An anonymous reader quotes a report from Ars Technica: On Wednesday, Google DeepMind announced two new AI models designed to control robots: Gemini Robotics and Gemini Robotics-ER. The company claims these models will help robots of many shapes and sizes understand and interact with the physical world more effectively and delicately than previous systems, paving the way for applications such as humanoid robot assistants. [...] Google's new models build upon its Gemini 2.0 large language model foundation, adding capabilities specifically for robotic applications. Gemini Robotics includes what Google calls "vision-language-action" (VLA) abilities, allowing it to process visual information, understand language commands, and generate physical movements. By contrast, Gemini Robotics-ER focuses on "embodied reasoning" with enhanced spatial understanding, letting roboticists connect it to their existing robot control systems. For example, with Gemini Robotics, you can ask a robot to "pick up the banana and put it in the basket," and it will use a camera view of the scene to recognize the banana, guiding a robotic arm to perform the action successfully. Or you might say, "fold an origami fox," and it will use its knowledge of origami and how to fold paper carefully to perform the task. In 2023, we covered Google's RT-2, which represented a notable step toward more generalized robotic capabilities by using Internet data to help robots understand language commands and adapt to new scenarios, then doubling performance on unseen tasks compared to its predecessor. Two years later, Gemini Robotics appears to have made another substantial leap forward, not just in understanding what to do but in executing complex physical manipulations that RT-2 explicitly couldn't handle. While RT-2 was limited to repurposing physical movements it had already practiced, Gemini Robotics reportedly demonstrates significantly enhanced dexterity that enables previously impossible tasks like origami folding and packing snacks into Zip-loc bags. This shift from robots that just understand commands to robots that can perform delicate physical tasks suggests DeepMind may have started solving one of robotics' biggest challenges: getting robots to turn their "knowledge" into careful, precise movements in the real world. DeepMind claims Gemini Robotics "more than doubles performance on a comprehensive generalization benchmark compared to other state-of-the-art vision-language-action models." Google is advancing this effort through a partnership with Apptronik to develop next-generation humanoid robots powered by Gemini 2.0. Availability timelines or specific commercial applications for the new AI models were not made available.

Read more of this story at Slashdot.

The abstract of a paper published on National Bureau of Economic Research: We document that, after remaining almost constant for almost 30 years, real labor productivity at U.S. restaurants surged over 15% during the COVID pandemic. This surge has persisted even as many conditions have returned to pre-pandemic levels. Using mobile phone data tracking visits and spending at more than 100,000 individual limited service restaurants across the country, we explore the potential sources of the surge. It cannot be explained by economies of scale, expanding market power, or a direct result of COVID-sourced demand fluctuations. The restaurants' productivity growth rates are strongly correlated, however, with reductions in the amount of time their customers spend in the establishments, particularly with a rising share of customers spending 10 minutes or less. The frequency of such 'take-out' customers rose considerably during COVID, even at fast food restaurants, and never went back down. The magnitude of the restaurant-level relationship between productivity and customer dwell time, if applied to the aggregate decrease in dwell time, can explain almost all of the aggregate productivity increase in our sample.

Read more of this story at Slashdot.

D-Wave researchers have published findings in Science demonstrating what they call "quantum supremacy" by showing their quantum annealers can solve problems beyond the reach of classical computers. The team, led by Andrew D. King, demonstrated area-law scaling of entanglement in model quench dynamics of two-, three- and infinite-dimensional spin glasses. The research shows quantum annealers rapidly generating samples that closely match solutions to the Schrodinger equation, supporting observed stretched-exponential scaling in matrix-product-state approaches. According to the paper, D-Wave's processors completed these magnetic materials simulations in under 20 minutes, while the same calculations would require nearly a million years on Oak Ridge National Laboratory's supercomputers. The claim hasn't gone unchallenged. Miles Stoudenmire from the Flatiron Institute's Center for Computational Quantum Physics argues that classical computers can achieve comparable results using methods developed since D-Wave's initial findings. "We're just saying, 'Look, this one problem at this one time didn't beat classical computers. Try again,'" Stoudenmire noted. The quantum computing community has increasingly shifted terminology from "supremacy" to "advantage" or "utility," focusing on solving practical business or scientific problems faster, more accurately, or more economically than classical alternatives.

Read more of this story at Slashdot.

Several Asian airlines have tightened restrictions on portable battery chargers amid growing concerns about fire risks, following a January blaze that destroyed an Air Busan aircraft in South Korea. South Korean airlines now require passengers to keep portable chargers within arm's reach rather than in overhead bins, a rule implemented March 1 to ease public anxiety, according to the Transportation Ministry. Taiwan's EVA Air and China Airlines have banned using or charging power banks on flights but still allow them in overhead compartments. Thai Airways announced a similar ban last Friday, citing "incidents of in-flight fires on international airlines." Battery-related incidents on U.S. airlines have increased from 32 in 2016 to 84 last year, with portable chargers identified as the most common culprit, according to Federal Aviation Administration data. The International Civil Aviation Organization has banned lithium-ion batteries from cargo holds since 2016, though no industry standard exists for regulating power banks.

Read more of this story at Slashdot.

Global sales of smartwatches have fallen for the first time, new figures indicate, in large part due to a sharp decline in the popularity of market leader, Apple. From a report: Market research firm Counterpoint says 7% fewer of the devices were shipped in 2024 compared to the year before. Shipments of Apple Watches fell by 19% in that period, Counterpoint says. It blames the slump on a lack of new features in Apple's latest devices, and the fact a rumoured high-end Ultra 3 model never materialised.

Read more of this story at Slashdot.

The Federal Trade Commission asked a judge in Seattle to delay the start of its trial accusing Amazon of duping consumers into signing up for its Prime program, citing resource constraints. CNBC: Attorneys for the FTC made the request during a status hearing on Wednesday before Judge John Chun in the U.S. District Court for the Western District of Washington. Chun had set a Sept. 22 start date for the trial. Jonathan Cohen, an attorney for the FTC, asked Chun for a two-month continuance on the case due to staffing and budgetary shortfalls. The FTC's request to delay due to staffing constraints comes amid a push by the Trump administration's Department of Government Efficiency to reduce spending. DOGE, which is led by tech baron Elon Musk, has slashed the federal government's workforce by more than 62,000 workers in February alone. "We have lost employees in the agency, in our division and on our case team," Cohen said.

Read more of this story at Slashdot.

Schools across the United States are increasingly using artificial intelligence to monitor students' online activities, raising significant privacy concerns after Vancouver Public Schools inadvertently released nearly 3,500 unredacted, sensitive student documents to reporters. The surveillance software, developed by companies like Gaggle Safety Management, scans school-issued devices 24/7 for signs of bullying, self-harm, or violence, alerting staff when potential issues are detected. Approximately 1,500 school districts nationwide use Gaggle's technology to track six million students, with Vancouver schools paying $328,036 for three years of service. While school officials maintain the technology has helped counselors intervene with at-risk students, documents revealed LGBTQ+ students were potentially outed to administrators through the monitoring.

Read more of this story at Slashdot.

IBM CEO Arvind Krishna has publicly disagreed with Anthropic CEO Dario Amodei's prediction that AI will write 90% of code within 3-6 months, estimating instead that only "20-30% of code could get written by AI." "Are there some really simple use cases? Yes, but there's an equally complicated number of ones where it's going to be zero," Krishna said during an onstage interview at SXSW. He argued AI will boost programmer productivity rather than eliminate jobs. "If you can do 30% more code with the same number of people, are you going to get more code written or less?" he asked. "History has shown that the most productive company gains market share, and then you can produce more products."

Read more of this story at Slashdot.

Roomba maker iRobot has warned it may cease operations within 12 months unless it can refinance debt or find a buyer, just one day after launching a new vacuum cleaner line. In its March 12 quarterly report, the company disclosed it had spent $3.6 million to amend terms on a $200 million Carlyle Group loan from 2023, as U.S. revenue plunged 47% in the fourth quarter. "Given these uncertainties and the implication they may have on the Company's financials, there is substantial doubt about the Company's ability to continue as a going concern for a period of at least 12 months from the date of the issuance of its consolidated 2024 financial statements," the company wrote. The robot vacuum pioneer has initiated a formal strategic review after a failed Amazon acquisition, the departure of founder Colin Angle, and layoffs affecting over half its workforce. iRobot cited mounting competition from Chinese manufacturers and expects continued losses for "the foreseeable future."

Read more of this story at Slashdot.

Morgan Stanley has reduced its iPhone shipment forecasts after Apple confirmed the delay of a more advanced Siri personal assistant, dampening prospects for accelerating phone upgrades. The investment bank now predicts 230 million iPhone shipments in 2025 (flat year-over-year) and 243 million in 2026 (up 6%), down from previous estimates. An upgraded Siri was the most sought-after Apple Intelligence feature among prospective buyers, according to the bank's survey data. "Access to Advanced AI Features" appeared as a top-five driver of smartphone upgrades for the first time, with about 50% of iPhone owners who didn't upgrade to iPhone 16 citing the delayed Apple Intelligence rollout as affecting their decision. The firm also incorporated headwinds from China tariffs in its assessment, noting Apple is unlikely to fully offset these costs without broader exemptions.

Read more of this story at Slashdot.

Amazon, Alphabet's Google and Meta Platforms on Wednesday said they support efforts to at least triple nuclear energy worldwide by 2050. From a report: The tech companies signed a pledge first adopted in December 2023 by more than 20 countries, including the U.S., at the U.N. Climate Change Conference. Financial institutions including Bank of America, Goldman Sachs and Morgan Stanley backed the pledge last year. The pledge is nonbinding, but highlights the growing support for expanding nuclear power among leading industries, finance and governments. Amazon, Google and Meta are increasingly important drivers of energy demand in the U.S. as they build out AI centers. The tech sector is turning to nuclear power after concluding that renewables alone won't provide enough reliable power for their energy needs. Microsoft and Apple did not sign the statement.

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Register: New York State has sued Allstate Insurance for operating websites so badly designed they would deliver personal information in plain-text to anyone that went looking for it. The data was lifted from Allstate's National General business unit, which ran a website for consumers who wanted to get a quote for a policy. That task required users to input a name and address, and once that info was entered, the site searched a LexisNexis Risk Solutions database for data on anyone who lived at the address provided. The results of that search would then appear on a screen that included the driver's license number (DLN) for the given name and address, plus "names of any other drivers identified as potentially living at that consumer's address, and the entire DLNs of those other drivers." Naturally, miscreants used the system to mine for people's personal information for fraud. "National General intentionally built these tools to automatically populate consumers' entire DLNs in plain text -- in other words, fully exposed on the face of the quoting websites -- during the quoting process," the court documents [PDF] state. "Not surprisingly, attackers identified this vulnerability and targeted these quoting tools as an easy way to access the DLNs of many New Yorkers," according to the lawsuit. The digital thieves then used this information to "submit fraudulent claims for pandemic and unemployment benefits," we're told. ... [B]y the time the insurer resolved the mess, crooks had built bots that harvested at least 12,000 individuals' driver's license numbers from the quote-generating site.

Read more of this story at Slashdot.

AmiMoJo shares a report from Electrek: The U.S. installed 50 gigawatts (GW) of new solar capacity in 2024, the largest single year of new capacity added to the grid by any energy technology in over two decades. That's enough to power 8.5 million households. According to the U.S. Solar Market Insight 2024 Year in Review report (PDF) released today by the Solar Energy Industries Association (SEIA) and Wood Mackenzie, solar and storage account for 84% of all new electric generating capacity added to the grid last year. In addition to historic deployment, surging U.S. solar manufacturing emerged as a landmark economic story in 2024. Domestic solar module production tripled last year, and at full capacity, U.S. factories can now produce enough to meet nearly all demand for solar panels in the U.S. Solar cell manufacturing also resumed in 2024, strengthening the U.S. energy supply chain. [...] Total US solar capacity is expected to reach 739 GW by 2035, but the report forecasts include scenarios showing how policy changes could impact the solar market. [...] The low case forecast shows a 130 GW decline in solar deployment over the next decade compared to the base case, representing nearly $250 billion of lost investment.

Read more of this story at Slashdot.

Longtime Slashdot reader schwit1 shares a report from Behind The Black: According to information at this tweet from anonymous sources, parts of Starship will likely require a major redesign due to the spacecraft's break-up shortly after stage separation on its last two test flights. These are the key take-aways, most of which focus on the redesign of the first version of Starship (V1) to create the V2 that flew unsuccessfully on those flights: - Hot separation also aggravates the situation in the compartment. - Not related to the flames from the Super Heavy during the booster turn. - This is a fundamental miscalculation in the design of the Starship V2 and the engine section. - The fuel lines, wiring for the engines and the power unit will be urgently redone. - The fate of S35 and S36 is still unclear. Either revision or scrap. - For the next ships, some processes may be paused in production until a decision on the design is made. - The team was rushed with fixes for S34, hence the nervous start. There was no need to rush. - The fixes will take much longer than 4-6 weeks. - Comprehensive ground testing with long-term fire tests is needed. [emphasis mine] It must be emphasized that this information comes from leaks from anonymous sources, and could be significantly incorrect. It does however fit the circumstances, and suggests that the next test flight will not occur in April but will be delayed for an unknown period beyond.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Ars Technica: On Tuesday, OpenAI unveiled a new "Responses API" designed to help software developers create AI agents that can perform tasks independently using the company's AI models. The Responses API will eventually replace the current Assistants API, which OpenAI plans to retire in the first half of 2026. With the new offering, users can develop custom AI agents that scan company files with a file search utility that rapidly checks company databases (with OpenAI promising not to train its models on these files) and navigate websites -- similar to functions available through OpenAI's Operator agent, whose underlying Computer-Using Agent (CUA) model developers can also access to enable automation of tasks like data entry and other operations. However, OpenAI acknowledges that its CUA model is not yet reliable for automating tasks on operating systems and can make unintended mistakes. The company describes the new API as an early iteration that it will continue to improve over time. Developers using the Responses API can access the same models that power ChatGPT Search: GPT-4o search and GPT-4o mini search. These models can browse the web to answer questions and cite sources in their responses. That's notable because OpenAI says the added web search ability dramatically improves the factual accuracy of its AI models. On OpenAI's SimpleQA benchmark, which aims to measure confabulation rate, GPT-4o search scored 90 percent, while GPT-4o mini search achieved 88 percent -- both substantially outperforming the larger GPT-4.5 model without search, which scored 63 percent. Despite these improvements, the technology still has significant limitations. Aside from issues with CUA properly navigating websites, the improved search capability doesn't completely solve the problem of AI confabulations, with GPT-4o search still making factual mistakes 10 percent of the time. Alongside the Responses API, OpenAI released the open source Agents SDK, providing developers free tools to integrate models with internal systems, implement safeguards, and monitor agent activities. This toolkit follows OpenAI's earlier release of Swarm, a framework for orchestrating multiple agents.

Read more of this story at Slashdot.

An anonymous reader quotes a report from TechCrunch: There's a power crunch looming as AI and cloud providers ramp up data center construction. But a new report suggests that a solution lies beneath their foundations. Advanced geothermal power could supply nearly two-thirds of new data center demand by 2030, according to an analysis by the Rhodium Group. The additions would quadruple the amount of geothermal power capacity in the U.S. -- from 4 gigawatts to about 16 gigawatts -- while costing the same or less than what data center operators pay today. In the western U.S., where geothermal resources are more plentiful, the technology could provide 100% of new data center demand. Phoenix, for example, could add 3.8 gigawatts of data center capacity without building a single new conventional power plant. Geothermal resources have enormous potential to provide consistent power. Historically, geothermal power plants have been limited to places where Earth's heat seeps close to the surface. But advanced geothermal techniques could unlock 90 gigawatts of clean power in the U.S. alone, according to the U.S. Department of Energy. [...] Because geothermal power has very low running costs, its price is competitive with data centers' energy costs today, the Rhodium report said. When data centers are sited similarly to how they are today, a process that typically takes into account proximity to fiber optics and major metro areas, geothermal power costs just over $75 per megawatt hour. But when developers account for geothermal potential in their siting, the costs drop significantly, down to around $50 per megawatt hour. The report assumes that new generating capacity would be "behind the meter," which is what experts call power plants that are hooked up directly to a customer, bypassing the grid. Wait times for new power plants to connect to the grid can stretch on for years. As a result, behind the meter arrangements have become more appealing for data center operators who are scrambling to build new capacity.

Read more of this story at Slashdot.

Sphere Entertainment Co, the company behind the Las Vegas Sphere, said they are considering opening scaled-down versions of the immersive venue in other cities. AV Magazine reports: While this has been been feasible for its high-profile residencies such as U2, the Eagles, Dead & Company and Anyma, smaller venues could attract a broader range of artists who might not have the budget or demand to fill the flagship Las Vegas location. By scaling down the size while retaining the signature technology, Sphere Entertainment Co can offer a similar spectacle at a more sustainable cost for artists and spectators. The possibility of mini-Spheres follows news that a full-scale venue will open in the UAE as a result of a partnership between Sphere Entertainment Co and the Department of Culture and Tourism -- Abu Dhabi. Beyond concerts, the Las Vegas Sphere has proven successful with immersive films such as V-U2: An Immersive Concert Film and the Sphere Expeience featuring Darren Aronofsky's Postcard from Earth, which In January passed 1,000 screenings. "As we enter a new fiscal year, we see significant opportunities to drive our Sphere business forward in Las Vegas and beyond," said Dolan. "We believe we are on a path toward realizing our vision for this next-generation medium and generating long-term shareholder value."

Read more of this story at Slashdot.

MojoKid writes: AMD just launched its latest flagship desktop processors, the Ryzen 9 9950X3D. Ryzen 9 9950X3D is a 16-core/32-thread, dual-CCD part with a base clock of 4.3GHz and a max boost clock of 5.7GHz. There's also 96MB of second-gen 3D V-Cache on board. Standard Ryzen 9000 series processors feature 32MB of L3 cache per compute die, but with the Ryzen 9 9950X3D, one compute die is outfitted with an additional 96MB of 3D V-Cache, bringing the total L3 up to 128MB (144MB total cache). The CCD outfitted with 3D V-Cache operates at more conservative voltages and frequencies, but the bare compute die is unencumbered. The Ryzen 9 9950X3D turns out to be a high-performance, no-compromise desktop processor. Its complement of 3D V-Cache provides tangible benefits in gaming, and AMD's continued work on the platform's firmware and driver software ensures that even with the Ryzen 9 9950X3D's asymmetrical CCD configuration, performance is strong across the board. At $699, it's not cheap but its a great CPU for gaming and content creation, and one of the most powerful standard desktop CPUs money can buy currently.

Read more of this story at Slashdot.