Physicists Discover Universal Law Governing How Objects Fracture

Sure! Here’s a rewritten version of the content while preserving the HTML tags:

How many pieces can a dropped vase break into?

Imaginechina Limited / Alamy

The physics behind a dropped plate, a crumbled sugar cube, and a shattered glass shows striking similarities regarding how many pieces result from each object breaking.

For decades, researchers have recognized a universal behavior related to fragmentation, where objects break apart upon falling or colliding. If one counts the fragments of varying sizes and plots their distribution, a consistent shape emerges regardless of the object that is broken. Emmanuel Villemaux from the University of Aix-Marseille in France has formulated equations to illustrate these shapes, thereby establishing universal laws of fragmentation.

Instead of concentrating on the appearance of cracks leading to an object’s breakup, Villermaux employed a broader approach. He considered all potential fragment configurations that could result in shattering. Some configurations produce precise outcomes, such as a vase breaking into four equal parts; however, he focused on capturing the most probable set that represents chaotic breakage, namely the one with the highest entropy. This mirrors methods used to derive laws concerning large aggregates of particles in the 19th century, he notes. Villermaux also applied the principles of physics that govern changes in fragment density during shattering, knowledge previously uncovered by him and his colleagues.

By integrating these two elements, they succeeded in deriving a straightforward equation that predicts the size distribution of fragments in a broken object. To verify its accuracy, Villermaux compared it against a number of earlier experiments involving glass rods, dry spaghetti, plates, ceramic tubes, and even fragments of plastic submerged in water and waves crashing during stormy weather. Overall, the fragmentation patterns observed in each of these experiments conformed to his novel law and reflected the universal distribution shapes previously noted by researchers.

He also experimented by dropping objects from varying heights to crush sugar cubes. “This was a summer endeavor with my daughters. I had done it a long time ago when they were young, and later revisited the data to further illustrate my concept,” Villermaux explains. He observes that this equation fails to hold when randomness is absent, or the fragmentation process is overly uniform, as occurs when a liquid stream divides into uniform droplets based on the deterministic rules of fluid dynamics, or in instances when fragments engage with each other during fragmentation.

Mr. Ferenc and his colleagues at the University of Debrecen in Hungary argue that the graphical pattern highlighted in Villermaux’s analysis is so fundamentally universal that it may derive from a more extensive principle. Simultaneously, they express surprise at how broadly applicable it is, as well as its adaptability to accommodate specific variations, such as in plastics where cracking can be “healed.”

Fragmentation is not merely a captivating challenge in physics; a deeper understanding could significantly impact energy expenditures in mining operations or guide preparations for increasing rockfalls in mountainous areas as global temperatures tend to rise, Kuhn remarks.

Looking ahead, it may prove beneficial to explore not only the sizes of the fragments but also their shape distributions, suggests Kuhn. Additionally, identifying the smallest conceivable size of a fragment remains an unresolved issue, according to Villermaux.

Topic:

Source: www.newscientist.com

Antibody Cocktails Could Serve as a Universal Treatment for Influenza

SEI 265484890

Illustration of an antibody targeting influenza virus particles

Science Photo Library/Alamy

Antibody cocktails may provide innovative strategies to tackle emerging strains that lead to seasonal flu and pandemics. While effective in shielding mice from a variety of influenza strains, these cocktails have yet to undergo testing in humans.

Conventional treatments and vaccines for influenza typically aim to stimulate the production of proteins known as neutralizing antibodies. These antibodies attach to specific virus strains and prevent the infection of cells. Though such medical strategies can be quite effective, they often require months for development and may become ineffective due to viral mutations. This explains the seasonal updates to influenza vaccines and the ongoing efforts for a universal vaccine that could guard against all flu variants or even a broader range of viruses.

Silke Paust at The Jackson Institute in Farmington, Connecticut, alongside her team, is exploring an alternative route. Their focus is on non-neutralizing antibodies—another type of protein that the immune system produces. Although these proteins have been largely overlooked for infection control, they empower the immune system to eliminate the virus by marking infected lung cells.

“We’re not just a vaccination; we aim to treat them. Our goal is to develop medications that can avert severe illness and fatalities, either as a preventive measure or therapeutically after infection,” Paust explains.

Paust and her research team investigated antibodies that target influenza virus proteins in a specific region termed M2E.

The researchers carried out a series of experiments assessing the efficacy of antibodies, both singularly and in combinations, on mice infected with the flu virus, discovering that a combination of three antibodies yielded the most promising results.

They evaluated antibody cocktails on mice exposed to two H1N1 strains, including the ones responsible for the 2009 swine flu pandemic. Currently circulating H1N1 alongside two avian strains: H5N1, which affects wildlife and livestock worldwide, and H7N9, which poses a significant threat to humans and other animals.

The findings indicated that the antibody cocktails diminished the severity of lung disease and reduced viral loads, leading to improved survival rates in both healthy and immunocompromised mice.

For instance, when treated with antibody cocktails within the first three days post-exposure to H7N9, all mice survived; 70% of those treated on day four survived, and 60% did on day five.

Paust highlighted this as a groundbreaking moment, noting it marked the first instance of widespread influenza protection in living subjects. The cocktail also proved effective when administered before infection, suggesting potential preventative uses.

Even after 24 days of treatment, there were no indicators of the virus mutating to develop resistance. “For the virus to evade treatment, it would need to avoid all three antibodies, which bind in different ways,” Paust states.

“This demonstrates the potential for using antibody cocktails to treat individuals during flu pandemics, in conjunction with vaccines,” says Daniel Davis from Imperial College London. “However, further testing in humans is crucial before considering this a true medical advancement.”

Paust’s next step involves modifying the antibodies aimed at M2E to resemble human proteins. This has been done with numerous antibodies in the past. If successful, the process will proceed to safety and efficacy evaluations.

Paust envisions a future where these antibody cocktails could be stockpiled as drugs to tackle seasonal flu outbreaks. “Ideally, this would be administered to high-risk individuals at the onset of the season,” she concludes. “This would ensure they remain relatively healthy.”

topic:

Source: www.newscientist.com

Universal Detectors Identify AI Deepfake Videos with Unprecedented Accuracy

Deepfake video showcasing Australian Prime Minister Anthony Albanese on a smartphone

Australia’s Associated Press/Alamy

Universal DeepFake Detectors have demonstrated optimal accuracy in identifying various types of videos that have been altered or entirely produced by AI. This technology can assist in flagging adult content, deepfake scams, or misleading political videos generated by unregulated AI.

The rise of accessible DeepFake Creation Tools powered by inexpensive AI has led to rampant online distribution of synthetic videos. Numerous instances involve non-consensual depictions of women, including celebrities and students. Additionally, deepfakes are utilized to sway political elections and escalate financial scams targeting everyday consumers and corporate leaders.

Nevertheless, most AI models designed to spot synthetic videos primarily focus on facial recognition. This means they excel in identifying a specific type of deepfake where a person’s face is swapped with existing footage. “We need a single video with a manipulated face and a model capable of detecting background alterations or entirely synthetic videos,” states Rohit Kundu from the University of California Riverside. “Our approach tackles that particular issue, considering the entire video could be entirely synthetically produced.”

Kundu and his team have developed a universal detector that leverages AI to analyze both facial features and various background elements within the video. It can detect subtle signs of spatial and temporal inconsistencies in deepfake content. Consequently, it identifies irregular lighting conditions for people inserted into face-swapped videos, as well as discrepancies in background details of fully AI-generated videos. The detector can even recognize AI manipulation in synthetic videos devoid of human faces, and it flags realistic scenes in video games like Grand Theft Auto V, independent of AI generation.

“Most traditional methods focus on AI-generated facial videos, such as face swaps and lip-synced content.” says Siwei Lyu from Buffalo University in New York. “This new method is broader in its applications.”

The universal detector reached an impressive accuracy rate of 95% to 99% in recognizing four sets of test videos featuring manipulated faces. This performance surpasses all previously published methods for detecting this type of deepfake. In evaluations of fully synthetic videos, it yielded more precise results than any other detectors assessed to date. Researcher I presented their findings at the 2025 IEEE Conference on Computer Vision and Pattern Recognition in Nashville, Tennessee, on June 15th.

Several researchers from Google also contributed to the development of these new detectors. Though Google has not responded to inquiries regarding whether this detection method would be beneficial for identifying deepfakes on platforms like YouTube, the company is among those advocating for watermarking tools that help label AI-generated content.

The universal detectors have room for future enhancements. For instance, it would be advantageous to develop capabilities for detecting deepfakes utilized during live video conference calls—a tactic some scammers are now employing.

“How can you tell if the individual on the other end is genuine or a deepfake-generated video, even with network factors like bandwidth affecting the transmission?” asks Amit Roy-Chowdhury from the University of California Riverside. “This is a different area we’re exploring in our lab.”

Topics:

Source: www.newscientist.com

Disney and Universal File Lawsuit Against AI Image Creator Midjourney for Copyright Infringement

Disney and Universal have filed a lawsuit against an artificial intelligence company, claiming copyright violations. The entertainment titans have described the image generator behind Midi Johnny’s popular AI as a “bottomless pit of plagiarism,” alleging it replicates the studios’ most iconic characters.

The lawsuit, lodged in federal court in Los Angeles, accuses Midi Joan of illegally accessing two Hollywood studio libraries and creating numerous unauthorized copies of key characters, including Darth Vader from Star Wars, Elsa from Frozen, and Minions from Despicable Me. Midjourney has not yet commented on the matter.

This legal action from Disney and Universal marks a new chapter in the ongoing battle over copyright issues related to artificial intelligence, following prior lawsuits focusing on text and music. So far, these two companies are among the largest industry stakeholders to address the implications for images and videos.

“We are optimistic about the potential of AI technology when used responsibly to enhance human creativity; however, it’s crucial to recognize that piracy and copyright infringement carried out by AI companies is unacceptable,” stated a company representative.

Kim Harris, vice-chair and legal counsel at NBCUniversal, emphasized the need to “entertain and inspire while protecting the hard work of all artists who invest significantly in content.”

The studios assert that the San Francisco-based company, one of the pioneers in AI-driven image generation, must either cease infringing upon copyrighted works or implement technical measures to prevent the creation of AI-generated images of copied characters.

Nonetheless, studios claim that Midjourney continues to release updates to its AI image service, promoting high-quality infringing images. The AI is capable of recreating animated visuals based on user prompts. These companies train their models using vast datasets, often sourced from millions of websites.

In a 2022 interview with Forbes, Midjourney CEO David Holz mentioned that he built the company’s database through extensive “internet scraping.”

The lawsuit, initiated by seven entities holding the copyrights to various Disney and Universal Pictures Film Units, includes examples of AI-generated animations with Disney characters like Yoda wielding lightsabers, as well as universal characters such as the Dragon from Kung Fu Panda, Toothless, and Shrek.

“By leveraging plaintiffs’ copyrighted materials and distributing images (and soon videos) that unmistakably incorporate beloved characters from Disney and Universal, Midi Joan exemplifies a typical copyright-free rider, creating a bottomless pit of flexible liability,” the studios claim.

Disney and Universal are seeking a preliminary injunction to prevent Midjourney from continuing to copy their works or providing image and video generation services without protective measures against infringement, as well as unspecified damages.

Founded in 2021 by David Holz, Midjourney operates on a subscription model, boasting a revenue of $300 million from its services last year alone.

Skip past newsletter promotions

This isn’t the first instance of Midjourney facing accusations of leveraging artists’ works to train AI systems. Approximately a year ago, a federal judge in California found that 10 artists, alongside Stability AI and others, were in litigation against Midjourney, alleging that these companies had copied and stored their works on their servers, rendering them potentially liable for unauthorized use. This ruling allowed the lawsuit to proceed based on misuse of images, and it is currently ongoing.

This case is part of a larger trend of lawsuits involving authors, media organizations, and record labels against high-tech firms over the utilization of copyrighted materials for AI training.

When asked whether the company sought consent from artists whose works are copyrighted, Holz remarked, “It’s practically impossible to gather 100 million images and trace their origins.” In a submission to the UK government last year, OpenAI stated, “Training today’s leading AI models without the use of copyrighted materials is unfeasible.”

In late 2023, the New York Times filed a lawsuit against OpenAI, the developer of ChatGPT, along with Microsoft (which holds a 49% stake in the startup), for allegedly misusing and regenerating text from its articles. That suit is still pending. Other media outlets, including The Guardian, have negotiated licensing agreements with AI companies to use their archives. Similarly, authors have sued Meta, claiming it used a vast database of pirated books to train the LLaMA AI model, although many of those claims were dismissed.

In June 2024, major record companies filed lawsuits against two AI companies for copyright infringement. Sony Music Entertainment, Universal Music Group Recordings, and Warner Records accused Suno and Udio of improperly using millions of songs to create a system capable of generating derivative music.

Source: www.theguardian.com

Disney and Universal Lawsuit Could Deal a Heavy Blow in the AI Copyright Battle

The Minion character originates from films produced by Universal Pictures.

Movie/Aramie

Disney and Universal have initiated a lawsuit against the AI image generator Midjourney, alleging widespread copyright infringement that enables users to produce images that “explicitly incorporate and mimic well-known Disney and Universal characters.” This lawsuit could mark a significant shift in the ongoing legal discourse surrounding AI-related copyright issues faced by book publishers, news outlets, and other content creators.

The Midjourney tool, which generates images based on textual prompts, boasts around 20 million users on its Discord platform. Users provide their input for creation.

In the lawsuit, the two film production giants provide examples where Midjourney can generate images surprisingly similar to characters it does not own rights to, like the Disney-owned Minions and characters from The Lion King. They assert that these results stem from the AI being trained on their copyrighted materials. They also contend that Midjourney “disregarded” their attempts to resolve these issues before resorting to legal action.

The complaint states, “Midjourney is a classic copyright-free rider and an endless source of plagiarism.” Midjourney has not yet issued a response to New Scientist‘s request for comment.

The lawsuit is applauded by Ed Newton Rex, a nonprofit advocate for fairer training practices within AI companies. “This is a monumental day for creators globally,” he comments. “The government has displayed unsettling tendencies toward legalizing intellectual property theft, potentially yielding to the intense lobbying from Big Tech.

Newton-Rex alleges that Midjourney engineers previously justified their actions on the grounds that the art had become “ossified.” “Fortunately, this absurd defense is unlikely to hold up in court,” he adds.

Legal experts express candid perspectives on Midjourney’s likelihood of success. “It’s Disney; thus, Midjourney is in a precarious position, please excuse my bluntness,” remarks Andres Guadams from the University of Sussex, UK.

Guadams emphasizes Disney’s resolute approach to safeguarding its intellectual property—rarely, but effectively—underscoring the necessity of this intervention. The film studio took action several months following other entities, such as news publishers, in their pursuit against AI companies for the alleged unauthorized use of their creations. Many of those disputes were resolved through licensing agreements between the AI firms and copyright holders.

“Media conglomerates are excited about potential breaches. The models have improved to such an extent that they can effortlessly create characters that come to mind,” states Guadams. He believes Disney is biding its time because “unlike publishers, they’re not simply seeking licenses to survive.”

The involvement of these two media powerhouses signals a pivotal moment at the intersection of AI and copyright, according to Guadams. “The fact that they are targeting Midjourney sends a clear message,” he states. Midjourney specializes in image generation exclusively, making it relatively small compared to major AI corporations. “This acts as a warning to larger entities, urging them to implement stronger protective measures.”

While many major AI companies incorporate image-generating features in their chatbots, they tend to impose stricter controls on users’ abilities to produce images featuring copyrighted characters through considerable limitations.

Disney, which generated $91 billion in revenue last year, is not seeking to profit from Midjourney. “This could act as a call for negotiations. Since AI is not going away, Disney may be setting a precedent for future business interactions,” notes Guadams.

Topic:

Source: www.newscientist.com

Universal Antivenom Could Emerge from a Volunteer Who Endures 200 Snake Bites

The video lasts under 2.5 minutes. A slender man with thick hair enters the room, pulling a long black mamba that can deliver a lethal bite within an hour. He bites his left arm and moments later, extends his right arm towards a Taipan from Papua New Guinea. “Thank you for watching,” he states, as he calms down to speak to the camera while his left arm bleeds before it is bandaged.

For nearly 18 years, 57-year-old Tim Friede has injected himself with over 650 precisely measured doses of venom from 16 deadly snake species to develop immunity. He has also allowed snakes—mostly one at a time, although occasionally two, as in the video—to bite him nearly 200 times.

This coincidence, also known as one of its names, could aid in addressing global health issues. With over 600 species of venomous snakes across the globe, they bite approximately 2.7 million people annually, causing around 120,000 fatalities and affecting 400,000 others.

Researchers report that in Friede’s blood, they have discovered antibodies capable of neutralizing toxins from multiple snake species. According to Journal Cell.

“I am genuinely proud to contribute something meaningful for humanity and make a difference for people 8,000 miles away.

Deforestation, urban expansion, and climate change have heightened the risk of snake bites in recent years, yet antivenom research has not kept pace with demand.

“This is a more significant issue than the first world might realize,” stated Jacob Glanville, founder and CEO of Centivax, which aims to create broad-spectrum vaccines and serves as the lead author of the study.

Dr. Glanville and his team discovered that two potent antibodies from Friede’s blood, when paired with drugs that inhibit neurotoxins, can protect mice from the venom of 19 deadly snake species from various families around the world.

Experts not involved in the research consider this an extraordinary achievement. Most antivenoms can only counteract venom from one or a few closely related snake species within a particular area.

Nicholas Casewell, a researcher at the Liverpool School of Tropical Medicine in the UK, indicates that antivenom cocktails could potentially prevent fatalities and injuries from all snake families.

“The principles from this study can be effectively applied to other snakes,” he remarked.

Friede’s journey with snakes began with a harmless bite from a five-year-old garter snake, which sparked his lifelong fascination. “If I had known what lay ahead, I would have laughed at the thought,” he reminisces with a chuckle.

However, it wasn’t until he married and started a family while working in construction that he seriously began to explore his interest in snakes. He started experimenting with scorpions around 2000 but soon transitioned to snakes. At one point, he kept 60 venomous snakes in his basement lab.

His experiments came to a sudden halt on September 12, 2001, when he was bitten by two cobras, his emotions heightened by the previous day’s terrorist attack and the recent loss of a friend. Those were his first encounters with snakes without adequate immunity built up. While he initially felt fine after the first bite, the second caused him to feel cold, his eyelids to droop, and he lost his ability to speak. He awoke four days later in a hospital after slipping into a coma.

His wife was furious, while he felt frustrated with himself. He resolved to adopt a more methodical approach, carefully measuring venom doses and timing the bites.

“I work during the day, spend time with my kids and family at home, then head downstairs to work late into the night, only to repeat it all over again the next day,” he explained.

He faced numerous accidents, including unintentional bites, anaphylactic shocks, and power outages. Friede describes himself as a self-taught scientist, asserting, “No university in the world provides training for this. I’ve done as much as I could.”

Two scientific teams studied Friede’s blood over the years, but those efforts led to little progress. By the time he met Dr. Glanville in 2017, he was ready to give up.

Dr. Glanville was investigating what scientists refer to as broadly neutralizing antibodies as a foundation for universal vaccines, inspired by his upbringing in Mayan villages in the Guatemalan Highlands. He was intrigued by the potential to apply the same strategy to develop a Universal Antivenom.

Initially, he said his “humble” goal was to find someone like a hapless snake researcher who had experienced multiple bites. His quest took a turn when he stumbled upon an article about Friede.

“I had been eagerly awaiting this opportunity,” shared Dr. Glanville, expressing gratitude towards Friede.

In partnership with Columbia University vaccine researcher Peter Kwon, Dr. Glanville isolated broadly acting antibodies from Friede’s blood and devised a combination treatment.

Tests on antibodies from Friede’s blood against the venom of 19 snake species showed that one broadly neutralizing antibody protected mice from six species. When combined with a small molecule known as Varespladib, the second antibody provided full protection against 13 snake species and partial defense against the remaining six.

Cobras and mambas produce toxins that paralyze nerve cells, while the venom from Viper family snakes damages tissue and induces bleeding in victims. Each snake species within its family produces a unique combination of dozens of toxins, which can vary by region, age, diet, and season. It can vary significantly.

Despite advances, antivenoms are still produced using methods from 130 years ago. A small amount of venom is injected into horses, camels, or sheep, and the antibodies produced in response are harvested. These antibodies are typically specific to the type of venom administered and offer minimal relief from other snake venom types.

In fact, many antivenoms can cause more severe reactions than the venom itself, as mammalian proteins can provoke fatal allergic reactions.

Researchers are seeking treatments that mitigate these side effects. Cocktail treatments involving small-molecule drugs and monoclonal antibodies targeting critical toxic families are being developed—crafted copies of human antibodies that neutralize toxins across numerous species, according to Dr. Casewell.

Future plans include testing the Australian treatment on dogs brought into veterinary clinics for snake bites. Researchers also aim to identify additional components from Friede’s blood that could expand protection across all 19 snake species under study.

However, Friede’s experimental days are over. His last bite came from a water cobra in November 2018. He has since divorced; his wife and children have moved away. “That felt like enough,” he recalled.

While he misses the thrill of interacting with snakes, he insists it was not driven by painful bites. “I might revisit this in the future,” he reflected, “but for now, I’m content with where I am.”

Source: www.nytimes.com

AI system used to detect UK benefits fraud exposed for bias | Universal Credit

The Guardian has uncovered that artificial intelligence systems utilized by the UK government to identify welfare fraud exhibit bias based on individuals’ age, disability, marital status, and nationality.

A review of a machine learning program used to analyze numerous Universal Credit payment claims across the UK revealed that certain groups were mistakenly targeted more frequently than others.

This revelation came from documents published under the Freedom of Information Act by the Department for Work and Pensions (DWP). A “fairness analysis” conducted in February of this year uncovered a significant discrepancy in outcomes within the Universal Credit Advance automated system.

Despite previous claims by the DWP that the AI system had no discrimination concerns, the emergence of this bias raises important questions about its impact on customers.

Concerns have been raised by activists regarding the potential harm caused by the government’s policies and the need for transparency in the use of AI systems.

The DWP has been urged to adopt a more cautious approach and cease the deployment of tools that pose a risk of harm to marginalized groups.

The discovery of disparities in fraud risk assessment by automated systems may lead to increased scrutiny of the government’s use of AI, emphasizing the need for greater transparency.

The UK public sector employs a significant number of automated tools, with only a fraction being officially registered.

The lack of transparency in the use of AI systems by government departments has raised concerns about potential misuse and manipulation by malicious actors.

The DWP has stated that their AI tools do not replace human judgment and that caseworkers evaluate all available information when making decisions related to benefits fraud.

Source: www.theguardian.com

White House instructs NASA to establish universal time for the moon

The White House requested NASA to establish a standard time for the moon and other celestial bodies, as the U.S. seeks to lead in space standards amidst growing competition for lunar resources among nations and private entities. The directive was issued by the director of the White House Office of Science and Technology Policy (OSTP).

NASA has been tasked with collaborating with other U.S. government agencies to create a plan by the end of 2026 to implement a timekeeping system in space, known as Coordinated Lunar Time (LTC), according to a memo obtained by Reuters.

Variations in gravity on celestial bodies, among other factors, influence the passage of time differently than on Earth. LTC serves as a reference for timing lunar missions that require extreme precision for spacecraft and satellites.

Kevin Coggins, NASA’s space communications and navigation chief, explained, “A clock that works accurately on Earth will run at a different pace on the moon.”

According to the memo from OSTP Director Arati Prabhakar, Earth-based clocks would lose approximately 58.7 microseconds per day for individuals on the Moon, leading to a further time deviation. Periodic fluctuations are also a factor.

Coggins likened the atomic clocks at the U.S. Naval Observatory to the heartbeat of the nation, emphasizing the importance of synchronization in space.

NASA’s Artemis program aims to land astronauts on the moon and establish a scientific base to aid future Mars missions. Numerous companies, spacecraft, and countries are involved in this venture.

Uniform lunar time standards are deemed crucial by OSTP officials to ensure data security between spacecraft and synchronize communications between Earth, lunar assets, and astronauts, preventing errors in mapping and location tracking.

The deployment of an atomic clock on the moon’s surface may be necessary as commercial activities expand to ensure operational coordination, transaction reliability, and logistical efficiency.

The United States, which has previously landed astronauts on the moon, faces competition from other countries like China, Japan, and India with their own lunar exploration plans.

The memo also highlights the importance of defining Coordinated Lunar Time through existing standards bodies and agreements like the Artemis Accords, emphasizing the need for international cooperation in space.

Coordinated Universal Time plays a role in implementing Coordinated Lunar Time, with the United Nations’ International Telecommunication Union defining it as an international standard.

Source: www.nbcnews.com