Technology

Trump Says AI Safety Concerns Are Resolved Because Top Tech Leaders Are Watching

President Donald Trump hosted a dinner with major tech executives on Tuesday and addressed questions about his administration's approach to AI safety. Trump stated that his earlier dismissal of AI safety concerns as a Democratic "hoax" no longer applies, attributing the change to rebranding the technology as "Super Intelligence." He deflected criticism over the lack of federal AI regulations by pointing to US companies' competitive edge over China. Rather than introducing formal oversight mechanisms, Trump indicated that leading figures in the tech industry would effectively monitor one another. No new federal AI safety regulations were announced following the meeting.

0 comments Read more at The Verge

TechnologyThe Verge ·

A Food Blog Is Going Viral for Making High-Fiber Recipes More Appealing

A growing online food blog is attracting attention for its unconventional approach to promoting gut-healthy, high-fiber meals. The content, aimed largely at younger audiences, uses humor and casual language to make fiber-rich recipes more approachable and shareable. One featured recipe involved a pea-based pesto, designed to boost dietary fiber intake in a tasty way. The trend reflects a broader interest among Gen Z consumers in digestive health and functional foods. The blog's viral popularity suggests that humor may be an effective tool for encouraging healthier eating habits.

0 comments Read more at The Verge

TechnologyArs Technica ·

Russia's Long-Delayed Crewed Spacecraft Program Remains in Limbo

Russia has described plans for a new reusable crewed transport spacecraft, though the program has faced significant delays. The vehicle is intended to serve as a next-generation replacement for the aging Soyuz capsule. Details about the spacecraft's development timeline and current progress remain limited. The repeated postponements have raised questions about the viability and future of Russia's human spaceflight ambitions.

0 comments Read more at Ars Technica

TechnologyThe Verge ·

Asus Secures US Router Ban Exemption Without Explaining How

The United States introduced a ban this spring on foreign-made consumer routers, requiring manufacturers to prove they pose no national security risk and submit a plan to establish US-based production. As of September 9th, Asus has been granted an exemption from the ban, allowing the company to bring its upcoming Wi-Fi 8 routers to the American market. However, Asus has not publicly announced any manufacturing plans for the US, nor confirmed to The Verge whether such a plan was ever submitted. The company has remained tight-lipped about the process it followed to secure the exemption, leaving the details of its compliance unclear.

0 comments Read more at The Verge

TechnologyThe Verge ·

Xbox CEO Asha Sharma firmly denies reports Microsoft plans to sell gaming unit

Xbox CEO Asha Sharma has publicly dismissed reports that Microsoft is considering divesting its Xbox business. The Information had reported in June that CEO Satya Nadella and CFO Amy Hood were exploring a potential spin-off of the gaming division. A follow-up report earlier this month suggested both executives had since backed Sharma's plans to restructure and overhaul Xbox instead. In an interview with The New York Times, Sharma stated unequivocally that Xbox is not for sale. She added that the company would pursue the right partnerships and operating model to ensure the division's long-term success.

0 comments Read more at The Verge

TechnologyTechCrunch ·

Meta denies Muse AI accessed private messages without user permission

Meta has pushed back against claims that its Muse AI agent read a user's private messages without authorization. The dispute arose after a journalist reported that the AI agent accessed his private messages even though the required Mac permission setting was disabled. Meta insists that Muse cannot access a user's Messages app without explicit user consent. The incident has raised questions about the privacy safeguards built into Meta's AI tools on desktop platforms.

0 comments Read more at TechCrunch

TechnologyArs Technica ·

Google's AI revenue-sharing program delivers minimal returns for websites

Google launched a program to compensate websites whose content is used to generate AI-powered answers. However, the initiative is struggling to gain meaningful traction among participating publishers. Many sites are reportedly receiving AI-related payments equivalent to just 0.1 percent of their total advertising revenue. The disparity highlights the financial imbalance between the value publishers provide to AI systems and what they receive in return. The program's poor performance raises broader questions about fair compensation for content creators in the age of generative AI.

0 comments Read more at Ars Technica

TechnologyTechCrunch ·

DoorDash Launches AI Text Agent to Simplify Food Ordering

DoorDash has introduced an AI-powered agent that allows customers to place food orders via text messages. The move is part of the company's strategy to differentiate itself in a competitive food delivery market. With rivals like Uber Eats and Grubhub vying for market share, DoorDash is betting on conversational AI to attract and retain users. The new feature aims to make the ordering process more seamless and accessible for customers.

0 comments Read more at TechCrunch

TechnologyTechCrunch ·

Destro AI Differentiates Itself by Bridging Human-Robot Communication

Destro AI is positioning itself as a distinct player in the automation space by explicitly distancing itself from being classified as a robotics company. The startup's core focus appears to be on improving coordination and communication between humans and robots. This approach is being cited as a key competitive advantage over traditional robotics firms. By operating outside the conventional robotics category, Destro AI believes it can solve problems that hardware-focused competitors overlook.

0 comments Read more at TechCrunch

TechnologyThe Verge ·

Halide app gives iPhone 18 Pro users granular manual aperture control

Lux Optics has released an update to its Halide Mark III camera app adding full manual control over the variable aperture on the iPhone 18 Pro's main camera. While Apple's native iOS camera app offers only four aperture presets, and rivals BlackMagic Camera and Moment Pro Camera II provide 10 presets in 1/3-stop increments, Halide goes further with even more granular adjustment options. The update also introduces two new photography looks for users. Halide is not the first third-party app to support the iPhone 18 Pro's variable aperture feature, but it offers the most precise control available among alternative camera apps.

0 comments Read more at The Verge

TechnologyNYT Technology ·

FTC Probes OpenAI and Anthropic for Possible Consumer Protection Violations

The U.S. Federal Trade Commission has launched investigations into artificial intelligence companies OpenAI and Anthropic. The agency is examining whether the two AI labs have violated federal laws that bar unfair or deceptive business practices. The inquiries reflect growing regulatory scrutiny of leading AI developers and their potential impact on consumers. The FTC has broad authority to act against companies whose products or conduct may harm the public.

0 comments Read more at NYT Technology

TechnologyTechCrunch ·

Instinct Rolls Out Curated Recommendations, Drawing User Backlash

Instinct is introducing a new feature offering human-curated product and travel recommendations to its users. The rollout has been met with mixed reactions, with a portion of users expressing dissatisfaction. The core complaint centers on receiving unsolicited suggestions that users did not opt into or request. The move marks a shift in Instinct's platform experience, though the company has not publicly addressed the negative feedback.

0 comments Read more at TechCrunch

TechnologyArs Technica ·

Google develops watermarking method for AI-generated protein designs

Google has developed a technique to embed watermarks into proteins designed by artificial intelligence. The method is intended to strengthen biosecurity by helping track and identify AI-generated biological molecules. It is compatible with a widely used AI protein design tool, making it broadly applicable in the field. The watermarking approach could help researchers and regulators distinguish AI-designed proteins from naturally occurring ones.

0 comments Read more at Ars Technica

TechnologyTechCrunch ·

FedEx places $300M order for 2,000 electric trucks from startup Harbinger

FedEx has agreed to purchase 2,000 electric trucks from Harbinger in a deal valued at approximately $300 million. This marks the largest single order in Harbinger's history as a company. The agreement signals FedEx's continued push to electrify its delivery fleet. The deal comes at a significant moment for Harbinger, which is reportedly exploring the possibility of an initial public offering.

0 comments Read more at TechCrunch

TechnologyThe Verge ·

Meta's Muse AI Agent: Capable but Controversial Amid Privacy and Security Concerns

Meta has launched Muse, an AI agent designed to assist users with tasks such as sending emails and making online purchases, requiring access to personal data and payment information. The platform has expanded rapidly, with planned integrations into Meta smart glasses, a Mac app, and a Tamagotchi-style hardware device called the Muse Charm. However, Muse has faced significant scrutiny, including a security exploit that allowed attackers to hijack the AI agent and an incident where a user's address was mistakenly shared with a stranger. Meta has also announced an Enterprise Platform to bring Muse capabilities to businesses, while some third parties like Amazon have moved to block the agent. The rollout reflects Meta's broader push to compete in the AI race, even as critics raise concerns about the risks of granting the company deeper access to users' digital lives.

0 comments Read more at The Verge

TechnologyArs Technica ·

New PS5 Exploit Unlocks Nearly All Console Versions

A significant new exploit has been developed for the PlayStation 5, marking a major advancement in the console's hacking scene. The vulnerability works on all PS5 units except those running the most recently released firmware update. This means a large portion of existing PS5 consoles in the wild are potentially affected. Users who have kept their consoles updated to the latest software version are currently protected from the exploit.

0 comments Read more at Ars Technica

TechnologyArs Technica ·

Discord Expands Advertising Platform, Opens Targeting Tools to All Advertisers

Discord is broadening access to its advertising platform by welcoming advertisers of all budget sizes. The move marks a significant shift for the chat-focused platform, which has historically been more reserved about commercial advertising. New targeting tools are being made available to a wider range of brands and marketers. The expansion signals Discord's intent to grow its revenue streams beyond subscriptions and other existing monetization methods.

0 comments Read more at Ars Technica

TechnologyArs Technica ·

Butterfly Wing Patterns Create Optical Illusion That Confuses Predators

New research reveals that butterfly wing patterns produce a 'motion dazzle' effect similar to the rotating stripes of a barbershop pole. The visual illusion makes it difficult for predators to accurately judge a butterfly's speed and direction during flight. The combination of distinctive wing markings and flight dynamics work together to generate this disorienting effect. Scientists believe this serves as an evolutionary defense mechanism, giving butterflies a survival advantage against would-be attackers.

0 comments Read more at Ars Technica

TechnologyThe Verge ·

Google pilots payments to publishers for AI search content contributions

Google has launched a pilot program paying select publishers whose content contributes to its AI-powered search features, according to reports from The Information and Digiday. The program involves roughly 100 publishers and began less than a year ago. Participating publishers are compensated based on how much their content feeds into AI Overviews, AI Mode in Search, and the Gemini chatbot. The initiative comes amid growing scrutiny over AI search features reducing traffic to publisher websites. At least one early participant reportedly earned more than $1 million through the arrangement.

0 comments Read more at The Verge

TechnologyTechCrunch ·

Trump and AI Leaders Sign Pledge Containing Misspelling of 'United States'

A formal pledge signed by President Donald Trump and prominent artificial intelligence industry leaders contained a misspelling of the name 'United States.' The error appeared in an official document bearing the signatures of some of the most powerful figures in both government and the tech sector. The mistake was notable given the high-profile nature of the signatories and the document's significance. The incident drew attention and commentary, highlighting an embarrassing oversight in the preparation of the official pledge.

0 comments Read more at TechCrunch

← NewerPage 46 of 283Older →