updating its advertising ecosystem to fit the new conversational environment.

New formats built with Gemini include conversational discovery ads and highlighted answers that provide personalized guidance and transparent recommendations during a user's research process. AI-powered shopping experiences aim to turn longer, context-rich queries into smoother paths toward purchase, with expanded direct-offer options and native checkout in select categories.

These changes allow brands to participate more naturally in the dialogue rather than appearing as isolated sponsored links, aligning advertising with the same agentic and multimodal shift happening in search itself.

Gemini 3.5 Flash is starting to roll out to everyone globally, free of charge.

Try it by selecting “3.5 Flash” from the model dropdown menu at https://t.co/382WL5xSvc or in the app.

— Google Gemini (@GeminiApp) May 20, 2026

Taken together, the announcements show Google attempting to bridge two worlds. It is reinforcing the enormous reach and reliability of its core search product while infusing it with the proactive, conversational, and autonomous qualities that have drawn users toward newer AI tools.

By expanding the search bar, enabling background agents, and evolving ad formats accordingly, the company hopes to keep people inside its ecosystem longer instead of sending them to separate chatbots or specialized apps.

Whether this strategy fully counters the appeal of competitors or simply slows the migration remains uncertain, Google is kind of . What is clear is that Google is treating the redesign as a foundational step rather than an incremental tweak, acknowledging that the way people search is changing and that its flagship product must change with them.

Published:

20/05/2026

The Annoying Copilot Floating Button Can Now Disappear, As Microsoft Admits Forcing It In Office Apps Was A Mistake

Aditya — Fri, 22 May 2026 03:18:34 +0000

Microsoft has been trying to push Copilot deeper into more of its products.

And one of the more visible efforts involved adding a floating button for the AI assistant directly into Word, Excel, and PowerPoint. The button sat in the bottom right corner of the screen by default, designed as a constant prompt to encourage users to interact with the feature. Many people quickly found it intrusive, especially in Excel where it covered up cells and made everyday work feel cramped.

Complaints piled up fast, with users calling the always present icon disruptive and unnecessary for those who preferred accessing Copilot through the regular ribbon menu instead.

The company later admitted that forcing the floating button on everyone had been a misstep.

Even though internal data showed that engagement with Copilot increased because of the prominent placement, the backlash highlighted how much it got in the way of normal workflows.

The stubborn floating Copilot button can now go away.

Microsoft had intended the button to act like a helpful partner that appeared at the right moment rather than an interruption, but the execution clearly missed the mark for a large portion of the audience.

After listening to the feedback, the team decided to give users more control rather than keep the rigid approach.

The latest update lets people right click the Copilot button to move it back to the ribbon at the top of the app or dock it neatly to the side of the window. Once docked, it now stays in place for the entire document session instead of bouncing back to the floating position every time someone uses it.

These changes started rolling out across Word, Excel, and PowerPoint on Windows, Mac, and the web in the final week of May 2026.

The adjustments apply whether users access the apps through the desktop versions or the browser based editions.

In a blog post announcing the roll back, Microsoft said that:

"We’ve been working to make Microsoft 365 feel more connected and integrated with Copilot, available as a helpful thought partner, when you need it. We recently updated Word, Excel and PowerPoint with a consolidated, more visible entry point (Dynamic Action Button) at the bottom right of the apps that suggests actions from Copilot based on your document. You may have noticed similar entry points in other apps across Microsoft 365, all with access to the same Copilot."

"As we continue to thoughtfully and intentionally shape how Copilot integrates with the tried-and-true workflows within Word, Excel and PowerPoint that you rely on, we’re listening, learning and improving as we go. To that end, we're already making a few updates based on feedback."

The move reflects a broader pattern with Copilot features where Microsoft experiments aggressively to boost adoption but then pulls back when the friction becomes too obvious.

For now, the floating button is no longer mandatory, which should ease some of the frustration that built up over the past months.

Published:

22/05/2026

Google Accidentally Exposes 3-Years-Old Chromium Browser Bug That Could Turn Browsers Into Silent Botnet Proxies

Aditya — Fri, 22 May 2026 01:53:07 +0000

A serious security issue in Chromium-based browsers has drawn widespread attention after Google inadvertently exposed details of a long-standing and still-unfixed vulnerability.

What began as an internal bug report turned into a public concern when information about the flaw, including proof-of-concept exploit code, briefly appeared on the Chromium issue tracker before being pulled. Although the company quickly restricted access again, the details had already spread, leaving millions of users potentially at risk while a proper patch remains unavailable.

The vulnerability centers on the Background Fetch API, a legitimate feature designed to let browsers continue downloading large files or videos even after a user closes a tab or the entire browser window.

Security researcher Lyra Rebane first reported the issue privately in late 2022, highlighting how malicious websites could abuse this mechanism.

Post by @rebane2001@infosec.exchange

View on Mastodon

By using clever JavaScript and service workers, attackers can keep code running in the background indefinitely, creating persistent connections that survive browser restarts and sometimes even device reboots.

In some browsers like Microsoft Edge, the process can occur almost silently without obvious pop-ups or warnings, making it especially stealthy.

This flaw does not grant full control over a device or access to personal files, emails, or the operating system. Instead, it turns a user's browser into a lightweight participant in a potential botnet.

Compromised devices could relay traffic anonymously, contribute to distributed denial-of-service attacks, or serve as proxies for other malicious activities.

While building a massive network would require significant effort, even smaller-scale exploitation is relatively straightforward once the code is public, and the invisible nature of the background activity means most people would never notice anything unusual happening on their phones or computers.

Post by @rebane2001@infosec.exchange

View on Mastodon

What makes the situation more troubling is the length of time the vulnerability went unaddressed.

Reported over three years ago, it was internally rated as high priority by Google engineers yet lingered without a fix. The bug tracker entry had even been marked as resolved at one point, which led to the accidental public exposure when access restrictions were lifted.

As of now, the vulnerability affects virtually all Chromium-based browsers, including Google Chrome, Microsoft Edge, Brave, Opera, Vivaldi, and others, putting billions of users worldwide in the crosshairs. Firefox and Safari remain unaffected since they do not rely on the same engine.

Post by @rebane2001@infosec.exchange

View on Mastodon

Google has acknowledged the leak and stated that work on a fix is underway, but no timeline has been provided and no emergency update has rolled out yet.

In the meantime, users are advised to exercise caution when visiting unfamiliar websites, avoid clicking suspicious links, and keep their browsers updated in hopes that a patch arrives soon.

This incident serves as a stark reminder of how even well-intentioned browser features can be twisted for harm when left unresolved, and it underscores the importance of prompt attention to security reports in an ecosystem that powers so much of our daily online activity. Until the patch lands, staying vigilant remains the best defense against this lingering threat.

Post by @rebane2001@infosec.exchange

View on Mastodon

Published:

22/05/2026

'Gemini Omni' From Google Can 'Create Anything From Anything': Huge Gain But Not Quite There Yet

Aditya — Thu, 21 May 2026 14:52:31 +0000

The AI landscape has been defined by the relentless large language models (LLMs) war.

Ever since OpenAI's ChatGPT debuted and reshaped expectations for what conversational intelligence could achieve, the battle involving giants that shook the entire tech sphere. And Google, has astonished the world with Gemini.

Then, the powerhouse in search and foundational models, has also astonished the world with its Veo video generation models, particularly the Veo 3 and Veo 3.1, which delivered photorealistic clips with sounds that seemed poised to dominate the emerging field of generative media.

Those early demonstrations highlighted Google's deep expertise in rendering realistic motion and environments, positioning the company as a frontrunner in turning text and images into coherent video sequences that captured everyday physics and visual detail with surprising fidelity.

Yet the competitive dynamics shifted rapidly as specialized tools like ByteDance's Seedance 2.0 and Alibaba's HappyHorse 1.0 emerged from other developers, pushing the boundaries of multimodal video creation with superior motion fidelity, audio synchronization, and creative control that left many observers noting how Google's earlier leads had been eclipsed.

Then comes 'Omni,' which according to Google in the announcement, is literally an "any-to-any" model that can create "anything from anything," and "everything from everything."

Omni brings together an improved understanding of physics with Gemini's knowledge of history, biology, and culture, bridging the gap from photorealism to meaningful storytelling.

Actions have consequences, environments respond to events, and narratives evolve logically. pic.twitter.com/ajQ3purg0g

— Google DeepMind (@GoogleDeepMind) May 19, 2026

Seedance 2.0 excelled in handling combined inputs of text, images, video clips, and audio to produce cinematic sequences with director-level precision over camera movements, lighting, and narrative flow, smf Happyhorse 1.0 stood out for its top-ranked performance in blind human evaluations, generating dialogue-driven scenes with natural lip sync and immersive sound that felt remarkably lifelike.

These models raised the bar not just on output quality but on usability, allowing creators to iterate on complex, multi-shot stories without restarting from scratch, and they quickly became the benchmarks against which every new release was measured.

With the introduction of Gemini Omni, Google appears to be mounting a serious effort to catch up in this fast-evolving space, integrating the reasoning prowess of its Gemini models with advanced generative capabilities to enable more intuitive video editing and world-aware storytelling.

You can even reimagine the action in a video you took by asking Gemini Omni.

Transform your world instantly - change the environment, add new objects, or create something completely unexpected. pic.twitter.com/nMUyN7jNzW

— Google DeepMind (@GoogleDeepMind) May 19, 2026

And thanks to Gemini, this Omni includes features like conversational editing that let users reimagine a clip by describing changes to environments or objects while preserving the original intent.

Gemini Omni doesn't just build scenes that look real, it reasons about what should happen next. It combines an intuitive understanding of physics with Gemini's knowledge of history, science, and cultural context.

Rolling out today starting with video outputs to Google AI Plus,… pic.twitter.com/EkLjv5O0dN

— Sundar Pichai (@sundarpichai) May 19, 2026

And also thanks to Gemini, users can use Omni to create their own avatars.

Create videos with your own voice and likeness using avatars with Gemini Omni.

When you create an avatar, you have an AI digital version of yourself so you can easily generate videos that look and sound like you. No need to upload your image every time.

— Google Gemini (@GeminiApp) May 19, 2026

Announced as the first step toward a system that can create anything from any input, Omni Flash focuses on video for now but emphasizes native multimodality, letting users feed in text prompts, reference images, audio clips, or existing footage and then refine outputs through natural conversation.

It promises stronger world understanding, where actions follow logical consequences, characters remain consistent across scenes and lighting conditions, and cultural or scientific context informs how narratives unfold, bridging pure photorealism with something closer to meaningful storytelling.

In Google's words, Omni is the "Nano Banana" for videos.

Easily edit videos by just describing what you want to see with Gemini Omni.

Chat with Gemini naturally to add details, make edits, and watch your ideas come to life.

Every instruction builds on the last. Your characters stay consistent, the physics hold up and the scene… pic.twitter.com/tVKn5dQZuh

— Google (@Google) May 22, 2026

Edit your own videos with Gemini Omni.

Change specific things, or change everything. Your video becomes the starting point for something you never could have filmed yourself. pic.twitter.com/YrqRiNyPcj

— Google (@Google) May 22, 2026

The model is already rolling out in the Gemini app, Google Flow, and YouTube Shorts for subscribers, with API access planned soon, and

Gemini Omni can create anything from any input, starting with video.

This means you can combine images, audio, video and text as input and generate high-quality videos.

Or use drawings to create in a way that matches your vision.#GoogleIO pic.twitter.com/JIQW7uGfAZ

— Google (@Google) May 19, 2026

Early demonstrations show Omni Flash producing clips with improved physics simulation and continuity, such as characters interacting naturally with their surroundings or styles being applied seamlessly from reference inputs.

Gemini Omni is so fun - insanely great at editing videos! pic.twitter.com/KDGzcU78Wt

— Josh Woodward (@joshwoodward) May 19, 2026

Yet while these advances represent a clear step forward for Google in blending intelligence with media generation, they have not yet closed the full gap left by the current leaders.

Comparisons shared shortly after launch highlight that Omni still trails Seedance 2 in overall cinematic polish and motion stability during extended sequences, and it falls short of Happyhorse 1's benchmark dominance in realistic character animation and synchronized audio.

Some commenters even said that Omni also trails behind Kuaishou's Kling 3.0.

It's worth noting here that Seedance, HappyHorse, and Kling are all developed by Chinese companies.

Some outputs reveal lingering inconsistencies in complex physical interactions or subtle artifacts that more mature competitors have largely resolved, suggesting that Omni excels more at editable, context-aware refinement than at generating flawless video from the ground up on its first attempt.

Your browser does not support the video tag.

In a field moving at breakneck speed, Google's latest entry keeps the company firmly in the race.

While its debut claims no outright victory, it underscores how quickly the balance of power can tilt even among the biggest players.

Published:

21/05/2026

WordPress 7.0 'Armstrong' Released: Offering Greater Flexibility In Web Development

Aditya — Thu, 21 May 2026 14:45:05 +0000

Today, in the era of social media and AI, the web has become more interactive, intelligent, and deeply integrated into everyday life.

And WordPress is still one of the most widely-used content management systems out there. With 'WordPress 7.0,' named "Armstrong" after the influential jazz musician Louis Armstrong, the platform wishes to provide users with greater flexibility to shape their online presence in personal and creative ways.

This major update introduces foundational AI capabilities into the core platform for the first time.

A new WP AI Client enables WordPress to connect with generative models from providers including OpenAI, Google Gemini, and Anthropic Claude.

WordPress 7.0 “Armstrong” is here.

This major release introduces foundational AI tools, a refreshed admin experience, expanded design controls, new blocks, and powerful developer APIs.

Explore what’s new, update when you’re ready, and start building with WordPress 7.0 today.… pic.twitter.com/eIVtWblxo5

— WordPress (@WordPress) May 20, 2026

These connections are managed centrally through a Connectors screen in the dashboard, where API keys can be added and organized securely.

The system supports features such as generating post titles and excerpts, suggesting alternative text for images, and working with image generation or editing tools. An accompanying AI plugin extends these options further, while a new client-side JavaScript package brings abilities directly into the browser environment, complete with a command palette interface for quick access.

The administrative interface has undergone a significant modernization.

The dashboard now features an updated color scheme, smoother transitions when navigating between sections, and a cleaner overall appearance that improves daily use. A global command palette, triggered by Ctrl+K on Windows or Cmd+K on Mac, allows users to search for and access tools from anywhere in the admin area without hunting through menus.

Font management has been unified into a dedicated section that works across block-based, hybrid, and classic themes, simplifying the process of installing, uploading, and organizing typefaces.

Revisions now include a visual scrubbing tool that displays changes at a glance, making it easier for editors to review versions side by side and restore earlier states with greater precision.

Several new blocks and design tools expand the possibilities within the editor.

The gallery block gains support for lightbox slideshows, allowing images to be viewed in an expanded, interactive format. Additional blocks for headings, breadcrumbs, and icons provide more granular control over common elements.

Responsiveness options have been enhanced so that blocks can be shown or hidden based on specific device types without impacting other viewports, and users can define custom breakpoints to fine-tune layouts.

Navigation overlays can now be constructed entirely with blocks and patterns, offering full styling control over columns, typography, and even custom close buttons.

Patterns themselves function more flexibly as cohesive units that can be inserted, detached, and modified individually. Custom CSS can also be applied directly at the block level within the editor, giving designers more immediate control over details.

On the development side, WordPress 7.0 introduces tools aimed at speeding up and simplifying building processes.

Blocks and patterns can now be created server-side using only PHP, with automatic registration through the block API. The Site Editor benefits from improved extensibility, including better routing support and a new boot package that allows plugins to construct custom pages within the editor environment. These changes, combined with various API enhancements and performance improvements, provide developers with a more capable foundation for extensions and custom solutions.

The release also raises the minimum PHP requirement, dropping support for older versions to encourage more secure and efficient hosting environments.

More than 875 contributors from countries around the world participated in bringing WordPress 7.0 to completion, among them over 200 first-time contributors.

Their combined efforts resulted in roughly 420 enhancements and bug fixes. Matias Ventura served as release lead, supported by a coordinated team handling release coordination, technical leadership, triage, and testing. The collaborative nature of the project remains central to its ongoing development as open-source software.

Sites running earlier versions can update through the standard process in the dashboard when administrators determine the timing is appropriate, ideally after testing in a staging environment to ensure compatibility with existing themes and plugins.

Published:

21/05/2026

From Consumer To Victim: The Man Who Loved Watching Voyeur Porn Discovers He And His Girlfriend Are Now The Main Attraction

Bess — Thu, 21 May 2026 10:12:26 +0000

06/02/2026

In the quiet Shenzhen hotel room, Eric and his girlfriend Emily thought they were sharing nothing more than a private escape.

They unpacked their bags, settled in for the night, and let the world fade away as they moved together in the kind of unguarded intimacy that most couples assume stays locked behind closed doors. But no. Unknown to them at the time, they were actually the stars of explicit content, streamed for paying subscribers and accessible to strangers worldwide.

Eric only realized this months later.

He was scrolling through the same Telegram channel he had long used to browse pornography, and suddenly he froze.

There they were, entering the room, setting down their luggage, undressing, and then having sexual intercourse, captured in crisp, unfiltered detail, then edited into an hour-long clip broadcast to thousands of strangers.

Eric and his girlfriend Emily. The couple have become victims of underground voyeurism in China. Traumatized, they wear hats in public out of fear of recognition.

"What drew me in is the fact that the people don't know they're being filmed," Eric later recalled of his old (and bad) habit. "I think traditional porn feels very staged, very fake."

Eric acknowledged, and confessed, that he was an avid viewer of this kind of voyeurism, and never knew that he and his girlfriend would become the victims themselves.

Now, realizing that he and his girlfriend are on the other side of the camera, the thrills had curdled into horror.

When he told Emily about this, her first thought that he was joking.

But after she watched the footage herself, she too was mortified. Emily became terrified that colleagues or family might stumble across it. The couple didn't speak to each other for weeks afterward.

When they finally reconciled, they went out wearing hats, constantly scanning faces in crowds, wondering who might recognize their most vulnerable moments turned into public entertainment.

Telegram is the go-to platform for this underground spy-cam trade and similar illicit content networks in China for a combination of practical, technical, and regulatory reasons.

What makes this harrowing is the fact that the footage wasn't a one-off hack or a bitter ex's revenge.

It was a snapshot of a sprawling, industrial-scale spy-cam operation thriving inside China's hotel industry, even though producing or distributing pornography has been strictly illegal there for years. An in-depth BBC investigation, published in February 2026, peeled back the layers on this underground trade, revealing at least 180 hotel-room cameras feeding live streams and archived clips across six different websites and apps.

Many of these operations are heavily promoted through Telegram groups, some swelling to as many as 10,000 members.

"Mask Park Treehole Forum" was one of the most prominent. The illicit Telegram chat channel that operated until it was exposed and shut down had over 100,000 members, predominantly Chinese-speaking men.

Subscribers in those Telegram chats aren't passive; they build entire communities around the thrill, trading tips on the best feeds, celebrating "good" performances, and bonding over the shared secret. It's a dark mirror of how mainstream social media works, except the currency here is stolen intimacy.

The cameras themselves are ingeniously discreet: some no larger than a pencil eraser, wired professionally, and directly into the room's electrical system, and set to activate the moment a guest inserts their key card into the power slot.

One journalist tracked a device to a hotel in Zhengzhou, central China, hidden inside a wall ventilation unit with its lens aimed squarely at the bed. When the team disabled it, word rippled instantly through the Telegram channel.

"Zhonghua got taken down!" one subscriber wrote, referring to the camera by its code name.

The operator, known online as "AKA," replied: "It’s such a shame; that room has the best sound quality!"

But the mourning was short-lived.

Within hours, AKA posted that a replacement camera in another hotel was already live. "This is the speed of… [our livestreaming platform]," he boasted to his subscribers. "Impressive right?"

Photos shared online through the Telegram group "MaskPark Treehole Forum".

The mechanics are ruthlessly efficient.

Subscribers pay a monthly fee, which is around 450 yuan, or about $65, in order to access multiple live feeds showing several hotel rooms at once. Subscribers can rewind the streams from the moment the electricity kicks in or download edited clips from vast archives stretching back to 2017, one of which held more than 6,000 videos.

In the live chats, the audience doesn't just watch in silence; they gossip in real time, rating bodies, making lewd comments, speculating about the guests' conversations, and cheering like sports fans when a couple starts having sex.

Lights off draws groans of complaint.

Women in the footage are routinely described with crude slurs, like "sluts," "whores," "bitches."

This so-called business is profitable, and AKA, one of roughly a dozen agents the BBC identified, wasn't working alone. He answered to "camera owners" higher up the chain, the people who install the devices and run the platforms.

Based on membership numbers and fees, the BBC estimates AKA alone pulled in at least 163,200 yuan, or about $22,000, in just a few months. This is about four times China's average annual income of 43,377 yuan. I

t’s a business model built on volume, low overhead, and the relentless churn of new victims.

Hidden cameras, carefully positioned, capturing exactly what strangers are willing to pay for.

The demand driving this ecosystem isn’t some niche fetish confined to the shadows of Chinese social media. It’s a lucrative slice of the global adult entertainment industry, which continues to balloon in value even as laws tighten in certain regions.

Recent market analyses peg the digital adult content sector at around $44 billion to $70 billion in 2025-2026, with projections climbing toward $75 billion or even $200 billion by the early 2030s, fueled by subscription models, mobile access, and ever-more-personalized feeds. But what sets voyeur or "spy-cam" content apart is its promise of raw authenticity.

While studio-produced porn often feels polished and performative, and that they create scripted moans, perfect lighting, actors and actresses who know exactly where the cameras are. Hidden-camera footage sells the opposite illusion: unscripted, unaware participants going about the most private acts of their lives.

That sense of forbidden intrusion taps into something primal: a mix of curiosity, power, and the adrenaline rush of transgression.

Your browser does not support the video tag.

Long story short, what keeps the wheels spinning is the market's insatiable appetite for the "real."

Traditional porn may satisfy some, but voyeur content delivers a different high: the knowledge that these are ordinary people: flawed bodies, awkward moments, genuine reactions, and the fact that they're unaware that they're being watched performing

On major tube sites, hidden-camera and voyeur categories have ranked among the most searched for years, drawing hundreds of millions of views monthly.

The hardware side has exploded alongside it. Miniature spy cameras, disguised as smoke detectors, chargers, or air fresheners, sell for pennies on electronics markets like Shenzhen's Huaqiangbei, where anyone with basic know-how can stock up and turn a quick profit.

Global projections for the broader spy-camera and surveillance gadget market hover around $300 million to $550 million in the coming decade, but the real money is in the content it generates.

China’s operation stands out for its seamless integration of supply, distribution, and monetization, but the problem is hardly isolated.

Similar spikes in voyeurism crimes have been reported worldwide, including in the West, as well as neighboring countries, like South Korea, where the term "molka" has become an endemic. In many of these countries, the police had logged thousands of incidents in homes, Airbnbs, changing rooms, and public spaces, including toilets.

Victims span the spectrum: couples mid-act, women in showers, families changing clothes, and even inside toilet bowls.

In other words, what are supposed to be normal acts of human beings, assumed to be performed alone or with a trusted partner, are now objectified for strangers' entertainment.

A screenshot of one of the possibly countless of 'molka' voyeurism videos depicting women inside public toilet cubicles in South Korea.

And here, Eric and Emily aren't the only victims. They're far from alone, but are part of only a handful of people who shared their story.

While tech platforms bear heavy responsibility, and that Telegram insists it moderates non-consensual content and bans it explicitly, yet channels like these persist for months, sometimes years, raking in fees before any action is taken. If it's taken at all.

As for the hotels, they are known for their slow response, probably due to insiders who profit from this kind of business, despite new Chinese regulations introduced in April 2025 requiring regular sweeps for hidden devices.

Since enforcement lags, replacements go up within hours, and budget properties remain especially vulnerable, travelers are always at risk.

They too have created communities to share their survival tips: like urging others to sweep rooms with phone flashlights to catch lens glints, probe ventilation grates and outlets, download RF detectors or Wi-Fi-scanning apps, or simply avoid unverified spots in high-risk cities. Some couples even pitched the idea of setting up a tent in their rooms, or book hotels only at chains with verified privacy protocols.

Portable privacy screens and white-noise machines have become unexpected travel essentials.

Read: How 'Molka' Voyeurism Through Spy Cameras Has Become 'A Part Of Daily Life' In South Korea

The human cost is the part that rarely makes it into the glossy market reports.

For every Eric who once consumed this content and now recoils from it, there are countless others whose lives are quietly upended. Bodies that were meant to be shared only with a partner become public property, turned into fapping material, dissected in comment threads, replayed for profit long after the guests have checked out.

The industry literally doesn't care about consent or consequences. After all, business is business, and all it cares about is retention and recurring revenue.

It’s worth noting that, more often than not, the footage was recorded by men. They were frequently the ones uploading the videos online as well, distributing them across platforms like Telegram.

Telegram itself is blocked by the Chinese government, meaning users in China typically need a VPN to access the platform.

"This has heightened the concerns of many women, as voyeuristic incidents seem to be ubiquitous," said Huang Simin, a Chinese lawyer who specializes in sexual violence cases.

Vivaldi Introduces Version 8.0, Which Makes The Browser Feels Like A Different Kind Of Browser

Irvan — Wed, 20 May 2026 14:10:12 +0000

Web browsers, the gateways to the web, have been evolving since the first time they were introduced.

While major browsers like Chrome, Firefox, Edge and Safari command huge percentage of the market, Vivaldi standa apart primarily through its extreme focus on customization and built-in productivity tools designed for power users. While Chrome and Edge prioritize simplicity and integration with their respective ecosystems, Vivaldi treats the browser as a highly adaptable workspace.

Users can rearrange almost every element of the interface, including toolbars, buttons, tab positions, and panels, without relying on extensions.

This level of control extends to themes, layouts, and behaviors that adapt to individual workflows, something the major browsers offer only in limited ways.

Vivaldi also boasts tab management, privacy features, built-in apps, and more. And since it's a Chromium-based browsers, Vivaldi delivers full compatibility with Chrome extensions and modern web standards, but it avoids the data collection practices common in Google Chrome and Microsoft Edge.

And now, comes version 8.0, which is an update that looks more like a huge upgrade.

New Vivaldi just dropped.
And by new, we mean NEW pic.twitter.com/9LNRJnhcjr

— Vivaldi Browser (@vivaldibrowser) May 21, 2026

First off, Vivaldi version 8.0 for desktop brings its most substantial visual update in years.

According to the announcement, the browser now uses a Unified design approach where toolbars, tabs, panels, and other elements sit on a single continuous surface rather than in separate layered regions. This creates a more cohesive appearance with precise alignment and intentional spacing that makes the interface easier to navigate and read.

Themes integrate more deeply across the entire window. Backgrounds and colors flow without interruption from the tab bar through panels and edges. Translucency and blur effects blend wallpapers into the environment more naturally. New default themes include options like Zen, Soria Moria, Sunset Forest, and Kawaii Clouds, alongside refreshed light and dark variants.

Users with custom themes can choose the new Unified coloring mode or stick with previous styles through the theme editor settings.

One of the several themes introduced in Vivaldi 8.0.

Onboarding and layout options have expanded to help both new and returning users.

Six preset layouts are available right at startup and remain accessible later in Settings under Appearance.

Simple provides a clean top-tab experience with minimal distractions. Classic retains the familiar toolbar and panel arrangement now updated with the Unified look.

Vertical Left and Vertical Right shift tabs and address bar to the sides for better use of wide screens. Bottom places navigation at the lower edge, while Auto Hide keeps the interface out of view until the cursor reaches a screen edge, maximizing content space.

Vivaldi 8.0 introduces six preset layouts, available right from onboarding.

These layouts serve as starting points that users can further customize using Vivaldi's extensive tools. Tab management remains powerful with stacking, tiling, and grouping options that adapt to individual workflows. Panels, workspaces, and built-in features like mail and calendar stay fully accessible and integrated.

The update also includes backend improvements such as a rewritten tabs management system, combined site preferences and tracker blocker dialogs, and refinements to translation and permission handling.

Since Chromium itself has been updated to version 148 for better compatibility and security, Vivaldi is able to bring various fixes address mail functionality, tab behaviors, auto-hide interactions, and platform-specific issues across Windows, macOS, and Linux.

The result is a browser that feels more unified in structure while preserving the depth of customization that has defined Vivaldi over time.

While AI dominates the headlines and social media consumes ever more of people's attention, Vivaldi 8.0 aims to make the good old web feel "worth coming back to."

https://t.co/AJ1hcxEUir

— Vivaldi Browser (@vivaldibrowser) May 21, 2026

Published:

21/05/2026

Sophie Rain Allegedly Turning Down A $15 Million Offer For Her Virginity Reveals The Billion Dollar Economics Of Internet Fantasy

Bess — Tue, 19 May 2026 10:44:37 +0000

19/05/2026

Sophie Rain has built an internet empire not only from physical attraction or character, but also on contradiction.

The 21-year-old Florida creator became one of the most talked about names on OnlyFans after repeatedly claiming to be the most highly-paid creator, while still a virgin. Her statements transformed what could have been a private personal detail into the foundation of a massively profitable online brand.

Now, she says a billionaire allegedly offered her $15 million for the rights to her first time, an offer she claims she rejected immediately because she has "ethics and morals" to uphold.

The number sounds outrageous, but the business logic behind it is not.

Sophie understands the economics of scarcity better than most creators online.

In a digital landscape oversaturated with explicit content and unlimited access, rarity becomes valuable. The less accessible something feels, the more audiences obsess over it.

Sophie Rain, born Izabella Blair on September 22, 2004, is an American internet personality, OnlyFans creator, and co founder of the viral creator collective Bop House.

Raised in Florida, she first gained mainstream attention in 2024 after claiming she earned more than $43 million during her first year on OnlyFans while still being a virgin and a devout Christian. Before becoming one of the platform's most recognizable creators, she reportedly worked minimum wage restaurant jobs and has spoken publicly about growing up on food stamps.

Her online persona, built around the contradiction between religion, sexuality, exclusivity, and internet fame, quickly turned her into one of the most controversial and commercially successful figures in the modern creator economy.

That is why Sophie Rain is not simply selling explicit content.

She is selling anticipation, fantasy, and the illusion that some part of her remains unattainable. Her virginity is no longer just a personal detail. It has become part of the product itself.

The mystery surrounding her identity is what keeps people subscribed, debating, reposting clips, making reaction videos, and emotionally invested in her story.

As long as audiences continue believing that narrative and attaching value to it, the market value of her brand continues to grow.

View this post on Instagram

A post shared by Sophie Rain (@sophieraiin)

That is what makes the alleged $15 million offer so fascinating. It was not merely an offer for sex. It was an offer to destroy the very narrative that helped make her one of the most profitable creators on the internet.

From a purely financial perspective, accepting it may have secured instant wealth, but long term, it could have damaged the mythology surrounding her brand. Sophie herself has acknowledged that her virginity plays a major role in why fans subscribe to her content, and reports about enormous offers tied to it have circulated online for months.

In many ways, this is influencer marketing pushed to its absolute extreme.

Traditional influencers sell products through aspiration. Sophie sells identity through contradiction. She presents herself as religious, reserved, morally selective, and unattainable while operating on one of the internet’s most sexualized platforms. That tension fuels endless conversation. Every debate over whether she is genuine or performing for attention only strengthens the brand further.

The modern creator economy no longer revolves solely around content itself.

Fans are paying for emotional investment, perceived intimacy, direct interaction, and participation in an ongoing narrative. Sophie appears to understand this better than most creators in the industry. The longer the mystery survives, the more profitable it becomes.

why is it so hard for people to believe im a virgin https://t.co/aCxvX1F6Oy

— Sophie Rain (@sophieraiin) May 19, 2026

That is also why her comments about ethics and morals resonated online.

Whether people believe her claims or not almost becomes secondary. Morality itself becomes part of the marketing structure. In an era where oversharing dominates internet culture, withholding something becomes powerful. The refusal becomes content. The boundary becomes content. Even the speculation becomes content.

Ironically, Sophie's strategy almost exists as the complete opposite of creators like Bonnie Blue and Lily Phillips, whose brands exploded through excess, shock value, and the constant escalation of explicit content. Where Bonnie and Lily built attention through pushing limits further and further in public, Sophie built hers through withholding.

One strategy thrives on overexposure. The other thrives on restraint.

Yet all three creators reveal the same reality about the modern internet economy. Attention is currency, and personal identity has become the product.

View this post on Instagram

A post shared by The Iced Coffee Hour Podcast (@icedcoffeehour)

Virginity, relationships, breakups, pregnancies, friendships, and even mental health struggles have increasingly become monetized storylines in influencer culture. The line between personal identity and commercial strategy has nearly disappeared. What once belonged to private life is now repackaged into engagement, virality, and profit for an audience constantly demanding something more intimate and emotionally revealing.

Reports surrounding Sophie's earnings have only amplified that phenomenon. She has claimed to earn close to a hundred million dollars through OnlyFans, while online discussions continue suggesting that individual subscribers have spent staggering amounts on her content alone.

Whether those numbers are fully accurate or partially exaggerated almost becomes irrelevant. The attention itself is the product.

What separates Sophie Rain from countless other creators is that she understands restraint can often be more profitable than unlimited access. In an online economy flooded with explicit content, suggestion can become more valuable than revelation. Mystery creates anticipation, anticipation creates obsession, and obsession keeps audiences invested far longer than instant gratification ever could.

That may ultimately be the real reason she allegedly turned down the $15 million offer. Not necessarily because the amount itself was unbelievable, but because ending the narrative could potentially destroy the very thing that made her so valuable in the first place.

View this post on Instagram

A post shared by The Iced Coffee Hour Podcast (@icedcoffeehour)

Whether Sophie was protecting her personal values, preserving her self respect, or simply safeguarding the long term power of her brand, she understands something many creators eventually learn too late.

In the attention economy, mystery is one of the most expensive assets a person can own.

View this post on Instagram

A post shared by The Iced Coffee Hour Podcast (@icedcoffeehour)

8-Year-Old Spots NASA Website Typo That Went Unnoticed For Over A Year

Danny Krisnanto — Tue, 19 May 2026 09:46:35 +0000

19/05/2026

NASA's website at nasa.gov is one of the most impressive and high-traffic government websites in the world.

It's extremely popular, and regularly pulls in tens of millions of visits per month (over 95 million in April 2026 alone). That's more traffic than many major news or tech sites. Since it's regularly updated too, it always has been a go-to for reliable space news, stunning images, mission updates, educational resources, and raw scientific data.

In other words, the website is considered one of the most trusted and authoritative sources for space-related information.

In the era where the choir of content writing has been elevated with the help of AI, NASA shows that behind the massive popularity of the website, there are humans who can still make mistakes.

And an 8-year-old boy was the one who realized this.

hey @NASA

my 8-year old just found a spelling mistake on your website.

he insists this is a "critical mission update" :) pic.twitter.com/QD4QQFm4sP

— musarat (@musaratali) May 18, 2026

The post came from X user @musaratali, who shared a screenshot from NASA's Solar System page showing the Jupiter section. A link shows "Expore Jupiter" instead of the correct "Explore Jupiter," with the missing "L" neatly circled in red. The parent noted that their child insisted this was a "critical mission update," and the post quickly gained huge engagement.

NASA responded with perfect humor and humility. Their official reply read: "Oops! Thanks for pointing it out to us. We're on our way to retrieve that missing letter."

The space agency fixed the typo within the hour.

Oops! Thanks for pointing it out to us. We're on our way to retrieve that missing letter.

— NASA (@NASA) May 18, 2026

The replies flooded in with jokes about Pluto's planetary status, budget concerns from the eight-year-old's discovery, and appreciation that NASA is still run by real humans rather than flawless AI.

This small incident highlights exactly why nasa.gov feels special among major government and science websites. It draws tens of millions of visitors each month for stunning images, live mission data, and reliable information, yet it still has room for a simple human error and the personality to laugh about it.

In an era of polished corporate content, NASA's quick, witty response reminded everyone that space exploration remains a human endeavor filled with curiosity, mistakes, and quick fixes.

The real star of the story is that sharp-eyed eight-year-old, who now knows that paying attention to details matters, even at NASA. Somewhere out there is a future astronaut who just completed his first official mission: retrieving a missing letter from Jupiter.

Moments like this make NASA’s site one of the most human and engaging major science platforms on the internet.

It's worth noting that the typo had been around for more than a year, and can be traced back to at leat 2024, when the page's design and content were last updated.

NASA corrected the typo in about an hour after the original post.

Further reading: One Tweet, One Reply, One Lesson: The NASA Internship That Went Viral On Twitter