Google has released (PDF) a new analysis of its AI's environmental impact, showing that it has cut the energy use of AI text queries by a factor of 33 over the past year. Each prompt now consumes about 0.24 watt-hours -- the equivalent of watching nine seconds of TV. An anonymous reader shares an excerpt from an Ars Technica article: "We estimate the median Gemini Apps text prompt uses 0.24 watt-hours of energy, emits 0.03 grams of carbon dioxide equivalent (gCO2e), and consumes 0.26 milliliters (or about five drops) of water," they conclude. To put that in context, they estimate that the energy use is similar to about nine seconds of TV viewing. The bad news is that the volume of requests is undoubtedly very high. The company has chosen to execute an AI operation with every single search request, a compute demand that simply didn't exist a couple of years ago. So, while the individual impact is small, the cumulative cost is likely to be considerable. The good news? Just a year ago, it would have been far, far worse. Some of this is just down to circumstances. With the boom in solar power in the US and elsewhere, it has gotten easier for Google to arrange for renewable power. As a result, the carbon emissions per unit of energy consumed saw a 1.4x reduction over the past year. But the biggest wins have been on the software side, where different approaches have led to a 33x reduction in energy consumed per prompt. The Google team describes a number of optimizations the company has made that contribute to this. One is an approach termed Mixture-of-Experts, which involves figuring out how to only activate the portion of an AI model needed to handle specific requests, which can drop computational needs by a factor of 10 to 100. They've developed a number of compact versions of their main model, which also reduce the computational load. Data center management also plays a role, as the company can make sure that any active hardware is fully utilized, while allowing the rest to stay in a low-power state. The other thing is that Google designs its own custom AI accelerators, and it architects the software that runs on them, allowing it to optimize both sides of the hardware/software divide to operate well with each other. That's especially critical given that activity on the AI accelerators accounts for over half of the total energy use of a query. Google also has lots of experience running efficient data centers that carries over to the experience with AI. The result of all this is that it estimates that the energy consumption of a typical text query has gone down by 33x in the last year alone.

Read more of this story at Slashdot.

Bluesky has blocked access to its service in Mississippi rather than comply with a new state law requiring age verification for all social media users. TechCrunch reports: In a blog post published on Friday, the company explains that, as a small team, it doesn't have the resources to make the substantial technical changes this type of law would require, and it raised concerns about the law's broad scope and privacy implications. Mississippi's HB 1126 requires platforms to introduce age verification for all users before they can access social networks like Bluesky. On Thursday, U.S. Supreme Court justices decided to block an emergency appeal that would have prevented the law from going into effect as the legal challenges it faces played out in the courts. As a result, Bluesky had to decide what it would do about compliance. Instead of requiring age verification before users could access age-restricted content, this law requires age verification of all users. That means Bluesky would have to verify every user's age and obtain parental consent for anyone under 18. The company notes that the potential penalties for noncompliance are hefty, too -- up to $10,000 per user. Bluesky also stresses that the law goes beyond child safety, as intended, and would create "significant barriers that limit free speech and disproportionately harm smaller platforms and emerging technologies." To comply, Bluesky would have to collect and store sensitive information from all its users, in addition to the detailed tracking of minors. This is different from how it's expected to comply with other age verification laws, like the U.K.'s Online Safety Act (OSA), which only requires age checks for certain content and features. Mississippi's law blocks anyone from using the site unless they provide their personal and sensitive information. The company notes that its decision only applies to the Bluesky app built on the AT Protocol. Other apps may approach the decision differently.

Read more of this story at Slashdot.

At its upcoming Connect conference next month, Meta is rumored to unveil its first consumer-ready smart glasses with a built-in display, alongside a neural wristband controller. The $800 device, codenamed Hypernova, will be able to show simple visual content like texts and support AI assistant interactions. CNBC reports: Connect is a two-day conference for developers focused on virtual reality, AR and the metaverse. It was originally called Oculus Connect and obtained its current moniker after Facebook changed its parent company name to Meta in 2021. The glasses are internally codenamed Hypernova and will include a small digital display in the right lens of the device, said the people, who asked not to be named because the details are confidential. The device is expected to cost about $800 and will be sold in partnership with EssilorLuxottica, the people said. CNBC reported in October that Meta was working with Luxottica on consumer glasses with a display. [...] With Hypernova, Meta will finally be offering glasses with a display to consumers, but the company is setting low expectations for sales, some of the sources said. That's because the device requires more components than its voice-only predecessors, and will be slightly heavier and thicker, the people said. [...] Although Hypernova will feature a display, those visual features are expected to be limited, people familiar with the matter said. They said the color display will offer about a 20 degree field of view -- meaning it will appear in a small window in a fixed position -- and will be used primarily to relay simple bits of information, such as incoming text messages. The Hypernova glasses will also come paired with a wristband that will use technology built by Meta's CTRL Labs, said people familiar with the matter. CTRL Labs, which Meta acquired in 2019, specializes in building neural technology that could allow users to control computing devices using gestures in their arms. [...] In addition to Hypernova and the wristband, Meta will also announce a third-generation of its voice-only smart glasses with Luxottica at Connect, one person said.

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Register: Microsoft has reportedly stopped giving Chinese companies proof-of-concept exploit code for soon-to-be-disclosed vulnerabilities following last month's SharePoint zero-day attacks, which appear to be related to a leak in Redmond's early-bug-notification program. The software behemoth gives some software vendors early bug disclosures under its Microsoft Active Protections Program (MAPP), which typically delivers info two weeks before Patch Tuesday. MAPP participants sign a non-disclosure agreement, and in exchange get vulnerability details so that they can provide updated protections to customers more quickly. According to Microsoft spokesperson David Cuddy, who spoke with Bloomberg about changes to the program, MAPP has begun limiting access to companies in "countries where they're required to report vulnerabilities to their governments," including China. Companies in these countries will no longer receive "proof of concept" exploit code, but instead will see "a more general written description" that Microsoft sends at the same time as patches, Cuddy told the news outlet. "A leak happened here somewhere," Dustin Childs, head of threat awareness at Trend Micro's Zero Day Initiative (ZDI), told The Register in July. "And now you've got a zero-day exploit in the wild, and worse than that, you've got a zero-day exploit in the wild that bypasses the patch, which came out the next day." Childs said the MAPP change "is a positive change, if a bit late. Anything Microsoft can do to help prevent leaks while still offering MAPP guidance is welcome." "In the past, MAPP leaks were associated with companies out of China, so restricting information from flowing to these companies should help," Childs said. "The MAPP program remains a valuable resource for network defenders. Hopefully, Microsoft can squelch the leaks while sending out the needed information to companies that have proven their ability (and desire) to protect end users."

Read more of this story at Slashdot.

Waymo has received its first permit from the New York City Department of Transportation to begin testing autonomous vehicles in Manhattan and Downtown Brooklyn, marking the city's first official rollout of self-driving car trials. The program will initially deploy up to eight vehicles with safety drivers through late September, with the potential to extend and expand into other boroughs. CNBC reports: New York state law requires the company to have a driver behind the wheel to operate. "We're a tech-friendly administration and we're always looking for innovative ways to safely move our city forward," [Mayor Eric Adams] said in a release. "New York City is proud to welcome Waymo to test this new technology in Manhattan and Brooklyn, as we know this testing is only the first step in moving our city further into the 21st century." The news comes just two months after the company said it filed permits to test its cars in the city with a trained specialist behind the wheel. [...] As part of the permit, Waymo must regularly meet and report data to DOT and work closely with law enforcement and emergency services.

Read more of this story at Slashdot.

Google has signed a six-year cloud computing deal with Meta worth over $10 billion, making it the second major partnership after a recent agreement with OpenAI. The deal will see Meta rely on Google Cloud's infrastructure to support its massive AI data center buildout, as the company ramps up capital spending into the tens of billions. The Information (paywalled) first reported the deal.

Read more of this story at Slashdot.

Clean-energy growth helped China's carbon dioxide (CO2) emissions fall by 1% year-on-year in the first half of 2025, extending a declining trend that started in March 2024. From a report: The CO2 output of the nation's power sector -- its dominant source of emissions -- fell by 3% in the first half of the year, as growth in solar power alone matched the rise in electricity demand. The new analysis for Carbon Brief shows that record solar capacity additions are putting China's CO2 emissions on track to fall across 2025 as a whole. Other key findings include: The growth in clean power generation, some 270 terawatt hours (TWh) excluding hydro, significantly outpaced demand growth of 170TWh in the first half of the year. Solar capacity additions set new records due to a rush before a June policy change, with 212 gigawatts (GW) added in the first half of the year. This rush means solar is likely to set an annual record for growth in 2025, becoming China's single-largest source of clean power generation in the process. Coal-power capacity could surge by as much as 80-100GW this year, potentially setting a new annual record, even as coal-fired electricity generation declines. The use of coal to make synthetic fuels and chemicals is growing rapidly, climbing 20% in the first half of the year and helping add 3% to China's CO2 since 2020. The coal-chemical industry is planning further expansion, which could add another 2% to China's CO2 by 2029, making the 2030 deadline for peaking harder to meet.

Read more of this story at Slashdot.

An anonymous reader quotes a report from the BBC: A lawyer representing the online message board 4chan says it won't pay a proposed fine by the UK's media regulator as it enforces the Online Safety Act. According to Preston Byrne, managing partner of law firm Byrne & Storm, Ofcom has provisionally decided to impose a 20,000-pound fine "with daily penalties thereafter" for as long as the site fails to comply with its request. "Ofcom's notices create no legal obligations in the United States," he told the BBC, adding he believed the regulator's investigation was part of an "illegal campaign of harassment" against US tech firms. "4chan has broken no laws in the United States -- my client will not pay any penalty," Mr Byrne said. Ofcom began investigating 4chan over whether it was complying with its obligations under the UK's Online Safety Act. Then in August, it said it had issued 4chan with "a provisional notice of contravention" for failing to comply with two requests for information. Ofcom said its investigation would examine whether the message board was complying with the act, including requirements to protect its users from illegal content. "American businesses do not surrender their First Amendment rights because a foreign bureaucrat sends them an email," law firms Byrne & Storm and Coleman Law wrote. "Under settled principles of US law, American courts will not enforce foreign penal fines or censorship codes. If necessary, we will seek appropriate relief in US federal court to confirm these principles." The statement calls on the Trump administration to intervene and protect American businesses from "extraterritorial censorship mandates."

Read more of this story at Slashdot.

An anonymous reader shares a report: Expect record-breaking temperatures to change the workplace, the World Health Organization (WHO) and World Meteorological Organization (WMO) warned today in a new report. When workers don't have adequate protections from heat stress, their health and productivity suffer. It's a risk employers and lawmakers have to take more seriously if they want to keep workers safe and businesses prosperous, the agencies say. That means finding ways to adapt in a warming world, and paying close attention to groups that might be more vulnerable than others. [...] More than 2.4 billion people around the world -- 71 percent of the working population -- experience workplace heat stress, according to estimates from the ILO. Each year, 22.85 million occupational injuries and 18,970 fatalities are linked to excessive heat at work. The report also says that worker productivity falls 2-3 percent with every degree increase above 20 degrees Celsius in wet-bulb globe temperature, a measure that takes humidity and other environmental factors into account.

Read more of this story at Slashdot.

Europe's share of global economic output has fallen from 33% to 23% since 2005 while its space launch capacity has nearly collapsed, launching just four rockets this year compared to over 100 for the United States and 40 for China. The continent's economic stagnation spans 15 years -- likely the longest streak since the Industrial Revolution according to Deutsche Bank calculations -- with Germany's economy growing just 1% since late 2017 versus 19% US growth. Per capita GDP gaps have widened dramatically: $86,000 annually in the US versus $56,000 in Germany and $53,000 in the UK. Industrial electricity costs have become prohibitive, running three times higher in Germany and four times higher in the UK than American rates. "America innovates, China imitates, Europe regulates," Italian Prime Minister Giorgia Meloni observed. The continent's largest company by market value, SAP, now ranks just 28th globally. Further reading: The Technology Revolution is Leaving Europe Behind.

Read more of this story at Slashdot.

President Donald Trump said on Friday the U.S. would take a 10% stake in Intel under a deal with the struggling chipmaker and is planning more such moves, the latest extraordinary intervention by the White House in corporate America. Reuters: The development follows a meeting between CEO Lip-Bu Tan and Trump earlier this month that was sparked by Trump's demand for the Intel chief's resignation over his ties to Chinese firms.

Read more of this story at Slashdot.

Organizations still using default Microsoft 365 email domains face severe throttling starting this October. The restrictions target the onmicrosoft.com domain that Microsoft 365 automatically assigns to new tenants, limiting external messages to 100 recipients per day starting October 15. Microsoft blames spammers who exploit new tenants for quick spam bursts before detection. Affected organizations must acquire custom domains and update primary SMTP addresses across all mailboxes -- a process that requires credential updates across devices and applications.

Read more of this story at Slashdot.

Apple is in early discussions about using Google Gemini to power a revamped version of the Siri voice assistant, marking a key potential step toward outsourcing more of its artificial intelligence technology. From a report: The iPhone maker recently approached Alphabet's Google to explore building a custom AI model that would serve as the foundation of the new Siri next year, according to people familiar with he matter. Google has started training a model that could run on Apple's servers, said the people, who asked not to be identified because the discussions are private. The work is part of an effort to catch up in generative AI, a field where the company arrived late and then struggled to gain traction. Earlier this year, Apple also explored partnerships with Anthropic and OpenAI, weighing whether Claude or ChatGPT could serve as Siri's new brain. Apple is still several weeks away from making a decision on whether to continue using internal models for Siri or move to a partner. And it hasn't yet determined who that partner may be.

Read more of this story at Slashdot.

BrianFagioli writes: Google is preparing to bring its television platforms in line with the rest of Android. Starting August 1, 2026, both Google TV and Android TV will require app updates that include native code to provide 64-bit support. The move follows similar requirements for phones and tablets, and it paves the way for upcoming 64-bit TV devices.

Read more of this story at Slashdot.

An anonymous reader shares a report: The National Oceanic and Atmospheric Administration is narrowing the capabilities and reducing the number of next-generation weather and climate satellites it plans to build and launch in the coming decades, two people familiar with the plans told CNN. This move -- which comes as hurricane season ramps up with Erin lashing the East Coast -- fits a pattern in which the Trump administration is seeking to not only slash climate pollution rules, but also reduce the information collected about the pollution in the first place. Critics of the plan also say it's a short-sighted attempt to save money at the expense of understanding the oceans and atmosphere better. Two planned instruments, one that would measure air quality, including pollution and wildfire smoke, and another that would observe ocean conditions in unprecedented detail, are no longer part of the project, the sources said.

Read more of this story at Slashdot.

Remote work policies designed to attract top talent are becoming security vulnerabilities as state-sponsored hackers seek employment at cryptocurrency firms. Coinbase has implemented mandatory in-person orientation and US citizenship requirements for sensitive roles after detecting North Korean IT workers attempting to infiltrate the company through remote positions. CEO Brian Armstrong revealed on Stripe cofounder John Collison's podcast that the exchange now requires fingerprinting and live video interviews after discovering coordinated efforts involving US-based facilitators who reship laptops and attend virtual interviews on behalf of foreign operatives.

Read more of this story at Slashdot.

An anonymous reader shares a report: As it tries to unseat Google, OpenAI is relying on search data from an unlikely source: Google. OpenAI has been using Google search results scraped from the web to help power ChatGPT responses, according to two people with knowledge of it. The Google search data helps answer ChatGPT queries on current events, such as news, sports and equity markets, one of the people said. OpenAI is getting the data from SerpApi, an eight-year-old web-scraping firm, which listed OpenAI as a customer on its website as recently as May last year. It removed the reference for reasons that couldn't be learned.

Read more of this story at Slashdot.

Professional services firms are engineering AI agents through massive prompt documents to automate complex knowledge work. KPMG Australia developed a 100-page prompt that transforms tax legislation and partner expertise into an agent producing comprehensive tax advice within 24 hours rather than the traditional two-week timeline. The TaxBot searches distributed internal documents and Australian tax code to generate 25-page draft reports after collecting four to five inputs from tax agents. Chief Digital Officer John Munnelly said the system operates on KPMG Workbench, a global platform combining retrieval-augmented generation with models from OpenAI, Microsoft, Google, Anthropic, and Meta.

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Guardian: Denmark is to stop charging VAT on books in an attempt to get more people reading. At 25%, the country's tax rate on books is the highest in the world, a policy the government believes is contributing to a growing "reading crisis." The culture minister, Jakob Engel-Schmidt, announced on Wednesday that the government would propose in its budget bill that the tax on books be removed. The move is expected to cost 330 million kroner ($51 million) a year. "This is something that I, as minister of culture, have worked for, because I believe that we must put everything at stake if we are to end the reading crisis that has unfortunately been spreading in recent years," Engel-Schmidt told the Ritzau news agency . "I am incredibly proud. It is not every day that one succeeds in convincing colleagues that such massive money should be spent on investing in the consumption and culture of the Danes." [...] "It is also about getting literature out there," said Engel-Schmidt. "That is why we have already allocated money for strengthened cooperation between the country's public libraries and schools, so that more children can be introduced to good literature." [...] If prices do not fall as a result of the measure, Engel-Schmidt said he would reconsider whether it was the right course of action. "I will of course monitor how prices develop. If it turns out that abolishing VAT only means that publishers' profits grow and prices do not fall, then we must consider whether it was the right thing to do," he said. Further reading: Denmark Ending Letter Deliveries Is a Sign of the Digital Times

Read more of this story at Slashdot.

Davis Lu, a former Eaton Corporation developer, has been sentenced to four years in prison for sabotaging his ex-employer's Windows network with malware and a custom kill switch that locked out thousands of employees once his account was disabled. The attack caused significant operational disruption and financial losses, with Lu also attempting to cover his tracks by deleting data and researching privilege escalation techniques. BleepingComputer reports: After a corporate restructuring and subsequent demotion in 2018, the DOJ says that Lu retaliated by embedding malicious code throughout the company's Windows production environment. The malicious code included an infinite Java thread loop designed to overwhelm servers and crash production systems. Lu also created a kill switch named "IsDLEnabledinAD" ("Is Davis Lu enabled in Active Directory") that would automatically lock all users out of their accounts if his account was disabled in Active Directory. When his employment was terminated on September 9, 2019, and his account disabled, the kill switch activated, causing thousands of users to be locked out of their systems. "The defendant breached his employer's trust by using his access and technical knowledge to sabotage company networks, wreaking havoc and causing hundreds of thousands of dollars in losses for a U.S. company," said Acting Assistant Attorney General Matthew R. Galeotti. When he was instructed to return his laptop, Lu reportedly deleted encrypted data from his device. Investigators later discovered search queries on the device researching how to elevate privileges, hide processes, and quickly delete files. Lu was found guilty earlier this year of intentionally causing damage to protected computers. After his four-year sentence, Lu will also serve three years of supervised release following his prison term.

Read more of this story at Slashdot.