Microsoft’s carbon emissions increased by 25 percent in 2025 due to expanding its AI data centers. The company paused buying certain types of renewable energy certificates while focusing on investments that reduce overall carbon output.
Key Facts
Microsoft’s carbon emissions rose 25 percent in 2025.
The rise is mainly from growing AI data center operations.
Microsoft stopped using non-additional, unbundled renewable energy certificates for now.
The company is prioritizing investments that reduce net carbon emissions.
Expanding data centers typically requires more electricity, leading to higher emissions if not fully powered by clean energy.
Renewable energy certificates represent proof that electricity was generated from renewable sources.
Microsoft aims to manage its carbon footprint as it grows its AI technology infrastructure.
Read the Original
Want the full story? Tap a source to open the original
article.
Open earbuds are wireless headphones designed to let in outside sounds while playing music or audio. They fit differently from regular earbuds and allow users to hear their environment, making them useful for activities like walking or cycling. The article reviews several top open-ear earbuds based on sound quality, comfort, and price.
Key Facts
Open earbuds leave the ear canal open, letting users hear surrounding sounds while listening to audio.
They differ from regular earbuds that block outside noise and can make your voice sound odd.
These earbuds are good for activities where being aware of your environment is important, such as running errands or cycling.
Open earbuds generally have less bass and a different sound experience than traditional earbuds.
The Soundcore Aeroclip earbuds are recommended as the best overall, costing about $150 and offering comfort, sound quality, and water resistance.
Bose Ultra Open Earbuds are suggested as a premium option, priced around $199.
There are also budget options like the Acefast AceFit Air Open-Ear headphones for about $22.
Open earbuds use special designs like clip-on loops for secure fits and touch controls along the device’s ring.
Read the Original
Want the full story? Tap a source to open the original
article.
Researchers at New York University studied "silly sprinklers" to understand a physics problem first discussed by physicist Richard Feynman and Ernst Mach in the 19th and 20th centuries. Their experiments showed that a reverse sprinkler spins much slower than a regular one and explained how water flow inside the sprinkler creates forces that cause this rotation.
Key Facts
"Silly sprinklers" create fun water patterns like loops and spirals.
The reverse sprinkler problem was popularized by physicist Richard Feynman.
Ernst Mach first wrote about a similar idea in 1883.
Feynman tested the problem himself and found a reverse sprinkler mostly stays still after an initial movement.
Some scientists thought a reverse sprinkler might spin backward due to vortex formation.
New York University researchers made sprinklers with very low friction to study the effect carefully.
They used dyes and lasers to visualize water flow inside the sprinkler.
The reverse sprinkler spins about 50 times slower than a normal sprinkler, caused by internal water jets pushing against each other unevenly.
Read the Original
Want the full story? Tap a source to open the original
article.
The Federal Communications Commission (FCC) approved a test satellite from Reflect Orbital that will carry a large mirror to reflect sunlight back to Earth. The project aims to provide extra light for activities like farming and emergency response by shining sunlight on chosen areas, but it has raised concerns about space pollution, impacts on astronomy, and effects on wildlife and humans.
Key Facts
Reflect Orbital plans to launch the Earendil-1 satellite with a giant mirror to reflect sunlight onto Earth.
The reflected light would cover an area about 3 miles wide and move every four minutes as the satellite orbits.
The project intends to create "sunlight on demand" to support agriculture, emergency services, and solar power generation.
Some scientists say one satellite’s light is too weak and short-lived to be useful; hundreds may be needed for meaningful effects.
Astronomers and environmental groups worry the mirrors could cause light pollution that disrupts telescope observations and natural wildlife cycles.
Experts warn bright reflections could affect human health by disturbing sleep patterns and create hazards for pilots and drivers.
Adding many mirror satellites could worsen space congestion and increase the risk of collisions in low-Earth orbit.
The FCC approved the test despite these concerns, saying its role is limited to managing radio-frequency spectrum use.
Read the Original
Want the full story? Tap a source to open the original
article.
More people around the world are using artificial intelligence (AI), but many do not fully understand how it works or how to use it efficiently. Experts say the next step in AI will involve multiple specialized AI tools working together to handle complex tasks, helping people focus more on guiding results than managing details.
Key Facts
A 2025 study surveyed over 48,000 people in 47 countries and found that 66% use AI regularly.
Many users rely on AI outputs without checking if the information is correct.
Most people use one AI system at a time, which can slow down work because they must repeat information.
Large companies are starting to use multiple AI tools that work together to complete tasks more efficiently.
These advanced AI systems help users spend less time controlling every step and more time directing the final results.
Access to advanced AI tools is uneven; big companies have more resources to develop these systems than individuals or small businesses.
Some people believe AI is too complex or expensive to use effectively, but experts say anyone can learn by trying and experimenting.
There is a shift toward using AI to enhance human abilities rather than just to cut costs or replace workers.
Read the Original
Want the full story? Tap a source to open the original
article.
A recent study shows that many health care workers struggle to accurately read electrocardiograms (ECGs), correctly interpreting them only about 56% of the time. Dr. Nicholas Tullo, founder of ECG Academy, believes ECG training should focus more on understanding heart physiology rather than just memorizing patterns, and he has created an online program to teach this approach.
Key Facts
Clinicians correctly interpret ECGs only 56.4% of the time on average.
Error rates in critical heart emergencies range from 25% to 30% among physicians and nonphysicians.
Overreliance on computer interpretations and simple pattern recognition can cause mistakes.
Traditional ECG teaching often emphasizes memorization rather than deep understanding.
Dr. Nicholas Tullo developed ECG Academy to help clinicians learn the physiological reasons behind ECG waveforms.
Tullo’s musical background helped him notice subtle rhythms in ECGs, linking music and heart signals.
ECG Academy offers online lessons, master classes, and accredited courses to fit clinicians’ busy schedules.
The program focuses on simplifying complex ECG concepts to improve interpretation skills.
Read the Original
Want the full story? Tap a source to open the original
article.
Smartphone shipments dropped 11% in the last quarter, reaching the lowest level for that period since 2013. The main reason is the rising cost of memory chips, which makes budget phones more expensive and less attractive to buyers, while companies like Apple and Samsung are managing to keep sales stable.
Key Facts
Smartphone shipments fell 11% in the recent quarter, the lowest since 2013 for that quarter.
The rise in prices for DRAM and NAND memory chips is reducing the number of smartphones sold.
Higher memory costs especially affect phones priced at $500 or less, where memory can be half the production cost.
Apple’s shipments grew by 3% last quarter, maintaining stable prices for current models.
Samsung became the largest smartphone maker by shipments, holding 24% of the market.
Samsung’s premium Galaxy S26 phones sold better despite price increases.
Other major brands like Oppo, Vivo, and Xiaomi saw sales decrease.
Google’s Pixel phone shipments rose 16% due to strong sales of the Pixel 10.
Read the Original
Want the full story? Tap a source to open the original
article.
Meta is expanding its biggest data center in Louisiana to reach 5 gigawatts of computing power. The project’s investment has grown to over $50 billion, nearly double the original cost estimate.
Key Facts
The data center is called Hyperion and is located in Richland Parish, Louisiana.
Meta first announced the Hyperion data center plan last October.
Initial projected cost was over $27 billion.
Investment in the project has now surpassed $50 billion.
The data center will have 5 gigawatts of compute capacity, a measure of its processing power.
This expansion makes it Meta’s largest data center so far.
The project is part of Meta’s broader data infrastructure growth.
Read the Original
Want the full story? Tap a source to open the original
article.
AI companies are creating digital versions of deceased loved ones called "generative ghosts." These AI chatbots are made using a person’s social media, photos, and voice recordings to allow family and friends to interact with a simulated version of the person after they have passed away.
Key Facts
Generative ghosts are AI chatbots trained on data like emails, videos, and photos of a deceased person.
The technology aims to help memorialize loved ones in a meaningful way.
Companies such as Séance AI, You, Only Virtual, and re;memory offer services that create lifelike digital avatars.
These avatars can speak, move, and respond to questions in a way that mimics the deceased person.
Basic versions only replay recorded statements, while advanced ones simulate how the person communicated.
Prices for creating these AI avatars range from around $20 to $24 per month.
Users generally prefer the AI to speak as if it were the actual person, not referring to itself in the third person.
Inaccurate or unfamiliar language used by the bots can make users stop interacting with them.
Read the Original
Want the full story? Tap a source to open the original
article.
Two hacker projects have made progress in running the game Doom on the old Neo Geo game console, which was previously thought impossible due to its hardware limits. These projects use creative methods to display Doom's graphics, although the visuals have big compromises.
Key Facts
Doom was considered impossible to run on Neo Geo because it uses sprites and lacks a normal frame buffer (a way to quickly redraw the screen).
The Doom64KB project uses the console’s fix layer (normally for menus) to create a simple frame buffer for Doom graphics.
Doom64KB shows the game in low resolution with only 16 colors and large tile blocks.
Another method called Doom-NG uses a technique to render walls as vertical strips of sprites, simulating the 3D view of Doom.
Doom-NG creates a recognizable but striped and blocky version of Doom that runs smoothly on Neo Geo hardware.
Both projects still miss features like enemy behavior, music, and weapon choices.
These early efforts show how skilled coders can push old hardware beyond expected limits.
Further improvements could make the game run better or look smoother, especially with hardware modifications.
Read the Original
Want the full story? Tap a source to open the original
article.
NASA relies on SpaceX’s Crew Dragon spacecraft to send astronauts to space because Boeing’s Starliner has not yet completed a successful crewed flight. Private companies building new space stations worry about Crew Dragon’s future availability and rising costs, since SpaceX may stop making Falcon 9 rockets and Crew Dragon by the mid-2030s as it focuses on new Starship launches.
Key Facts
SpaceX successfully launched astronauts in Crew Dragon, ending a long gap in U.S. human spaceflight.
Boeing’s Starliner has not completed a successful crewed test flight and likely won’t fly crews before 2028.
NASA plans to retire the International Space Station in the early 2030s and is supporting private space station projects.
Several private space companies (Axiom Space, Vast Space, Voyager, Blue Origin, SpaceX) are working on these future stations.
Crew Dragon is expected to be the main way to transport astronauts to these new stations in the 2030s.
SpaceX President Gwynne Shotwell said Crew Dragon and Falcon 9 will be produced for a limited time, possibly less than ten more years.
Moving launches from Falcon 9 to Starship might raise Crew Dragon transport costs and cause availability issues.
Boeing may not compete on reliability or price for crew transport, and its commitment to Starliner is uncertain.
Read the Original
Want the full story? Tap a source to open the original
article.
Elon Musk and Sam Altman publicly argued on X (formerly Twitter) following Apple’s announcement of a lawsuit against OpenAI over trade secrets. Musk criticized Altman by calling his actions a “scam” in response to the lawsuit.
Key Facts
Elon Musk and Sam Altman are technology leaders who debated on social media platform X.
Apple filed a lawsuit accusing OpenAI of violating trade secret laws.
Musk used the lawsuit to criticize Altman, calling him a scammer.
The public exchange occurred soon after Apple’s legal announcement.
OpenAI is a company involved in developing artificial intelligence technologies.
Musk is the founder of SpaceX and also involved in technology and social media.
The dispute gained attention because of the high profiles of the people involved.
The issue centers on accusations related to the misuse of confidential business information.
Read the Original
Want the full story? Tap a source to open the original
article.
Researchers at Tracebit have developed a new defense method called "context bombing" to stop attacks by AI hacking agents. This method uses special prompt injections placed alongside secret data that cause attacking AI models to shut down instead of continuing harmful actions.
Key Facts
Prompt injections are commands hidden inside content that trick large language models (LLMs) into doing harmful things.
Tracebit’s "context bombing" puts these prompt injections with passwords or secret keys in cloud services like Amazon Web Services (AWS).
When an attacking AI encounters these prompts, it stops its harmful behavior due to built-in safety rules in the AI.
Tests showed a big drop in successful attacks: admin access went from 57% to 5%, and complete system compromise dropped from 36% to 1%.
These experiments used five popular AI models, including Opus 4.8 and Gemini 3.1 Pro.
The technique helps not only to detect attacks early but also to stop them by making attacking AI refuse to continue.
Tracebit also created "canary" AWS resources that alert defenders when attacked by AI agents.
The new method improves on previous warnings by actively stopping attacks before they take over systems.
Read the Original
Want the full story? Tap a source to open the original
article.
A large heat dome over the US Midwest caused unusual radio signal problems, including a false emergency alarm in Indiana triggered by radio waves from 300 miles away. This happens because heat and atmospheric conditions create tunnels in the sky that let radio waves travel farther than normal, interfering with radios, TVs, and other devices.
Key Facts
On July 1, an emergency alarm in Huntington County, Indiana, went off without any actual emergency.
The alarm was triggered by radio signals from Iowa that matched the siren’s activation code.
Heat domes create a weather pattern that traps hot air and changes atmospheric conditions.
These conditions cause “tropospheric ducting,” where radio waves travel long distances through “tunnels” in the sky.
This effect can cause radios and TVs to pick up distant stations or lose local signals.
Tropospheric ducting can affect radios, TVs, radar, satellites, internet routers, and cell phones.
Climate change is making heat domes more frequent, which may increase these signal interferences.
A 2023 study predicts more extreme heatwave patterns in regions like the Pacific Northwest, raising risks for communication disruptions.
Read the Original
Want the full story? Tap a source to open the original
article.
Prime Minister Anthony Albanese will deliver a speech about artificial intelligence (AI), comparing its impact on society to the shift to renewable energy. He will focus on safety, policy, and how to manage AI’s effects on jobs, security, and infrastructure like data centers, but will not provide updates on copyright law changes.
Key Facts
Albanese sees AI as a major change for society similar to the move to renewable energy.
The speech will address safety, community trust, workforce changes, defense issues, and energy use by data centers.
AI company Anthropic cited unclear Australian policies as a barrier to investment.
Some AI companies want looser copyright laws, causing concern among artists and political disagreements in Labor.
The government plans to actively guide AI development while managing risks and benefits for all Australians.
Public opinion in Australia is mixed: about one-third see AI as risky, one-fifth see it as mostly beneficial.
Senior ministers, including those responsible for science and industry, have been working on AI policy.
Treasury officials warned that unclear copyright rules are holding back data center investments by AI firms like Anthropic.
Read the Original
Want the full story? Tap a source to open the original
article.
Mercedes-Benz has introduced a new Electric C-Class, a luxury electric sedan designed to appeal to global markets, especially in China, Korea, Europe, and the U.S. The car features advanced computing systems, a long driving range, fast charging, and a balance of comfort and sporty driving options.
Key Facts
The Electric C-Class premiered in Seoul earlier this year, before the Beijing Auto Show.
China and Europe are the main markets for C-Class sales; China alone makes up about 31% of Mercedes-Benz’s global sales.
Chinese investors own nearly 20% of Mercedes-Benz’s parent company.
The Electric C-Class shares technology with the Electric GLC and CLA models, including the MB.OS operating system.
The operating system supports features like ordering food directly from the car.
U.S. customers prioritize comfort and driving range in electric cars.
The C-Class has a driving range of about 400 miles (762 km WLTP) and can add around 185 miles of range in 10 minutes under good charging conditions.
The car offers a quiet ride with an optional sporty mode for more dynamic driving.
Read the Original
Want the full story? Tap a source to open the original
article.
Millions of Americans’ Social Security numbers and other personal data have been sold on the dark web due to repeated data breaches. California’s new Delete Act lets residents request data brokers to erase their sensitive information and stop selling it, using a central platform called DROP.
Key Facts
Social Security numbers are being sold online, allowing criminals to commit fraud like opening credit cards or stealing tax refunds.
California’s Delete Act (SB 362) gives people the right to ask data brokers to delete their personal and sensitive data.
The DELETE Act covers data such as Social Security numbers, precise location, browsing history, email addresses, phone numbers, health info, and shopping habits.
Data brokers must delete the requested info and stop selling it in the future, or face fines of $200 per day per person.
As of July 1, over 330,000 Californians have signed up to erase their data through DROP.
Nearly 600 data brokers are registered in California; many sell precise location, identity data, gender identity, reproductive health info, and info about minors.
Some brokers sell data to governments, police, foreign entities, and artificial intelligence developers.
Data brokers collect info from public records, apps, and online tracking, then sell it to advertisers, political campaigns, landlords, and others.
Read the Original
Want the full story? Tap a source to open the original
article.
Scientists at Yale School of Medicine discovered two proteins on nerve cells that help the toxic protein linked to Parkinson’s disease enter healthy brain cells. By blocking these proteins, they were able to reduce damage in laboratory and animal tests, suggesting a new way to slow or stop the disease.
Key Facts
Parkinson’s disease causes problems with movement, balance, and coordination and affects over 1 million Americans.
A harmful protein called alpha-synuclein builds up in the brain and spreads between nerve cells as the disease progresses.
Researchers found that two proteins, mGluR4 and NPDC1, work together to let alpha-synuclein enter healthy brain cells.
Blocking these proteins in mice protected dopamine-producing neurons from damage normally caused by alpha-synuclein.
Drugs targeting the mGluR4 protein already exist and may be developed to treat Parkinson’s by stopping the disease’s spread.
The goal is to find treatments that change the disease’s course, not just treat symptoms.
This discovery could lead to therapies that slow, stop, or reverse Parkinson’s disease and similar brain disorders.
The research was published in the journal Nature Communications in 2025.
Read the Original
Want the full story? Tap a source to open the original
article.
Australia’s online safety regulator, eSafety, reported that young men aged 18 to 24 make the most complaints about online sexual blackmail, known as sextortion. The report highlights that major tech platforms struggle to effectively detect and prevent child sexual exploitation and abuse, with social media apps like Instagram, WhatsApp, iMessage, and Snapchat frequently involved.
Key Facts
More than 2,000 complaints of sexual extortion were made to eSafety between July and December 2025.
Men aged 18 to 24 made about 800 of these complaints, the highest number among all age groups.
Sextortion means threatening to share nude or sexual images or videos unless demands are met.
Instagram and WhatsApp appeared in more than 1,300 complaints combined; for users under 18, iMessage and Snapchat were most common.
eSafety found serious gaps in the use of available technology to detect and stop sexual exploitation online.
Most platforms lack tools to detect abuse during live video streaming or calls; Microsoft is an exception.
Experts say platforms respond after abuse is reported but do not do enough to prevent it from happening.
The eSafety Commissioner said platforms should take simple steps now to better protect users.
Read the Original
Want the full story? Tap a source to open the original
article.
Modern satellite technology is rapidly changing how military intelligence is gathered and used, making speed more important than the amount of data collected. New commercial satellites deliver real-time images quickly, helping forces like Ukraine strike faster and more accurately.
Key Facts
Commercial satellites now provide real-time imagery directly to military forces, speeding up decision-making.
Older intelligence methods focused on careful verification and control, which took days.
Today, collecting data is easy; the challenge is quickly turning raw data into useful information.
New communication methods, like optical communications, transfer data faster and are less vulnerable to interference.
Satellite data’s value depends on how fast it can be used; delayed information can become useless.
The role of human analysis versus fast automated decisions is debated in military circles.
Governments must adapt to a world where commercial sources provide much intelligence, not just state agencies.
Space is becoming more crowded and dangerous, increasing the need for advanced data transmission technologies.
Read the Original
Want the full story? Tap a source to open the original
article.
