An AI model developed by OpenAI has solved the Erdős unit distance problem, a famous math question that had remained unsolved for 80 years. Human mathematicians have reviewed and extended the AI's proof, marking a new milestone where AI and humans work together in mathematical research.
Key Facts
The Erdős unit distance problem asks how many pairs of points exactly one unit apart can be arranged in a plane.
It was introduced by mathematician Paul Erdős in 1946 and remained unsolved for 80 years.
OpenAI’s internal AI model found a proof that disproves the conjecture related to this problem.
Mathematicians, including Fields Medal winner Tim Gowers, confirmed the importance of the AI’s solution.
The AI combined ideas from different math fields but did not create new techniques on its own.
Human experts cleaned up and extended the AI’s proof to make it publishable.
This event shows progress in AI’s ability to solve complex math problems and suggests a future where AI and humans collaborate in research.
AI’s rapid improvement in math may change the role of human mathematicians in the coming decade.
Read the Original
Want the full story? Tap a source to open the original
article.
A new report from cybersecurity firm Check Point Software says hackers are focusing more on spreading false information to confuse voters than on changing voting machines or ballots for the 2026 midterm elections. The main tactics include phishing (fake messages to steal information), pretending to be others, and spreading misleading stories.
Key Facts
Hackers aim to confuse voters with false information rather than directly change votes.
The report focuses on the 2026 midterm elections.
Common hacker methods include phishing and impersonation.
Cybersecurity firm Check Point Software published the report.
Foreign influence operators are also involved in spreading misinformation.
The goal is to mislead voters during the election process.
Direct tampering with voting machines or ballots is currently less common.
The report highlights the growing threat of online electoral misinformation campaigns.
Read the Original
Want the full story? Tap a source to open the original
article.
A new report warns that the 2024 U.S. midterm elections could face increased cyber threats from websites related to voting and elections. The report highlights a surge in newly registered domains using words like "election" and "vote," which may be used for phishing, fraud, or spreading false information.
Key Facts
In early 2024, about 1,300 new domains contained "election" and 2,957 contained "vote."
Between April 13 and May 14, new registrations rose to 1,140 with "election" and 4,010 with "vote."
These domain registrations do not prove bad intent but increase the risk of scams and misinformation.
Major fundraising platforms ActBlue and WinRed experienced leaked user credentials, with 9,600 and 6,500 exposed respectively.
Political campaign websites for swing-state candidates showed very few leaked credentials.
The main cyber threats include phishing (tricking people to give private info), impersonation, AI-generated false content, and attempts to disrupt voting.
Around 82% of malicious attacks come through phishing emails.
Election officials conduct training to help recognize and prevent phishing attacks.
Read the Original
Want the full story? Tap a source to open the original
article.
New research shows that popular artificial intelligence (AI) systems often leave out religion when giving advice on important personal and moral questions. The studies also found that AI sometimes favors certain religions over others when users ask about changing their faith.
Key Facts
AI models mention religion only 5% to 16% of the time, even though people expect religion to be part of 45% to 59% of answers on moral and life topics.
The AI systems tested show a positive bias towards Catholicism, Baha'i, and Sikhism, and a negative bias against Jehovah's Witnesses, atheism, and agnosticism.
When answering questions about grief and loss, humans expect religion in 59% of cases, but AI includes it only 16% of the time.
AI rarely includes religion in answers about family, parenting, forgiveness, and ethics, despite people expecting it often.
The research comes from the Consortium for Evaluating Faith and Ethics in AI, involving several U.S. universities.
The studies reviewed 27 AI language models, including OpenAI’s GPT 5.5, Anthropic’s Claude 4.7, and Google’s Gemini 3.1.
AI is increasingly used in religious settings, such as church chatbots and apps for prayer and sermons.
Researchers suggest AI should better recognize when religious or spiritual advice is relevant, without assuming every user wants it.
Read the Original
Want the full story? Tap a source to open the original
article.
A nurse at Erlanger Baroness hospital in Chattanooga stole and abused fentanyl for months while using Sentri7, an AI-powered software designed to detect missing drugs. Despite this, the software failed to alert staff to the theft, highlighting potential problems with AI drug-monitoring systems that are not typically transparent or publicly reviewed.
Key Facts
A nurse at Erlanger Baroness hospital was caught slurring words and failing a drug test, leading to his firing for stealing fentanyl.
The nurse admitted to stealing leftover fentanyl from surgeries over several months.
The hospital used Sentri7, an AI-based software meant to detect drug diversion faster than humans.
The AI system did not flag the missing fentanyl or other related inconsistencies.
Hospitals are not required to publicly report AI software failures related to drug theft or monitor transparency.
Sentri7 is developed by Wolters Kluwer, a Dutch technology company, which expressed confidence in its software but did not comment on this case.
Experts say lack of transparency in AI drug-monitoring systems can hide errors and risks being repeated elsewhere.
The theft of fentanyl, a highly potent painkiller, is a common drug diversion method in hospitals.
Read the Original
Want the full story? Tap a source to open the original
article.
Nvidia has introduced a new AI chip called the RTX Spark for personal computers, aiming to bring advanced artificial intelligence capabilities to everyday devices. The chip will be in upcoming Windows PCs from several major brands and is expected to be available this autumn.
Key Facts
Nvidia announced the RTX Spark AI chip for personal computers.
The chip aims to help PCs function more like intelligent assistants rather than just tools.
The RTX Spark will be included in new Windows PCs from companies such as Asus, Dell, HP, Lenovo, Microsoft Surface, and MSI.
These PCs are expected to be available starting this autumn, with more models from Acer and Gigabyte later.
Nvidia’s CEO Jensen Huang compared the impact of this chip to the transformation of phones into smartphones.
Nvidia is currently the world’s most valuable company, worth over $5 trillion.
The US government recently tightened rules to limit the sale of Nvidia’s most advanced AI chips to Chinese companies.
This move is part of efforts to restrict China’s access to high-tech AI components.
Read the Original
Want the full story? Tap a source to open the original
article.
Nvidia introduced Cosmos 3, a new AI model that helps robots and autonomous vehicles understand and predict what will happen in real-world environments. This open model generates data about how machines move, which developers can use to train robots and other machines to navigate and act better in physical spaces.
Key Facts
Cosmos 3 is trained on 20 trillion pieces of data, including images, videos, sounds, text, and robot movements.
Unlike regular AI that creates images or videos, Cosmos 3 models machine actions like joint movements and paths.
Developers can use Cosmos 3 to simulate physical tasks and build specialized models for robots and vehicles.
The model is open, allowing hardware makers to customize it for their needs.
Nvidia is working with companies like Agile Robots and Runway to support Cosmos 3.
Cosmos 3 can create rare or risky scenarios for training that are hard to replicate in real life, such as robot accidents.
Nvidia offers two versions now: a high-accuracy "super" model and a fast "nano" model, with a local "edge" version coming soon.
Nvidia aims to make AI that not only answers questions or creates images but can plan and act in the physical world.
Read the Original
Want the full story? Tap a source to open the original
article.
The U.S. Department of Commerce has clarified that export controls on advanced AI chips apply to Chinese companies and their subsidiaries even if they are outside China. This guidance closes a loophole that allowed some Chinese firms abroad to receive restricted chip shipments without licenses.
Key Facts
The U.S. restricts shipments of advanced AI chips to Chinese companies and their subsidiaries worldwide.
The Bureau of Industry and Security (BIS) issued the clarification after questions about export rules arose.
The Trump administration canceled a previous global licensing plan from the Biden era aimed at controlling AI chip exports.
Nvidia, a leading chip maker, says it already follows licensing rules for sales to China-based companies.
Some experts say the previous lack of clear rules allowed Chinese companies to buy restricted chips legally.
The updated guidance means shipments of banned chips to Chinese firms outside China are now illegal.
The U.S. and China are competing closely in advanced AI technology.
In late 2023, President Trump eased some export limits by allowing Nvidia to sell a powerful chip (H200) to China.
Read the Original
Want the full story? Tap a source to open the original
article.
South Korea is using robots with artificial intelligence (AI) to help older people feel less lonely. These robot companions remind seniors to take their medicine, check on their health, and can call for help if there is an emergency.
Key Facts
South Korea has a growing number of older adults who may experience loneliness.
The country uses AI-powered companion dolls to support these seniors.
The robots remind users to take their medications on time.
They monitor the health and wellbeing of the seniors.
The dolls can call for help in case of emergencies.
There are ongoing questions about whether AI companions can fully replace human relationships.
Read the Original
Want the full story? Tap a source to open the original
article.
For over a year, top technology company leaders have been trying to gain the support of President Donald Trump as he began his second term. They have met with him at locations like Mar-a-Lago and made large donations during his inauguration.
Key Facts
Tech leaders have actively sought a positive relationship with President Trump.
Meetings have occurred at Mar-a-Lago, President Trump’s private club.
CEOs such as Elon Musk and Mark Zuckerberg have participated in these efforts.
Large donations were given by tech companies for President Trump’s inauguration.
These actions occurred since President Trump returned to the White House for his second term.
Tech companies view maintaining good ties with the government as important.
Efforts show a blend of direct meetings and financial support to gain favor.
Read the Original
Want the full story? Tap a source to open the original
article.
KSI, a British social media star and judge on Britain's Got Talent, announced he is leaving the YouTube group Sidemen after 13 years. He shared that May 31 would be his last video with the group, ending his long-time collaboration.
Key Facts
KSI’s real name is Olajide Olatunji.
He joined the Sidemen collective in 2013 and helped create football, comedy, and challenge videos.
The Sidemen group has 23 million subscribers on YouTube.
KSI announced his departure on social media, calling it a difficult decision after months of thought.
He has over 18 million subscribers on his personal YouTube channel.
KSI is also a musician, former boxer, and part owner of a football club.
His spokesperson confirmed he was leaving after fans suspected it might be a stunt.
Read the Original
Want the full story? Tap a source to open the original
article.
Two developers created Mira, a pair of AI-powered glasses that transcribe and summarize conversations without using a camera. The glasses aim to help users create a searchable record of their daily life, acting like a "second brain."
Key Facts
The developers, Anhphu Nguyen and Caine Ardayfio, previously added facial recognition to Meta's Ray-Ban glasses.
Mira glasses cost $650 and do not have a camera.
These glasses quietly record spoken information and turn it into searchable notes.
Technology Editor Peter O'Brien tested Mira for two weeks.
The glasses raise privacy concerns about recording people nearby.
The idea is to help users easily remember and find information from daily conversations.
Mira build on previous experiments with wearable tech and AI.
Meta’s involvement is indirect; the creators originally hacked Meta’s smart glasses but now have their own product.
Read the Original
Want the full story? Tap a source to open the original
article.
Skyscrapers have hidden floors that are not seen by the public. These floors hold important systems like heating, water, and electrical equipment that help keep the building safe and working well. These hidden levels also include special structural floors that support the building’s weight and fire refuge floors for emergencies.
Key Facts
Hidden floors in skyscrapers house mechanical systems such as heating, ventilation, air conditioning (HVAC), water tanks, pumps, and electrical equipment.
These floors are not counted in the public floor numbering and are usually not accessible to occupants.
Mechanical floors are placed throughout tall buildings to properly distribute essential services from top to bottom.
Some hidden floors, called structural transfer floors, contain beams and support systems instead of regular rooms.
Fire refuge floors are included for safety during emergencies but are also not publicly acknowledged.
In very tall buildings like the Burj Khalifa, multiple mechanical levels have larger heights to hold bulky equipment.
These hidden floors allow skyscrapers to operate like “vertical cities” with systems working efficiently across many floors.
The building industry is focusing on making skyscrapers more energy-efficient and sustainable due to their large impact on global CO₂ emissions.
Read the Original
Want the full story? Tap a source to open the original
article.
Peacock announced its list of movies and shows coming in June 2026. Highlights include the Season 8 premiere of "Love Island" on June 2 and a new documentary called "Lorne" about the creator of Saturday Night Live.
Key Facts
"Love Island" Season 8 premieres on June 2, 2026, on Peacock.
The Season 4 finale of "Below Deck Down Under" airs on June 2.
The documentary "The Alpinist" about climber Marc-André Leclerc is available starting June 2.
Peacock adds many older movies on June 1, including "Jurassic Park," "The Fast and the Furious" series, and "Being John Malkovich."
A new documentary called "Lorne" premieres on June 5; it focuses on Lorne Michaels, creator of Saturday Night Live.
Other shows premiering or concluding in June include "Summer House," "America’s Got Talent," "Top Chef," "American Ninja Warrior," and "The Real Housewives of Rhode Island."
The list features both new seasons and reunions of popular reality TV series.
Peacock provides a mix of movies, documentaries, reality shows, and competition series throughout June.
Read the Original
Want the full story? Tap a source to open the original
article.
Elon Musk’s company SpaceX plans to grow by using its Starship rocket and artificial intelligence (AI). The company also focuses on Starlink, its satellite internet service, and rocket launch services as part of its growth strategy.
Key Facts
SpaceX is developing the Starship rocket for future missions.
The company uses AI to help with its near and mid-term growth goals.
Starlink is SpaceX’s satellite internet service, which provides global internet coverage.
SpaceX offers rocket launch services for various customers.
Elon Musk leads the company’s innovation efforts in space and technology.
AI refers to computer systems that can perform tasks usually requiring human intelligence.
The combination of Starship, AI, Starlink, and launch services supports SpaceX’s overall business growth.
Read the Original
Want the full story? Tap a source to open the original
article.
Leaders in the technology field, like Sam Altman of OpenAI and Elon Musk, imagine a future where humans and artificial intelligence (AI) merge or where AI becomes the dominant species beyond Earth. They believe this next stage of evolution, called “transhumanism,” could let humans live longer and travel through space by turning consciousness into digital form.
Key Facts
Sam Altman predicts humans may merge with AI within the next 50 years.
Elon Musk describes humans as a “biological bootloader” for digital superintelligence.
Some tech leaders think future humans will be partly or fully digital to explore the galaxy.
This idea is called transhumanism, aiming to extend life and expand human abilities.
The concept involves turning human consciousness into digital code stored on computers or transmitted through space.
These beliefs shape technology development, focusing on creating superhumans rather than just useful tools.
The tech elite behind these ideas have gained huge wealth and influence from advances in information technology.
Critics warn this vision may ignore the needs and desires of regular people.
Read the Original
Want the full story? Tap a source to open the original
article.
High-resolution photos and videos shared on social media can reveal detailed biometric data like fingerprints and voice patterns. Cybercriminals can use artificial intelligence (AI) tools to steal these details and create fake images or voices, which may lead to identity theft, fraud, or other crimes.
Key Facts
Modern phone cameras capture biometric details, such as fingerprint ridges, in near forensic quality.
AI can reconstruct usable biometric templates from social media photos and videos.
Voice cloning technology can create realistic fake voices from just a few seconds of recorded audio.
Deepfake technology can make convincing videos of people doing or saying things they never did.
Biometric data like fingerprints and voice patterns cannot be changed once stolen, unlike passwords.
In 2024, the FBI recorded over 859,000 cybercrime complaints with losses exceeding $16 billion.
Phishing emails remain a major cyber threat, with 3.4 billion malicious emails sent daily.
Experts warn that the risk of biometric theft is growing as AI lowers the skill barrier for attackers.
Read the Original
Want the full story? Tap a source to open the original
article.
The article looks back on the 1986 movie SpaceCamp and its connection to NASA's Space Shuttle program. It explains how the shuttle was meant to make space travel routine but was set back by the Challenger disaster. The movie, aimed at young audiences, was released shortly after the disaster and had mixed success.
Key Facts
NASA’s Space Shuttle was designed to be a reusable spacecraft for frequent trips to low Earth orbit.
The shuttle was expected to make space travel common, with plans for monthly or even weekly flights.
The Challenger disaster in January 1986, which killed educator Christa McAuliffe, ended hopes for quick civilian space travel.
SpaceCamp, a movie made before Challenger’s loss, showed kids training for space missions and used real NASA footage.
The film struggled at the box office, earning $9.6 million on a $25 million budget.
Audiences at the time were reluctant to see a film about kids in danger on a space shuttle, likely due to the disaster’s recent memory.
Today, SpaceCamp is mostly remembered by fans who saw it as children and is not currently available on streaming platforms.
The article’s authors recently rewatched the film and found it still has elements of charm despite its flaws.
Read the Original
Want the full story? Tap a source to open the original
article.
NASA selected two companies, Astrolab and Lunar Outpost, to build new electric moon buggies for its upcoming lunar base project. These vehicles will carry astronauts and supplies, work remotely, and operate for about a year on the moon's surface.
Key Facts
Astrolab and Lunar Outpost each received about $220 million from NASA to build moon rovers.
The rovers will be four-wheel drive electric vehicles that can carry two astronauts and supplies.
The vehicles will be designed to work on their own or be controlled remotely from Earth.
NASA plans to build a moon base near the lunar south pole over the next seven years.
The rovers will combine features from the old Apollo Lunar Rover and the modern Mars Perseverance Rover.
Blue Origin’s New Glenn Rocket, important for delivering these rovers, recently exploded in a test.
NASA announced a $20 billion plan to build the moon base.
Lunar Outpost aims to sell its rover to customers beyond NASA to support various activities on the moon.
Read the Original
Want the full story? Tap a source to open the original
article.
NASA is moving forward with its plan to return to the Moon by ordering lunar rovers for a future Moon base. This progress comes despite a recent setback involving an accident with Blue Origin, a company working with NASA.
Key Facts
NASA plans to build a base on the Moon.
They have ordered special vehicles called lunar rovers to use at the base.
Lunar rovers are designed to move around the Moon's surface.
Blue Origin, a private space company, had an explosion that caused a delay.
Despite this, NASA’s Moon mission is making significant progress.
The goal is to support astronauts living and working on the Moon for longer periods.
The development shows ongoing efforts to explore space beyond Earth.
Read the Original
Want the full story? Tap a source to open the original
article.
