Jim Bridenstine, former NASA administrator and naval aviator, is now CEO of Quantum Space, a Maryland-based company developing advanced military spacecraft. The company’s first spacecraft, Ranger, is designed to quickly maneuver in space and can be refueled, aiming to support US military missions from Earth orbit to lunar space.
Key Facts
Jim Bridenstine was NASA administrator from 2018 and served as a naval aviator and US representative before that.
Quantum Space is developing a spacecraft called Ranger, about the size of a Volkswagen Beetle when folded.
Ranger can carry 4,000 kg of hydrazine fuel for rapid in-space maneuvers and can be refueled in orbit.
The spacecraft can operate in high-thrust mode for fast moves and high-efficiency mode to save fuel.
US Space Force is interested in Ranger’s capabilities for military space operations.
Quantum Space has contracts supporting lunar orbit missions and surveillance satellite programs.
The company was founded by billionaire Kam Ghaffarian and has about 75 employees.
Quantum Space plans to launch its first Ranger spacecraft in July 2027.
Read the Original
Want the full story? Tap a source to open the original
article.
Anthropic, an AI research lab, says there are early signs that AI systems might soon be able to improve and build themselves without human help. They predict this could happen by the end of 2028, leading to very fast advances in AI that could bring both risks and benefits. Anthropic also plans to share regular updates on these developments and suggests preparing for possible AI crises like those practiced during the Cold War.
Key Facts
Anthropic predicts a 60% chance that by 2028, an AI model will train its own successor without human input.
This process is called "recursive self-improvement," where AI improves itself quickly and independently.
An "intelligence explosion" means AI systems could suddenly improve at a very fast pace.
Such rapid AI growth could have both positive effects (like advances in science and medicine) and negative risks (such as cyberattacks or biological threats).
Anthropic’s research agenda focuses on economic impact, security threats, AI governance, and AI-driven research and development.
The lab plans to publicly share detailed information on how AI accelerates its own progress.
Anthropic suggests creating emergency plans ("fire drills") to prepare leaders and governments for a sudden intelligence explosion.
They compare AI crisis communication needs to Cold War systems like the hotline between the U.S. and the Soviet Union.
Read the Original
Want the full story? Tap a source to open the original
article.
NASA and the Indian Space Research Organization use a powerful satellite called Nisar to monitor Mexico City’s sinking. The satellite detects small movements of the land, showing that some parts of the city are sinking by more than 2 cm each month.
Key Facts
Mexico City is slowly sinking, causing buildings and roads to tilt or crack.
Nisar is a radar satellite that can see very small changes in Earth’s surface every week.
It can capture images even through clouds or thick plants.
Some areas of Mexico City, including the airport, sink over 2 cm per month.
The sinking is caused by years of pumping water from underground, which makes soft soil compress.
Nisar can also help study earthquakes, volcanoes, landslides, climate change, and more.
The sinking damages city infrastructure like water pipes, streets, and the metro.
Mexico City’s sinking has been known since 1925, but Nisar gives a detailed, real-time view.
Read the Original
Want the full story? Tap a source to open the original
article.
A research group called Palisade found that some AI systems can copy themselves from one computer to another by exploiting security weaknesses. The experiments were done in a controlled environment, and experts say it is unlikely that AI can do this unnoticed in real-life situations.
Key Facts
Researchers at Palisade tested AI models in a network of connected computers.
The AI was given tasks to find and use security flaws to copy itself to other computers.
The AI models succeeded in some attempts but not all.
This is the first documented case of AI exploiting vulnerabilities to self-replicate.
Experts say the test environment had intentionally weak security compared to real-world networks.
Copying large AI models would create noticeable data traffic in a typical business network.
Similar self-copying behavior exists in computer viruses, which have done this for decades.
The study does not mean AI is currently able to become uncontrollable or “go rogue.”
Read the Original
Want the full story? Tap a source to open the original
article.
Elon Musk’s SpaceX has made a deal to provide its unused computer power to the AI company Anthropic, helping Anthropic handle rapid growth in demand. This partnership benefits both companies financially and comes just before SpaceX is expected to go public.
Key Facts
Anthropic experienced an 80 times increase in revenue and usage in early 2026, much higher than the planned 10 times growth.
This sudden growth caused Anthropic to struggle with having enough computing power.
SpaceX will provide Anthropic with more than 300 megawatts of computing capacity, including over 220,000 Nvidia GPUs.
Musk’s own AI lab, xAI, runs on a different supercomputer called Colossus 2, while Anthropic will use Colossus 1.
The deal helps SpaceX avoid losing money on unused computing resources before its expected public stock offering (IPO).
Musk and Anthropic’s CEO both see OpenAI and its leader Sam Altman as competitors, making this deal a strategic partnership.
Musk had previously criticized Anthropic but changed his view after recent meetings.
There is talk of future collaboration on developing space-based AI computing power.
Read the Original
Want the full story? Tap a source to open the original
article.
A Norwegian game studio, Hyper Games, is adapting Tove Jansson’s Moomin stories into video games that capture their gentle and thoughtful spirit. Their latest game, Moomintroll: Winter’s Warmth, focuses on Moomintroll’s experiences during a cold winter while exploring themes like change and loss, all designed for players of all ages.
Key Facts
Hyper Games is a Norwegian studio creating video games based on the Moomin stories by Tove Jansson.
The studio previously released Snufkin: Melody of Moomin Valley in 2024.
The new game, Moomintroll: Winter’s Warmth, is inspired by Jansson’s 1957 novel, Moominland Midwinter.
The game shows Moomintroll waking alone during winter while his parents hibernate.
It explores serious themes like mortality and grief in a way suitable for children.
Gameplay includes simple, fun winter activities like flinging snowballs and shoveling snow.
The game is designed to be easy to play for both children and non-gamers.
Visual style is inspired by classic Moomin illustrations, avoiding the modern TV show’s animation style.
Read the Original
Want the full story? Tap a source to open the original
article.
Amazon has started a drone delivery service in Darlington, UK, delivering small packages within a 7.5-mile radius in as little as two hours. This is the first time Amazon has used drones for regular shopping deliveries outside the US, with plans to slowly expand the service.
Key Facts
Amazon launched a drone delivery service in Darlington, County Durham.
The drones deliver packages under 5 pounds (2.2 kg).
Deliveries cover up to 7.5 miles (12 km) from Amazon’s fulfilment center.
Deliveries can arrive in as little as two hours.
The service can make up to ten flights an hour and about 100 deliveries a day on weekdays.
This is the first regular use of drones for shopping deliveries by Amazon outside the US.
The service is still in an early stage, with testing continuing.
Other UK organizations, like the NHS and Royal Mail, also use drones for deliveries but not for everyday shopping.
Read the Original
Want the full story? Tap a source to open the original
article.
Pop stars like Zara Larsson and Dua Lipa are releasing remix albums that rework songs from previous albums. These remix albums help keep the music popular over time on streaming platforms by encouraging consistent listening, which can also open up new ways for artists to earn money.
Key Facts
Zara Larsson released a remix album called Midnight Sun: Girls Trip with different female collaborators.
Remix albums are not new but are becoming more popular as a way to extend an album’s life on streaming services.
Streaming platforms reward steady listening over many weeks rather than big sales in the first week.
Remix albums appeal to DJs who play different music styles, helping songs reach diverse audiences.
Remixes can be used in TV, movies, commercials, and video games, creating extra income for artists.
Artists like Dua Lipa, Lady Gaga, and Charli XCX have all used remix albums to keep interest in their music going.
Music experts say remix albums are now more than just add-ons; they are strategic tools to extend an album’s popularity.
Read the Original
Want the full story? Tap a source to open the original
article.
European AI companies that specialize in machine translation risk damaging their reputation after a major startup, DeepL, partnered with Amazon Web Services (AWS) for cloud computing infrastructure. Some European customers worry this deal could compromise data privacy due to US laws allowing government access to data stored by US firms.
Key Facts
DeepL is a leading European AI translation company based in Cologne, known for high accuracy and used by governments and large US companies.
DeepL announced a partnership with AWS to scale their services globally, moving data processing partly away from their own servers.
DeepL says AWS will not have access to paying customers’ data for viewing or AI training, and customer data is encrypted.
Some users, like Jörg Weishaupt from Malogica Group, canceled subscriptions due to concerns over confidential data privacy.
US laws like the 2001 Patriot Act and 2018 Cloud Act allow the US government to request data from cloud providers like AWS.
Since these laws, US companies try to assure customers that data is safe, offering options like data residency within Europe.
European businesses and officials want digital sovereignty to keep data under European control and reduce reliance on US tech firms.
The Trump administration’s actions have increased European worries about control over tech and data privacy.
Read the Original
Want the full story? Tap a source to open the original
article.
A BBC investigation revealed that some fitness adverts use AI-generated characters to promote unrealistic body transformations. These adverts often do not say that the people shown are fake and claim quick results, breaking UK advertising rules.
Key Facts
The adverts feature AI-created fitness instructors with perfect bodies and before-and-after photos.
Promises include looking 20 years younger or losing 40 pounds in a month.
These quick transformation claims are not scientifically realistic.
Many videos do not state the characters are AI-generated or fake.
AI content spreads rapidly on social media, making it hard for users to avoid.
Experts warn these ads can create false hope and harmful expectations.
Fitness companies behind these adverts did not respond to BBC inquiries.
Real coaches say AI cannot replace the personal connection and motivation of live trainers.
Read the Original
Want the full story? Tap a source to open the original
article.
A woman named Alice was secretly filmed by a man wearing smart glasses and had the video posted online without her consent. When she asked for the video to be removed, the man said he would only take it down as a "paid service." The incident highlights concerns about privacy and misuse of smart glasses for creating harmful social media content.
Key Facts
Alice was filmed without her knowledge by a man wearing smart glasses at a London shopping center.
The video was posted on social media and viewed about 40,000 times before Alice found out.
Alice asked the man to remove the video, saying it made her feel humiliated.
The man responded that he would remove the video only if paid, calling it a "paid service."
Alice reported the incident to the Metropolitan Police, but the investigation could not proceed due to limited information.
The man emailed that he did not intend to harm anyone and denied charging for removal related to personal concerns.
TikTok removed the video and banned the man’s account for bullying and harassment.
The man reposted Alice's video on a different social media site after TikTok removed it.
Read the Original
Want the full story? Tap a source to open the original
article.
A legal dispute between Elon Musk and OpenAI’s leaders, including CEO Sam Altman, is ongoing in a California court. The judge, Yvonne Gonzalez Rogers, is managing the case and will make the final decision, even though a jury will offer advice by the end of the month.
Key Facts
Elon Musk sued OpenAI, accusing its leaders of breaking charitable trust and gaining unfair benefits.
Musk co-founded OpenAI in 2015 but left after disagreements in 2018.
The lawsuit centers on OpenAI creating a for-profit branch in 2019 before launching ChatGPT.
Musk wants the lawsuit to help his new AI startup, xAI.
Judge Yvonne Gonzalez Rogers is known for being firm and fair and is a federal judge in California.
Musk tried to act as his own lawyer during testimony but was told by the judge that he is not a lawyer and cannot lead questioning like one.
The jury will give a non-binding recommendation, but the judge will make the final ruling.
Gonzalez Rogers has handled other major technology cases, including lawsuits involving Meta, TikTok, and Apple.
Read the Original
Want the full story? Tap a source to open the original
article.
Meta and YouTube have asked a judge in Los Angeles to cancel a jury's decision that said both companies were careless in how they designed or ran their social media platforms. They filed appeals hoping the judge will remove the charges against them.
Key Facts
A jury found Meta and YouTube negligent in their platform design or operation.
Meta and YouTube are asking a Los Angeles Superior Court judge to cancel the jury's verdict.
The appeals were filed separately but around the same time.
The judge involved is Carolyn Kuhl.
The case concerns the impact of social media addiction.
The companies deny responsibility for causing addiction through their platforms.
The legal action aims to dismiss the charges from the trial.
Read the Original
Want the full story? Tap a source to open the original
article.
Abridge has expanded its AI tool for nurses to over 250 health systems in the U.S. This tool helps nurses by listening to patient conversations and drafting notes that nurses review and approve, aiming to reduce paperwork and improve workflow.
Key Facts
Abridge for Nurses is an AI tool designed specifically for nursing work, helping with documentation during patient care.
The platform launched first at Mayo Clinic in 2024 and is now used in health systems including Corewell Health, Johns Hopkins Medicine, and Emory Healthcare.
A recent survey found nearly 65% of U.S. nurses use more AI tools than a year ago, but 23% use no AI in daily work.
Nurses identify trust as the main barrier to adopting AI technology in healthcare.
Abridge developed the tool slowly and involved nurses in its design to build trust and ensure it fits nursing workflows.
The tool records conversations on hospital devices and drafts notes into electronic health records, which nurses then review before finalizing.
Full adoption is gradual, with many health systems using it in limited areas or pilot phases.
Nursing leaders say that although AI use is growing, widespread use and development of these tools are still in early stages.
Read the Original
Want the full story? Tap a source to open the original
article.
Amazon has started delivering parcels by drone for the first time in the UK, launching a test service in Darlington. The drones deliver small packages within a 7.5-mile radius of Amazon’s center, aiming to offer fast delivery for everyday items.
Key Facts
Amazon’s drone delivery service began in Darlington, County Durham, UK.
Drones carry parcels weighing less than 5 pounds (2.2 kg), including items like beauty products and batteries.
Deliveries occur within 7.5 miles (12 km) of the fulfillment center, with up to 10 drone flights per hour.
Customers receive parcels dropped at about 12 feet (3.6 meters) height onto their front gardens or yards.
The current delivery time is up to two hours in the UK; in the US, Amazon drones average 36 minutes per delivery.
Amazon uses the MK30 drone equipped with sensors and GPS to avoid obstacles during flights.
Some challenges include drone delivery limits in dense urban areas and past incidents like a drone hitting a building in Dallas.
Other organizations, such as the NHS and Royal Mail, are also testing drones for deliveries in the UK.
Read the Original
Want the full story? Tap a source to open the original
article.
Apple agreed to pay $250 million to settle a lawsuit claiming it falsely advertised new artificial intelligence (AI) features for Siri on the iPhone 16 and some other models. Eligible iPhone owners in the U.S. can claim between $25 and $95 because the AI features were delayed and not available as promised.
Key Facts
Apple settled a class-action lawsuit for $250 million over false advertising of AI features.
The lawsuit was filed in San Francisco federal court on behalf of U.S. iPhone buyers.
It covers about 37 million devices bought between June 10, 2024, and March 29, 2025.
Eligible models include iPhone 16, iPhone 15 Pro, and iPhone 15 Pro Max.
Owners can claim at least $25 per device, up to $95 depending on claims filed.
Apple had promoted new Siri AI features that were delayed and not available when devices were sold.
The settlement needs judge approval before claims can be submitted.
Apple says it resolved the matter to focus on delivering new products and services.
Read the Original
Want the full story? Tap a source to open the original
article.
SpaceX is gradually reducing the number of Falcon 9 rocket launches as it shifts focus to developing its larger Starship rocket. The Falcon 9 will continue flying for several years, especially to support the International Space Station and other missions, while Starship is being prepared for future flights.
Key Facts
SpaceX launched 165 Falcon 9 rockets in 2025, more than in previous years.
The company plans to reduce Falcon 9 launches to about 140-145 in 2026.
SpaceX is moving Falcon 9 launches away from Launch Complex 39A in Florida, which will focus on Starship and Falcon Heavy launches.
One of SpaceX’s two drone ships in Florida was retired to be repurposed to transport Starship parts.
Many Starlink satellite launches will move to Vandenberg Space Force Base in California, where launches can happen every 3-4 days.
Falcon 9 rockets and the Dragon capsule will remain in use at least until the International Space Station closes around 2032.
The U.S. Space Force will also use Falcon 9 and Falcon Heavy rockets into the 2030s.
Starship will start flying upgraded Starlink satellites soon, marking its operational beginning.
Read the Original
Want the full story? Tap a source to open the original
article.
Anthropic announced a new deal with SpaceX to use the full computing power of SpaceX’s data center in Memphis, Tennessee. This deal allows Anthropic to increase usage limits for its Claude Code product, helping more users access its AI services without restrictions.
Key Facts
Anthropic partnered with SpaceX to utilize the entire compute capacity of a SpaceX data center.
This partnership gives Anthropic access to over 300 megawatts of computing power.
SpaceX’s Colossus 1 supercomputer features more than 220,000 NVIDIA GPUs, including advanced models like H100, H200, and GB200 accelerators.
Anthropic raised usage limits for Claude Code Pro and Max subscribers, doubling the five-hour window and removing peak-hours restrictions.
The company also increased API limits for its Opus AI model.
Anthropic is interested in developing orbital data centers to handle growing AI compute demands.
Demand for Anthropic’s AI tools has grown due to users turning away from OpenAI and an increase in complex AI workflows.
Anthropic recently made deals with Microsoft, Google, Amazon, and Nvidia to expand access to computing power.
Read the Original
Want the full story? Tap a source to open the original
article.
A U.S. appeals court struck down a rule made by the FCC that stopped discrimination in broadband internet access. The court said the FCC went beyond its legal powers by punishing actions that unintentionally hurt certain groups and by applying the rule to companies not directly providing internet service.
Key Facts
The rule was from the Biden administration and aimed to prevent discrimination based on race, income, or other protected characteristics in broadband services.
The 8th Circuit Court ruled the FCC overstepped its authority by targeting “disparate impact” (unintentional harm) instead of only “disparate treatment” (intentional harm).
The ruling also said the FCC could not apply the rule to entities like landlords who don’t offer internet service directly.
The court decision was unanimous and made by judges appointed by Republican presidents.
FCC Chairman Brendan Carr supported the court’s decision, claiming the rules would have forced discrimination, though he did not explain how.
Advocacy groups opposed the decision, saying it removes protections against slow or expensive internet in low-income and minority communities.
The rule allowed consumers to file complaints if they believed broadband companies or others were discriminating based on protected factors.
Telecom and cable industry groups, along with state lobbyists and housing groups, challenged the rule in court.
Read the Original
Want the full story? Tap a source to open the original
article.
