Google users have discovered that adding curse words to search queries disables the company's AI-powered overview feature. While Google's Gemini AI system typically avoids profanity, inserting expletives into search terms bypasses AI summaries and delivers traditional web results instead. Users can also disable AI overviews by adding "-ai" or other text strings after a minus sign to their queries.

Read more of this story at Slashdot.

OpenAI today launched o3-mini, a specialized AI reasoning model designed for STEM tasks that offers faster processing at lower costs compared to its predecessor o1-mini. The model, priced at $1.10 per million cached input tokens and $4.40 per million output tokens, performs fact-checking before delivering results to reduce errors in technical domains like physics and programming, the Microsoft-backed startup said. (A million tokens are roughly 750,000 words) OpenAI claims that its tests showed o3-mini made 39% fewer major mistakes than o1-mini on complex problems while delivering responses 24% faster. The model will be available through ChatGPT with varying access levels -- free users get basic access while premium subscribers receive higher query limits and reasoning capabilities.

Read more of this story at Slashdot.

An anonymous reader shares a report: At first blush, stock trading this week is hardly a paragon of the market-efficiency theory, an oft-romanticized idea in Economics 101. After all, big equity gauges plunged on Monday, spurred by fears of an AI model released a week earlier, before swiftly rebounding. A fresh academic paper suggests the rise of passive investing may be fueling these kind of fragile market moves. According to a study to be published in the prestigious American Economic Review, evidence is building that active managers are slow to scoop up stocks en masse when prices move away from their intrinsic worth. Thanks to this lethargic trading behavior and the relentless boom in benchmark-tracking index funds, the impact of each trade on prices gets amplified, explaining how sell orders, like on Monday perhaps, can induce broader equity gyrations.ÂAs a result, the financial landscape is proving less dynamic and more volatile in the era of Big Passive, according to authors at the UCLA Anderson School of Management, the Stockholm School of Economics and the University of Minnesota Carlson School of Management.

Read more of this story at Slashdot.

More than 8,000 of the world's most-cited scientists have at least one retraction, according to a database that links retractions to top-cited papers. From a report: An analysis of the database, published in PLOS Biology on 30 January, attempts to map the scale of retractions and understand how they manifest. "Not every retraction is a sign of misconduct," says John Ioannidis, an epidemiologist at Stanford University in California, who led the study. "But it is important to have a bird's eye view, across all scientific fields, [of] people who are most influential in science." Retracted papers had a higher number of self-citations than did non-retracted papers. And papers with higher co-authorship numbers were more likely to be retracted than those with fewer co-authors. [...] In the study, the authors split the most-cited scientists into two groups. The first featured the 217,097 authors who were among the top 2% most-cited in their fields over their careers. The second group comprised the 223,152 scientists who made up the top 2% for citation impact in 2023, the most recent year for which there were data. The authors found that 8,747 (4%) of the most highly cited researchers in 2023 had at least one retraction during their career, as did 7,083 (3.3%) of the researchers who were most-cited over their careers.

Read more of this story at Slashdot.

DeepSeek has emerged as the leading open-source AI model developer, surpassing Meta's Llama and Mistral, after releasing its latest model V3 with breakthrough cost efficiencies, research and consultancy firm SemiAnalysis reported on Friday. The Chinese startup, backed by hedge fund High-Flyer, reached this milestone through innovations in Multi-head Latent Attention technology, which cut inference costs by 93.3% versus standard methods. Despite offering services below cost to gain market share, its performance matches or exceeds OpenAI's GPT-4.

Read more of this story at Slashdot.

A study has found microplastic and nanoplastic pollution to be significantly higher in placentas from premature births than in those from full-term births. From a report: The levels were much higher than previously detected in blood, suggesting the tiny plastic particles were accumulating in the placenta. But the higher average levels found in the shorter pregnancies were a "big surprise" for the researchers, as longer terms could be expected to lead to more accumulation. Preterm birth is the leading cause of infant death worldwide, and the reasons for about two-thirds of all preterm births were unknown, said Dr Enrico Barrozo, of Baylor College of Medicine in Texas, US. The established link between air pollution and millions of premature births had spurred the research team to investigate plastic pollution. The new study only demonstrates an association between microplastics and premature births. Further research is needed in cell cultures and animal models to determine if the link is causal. Microplastics are known to cause inflammation in human cells, and inflammation is one of the factors that prompts the start of labour.

Read more of this story at Slashdot.

An anonymous reader shares a report: Taiwan's digital ministry said on Friday that government departments should not use Chinese startup DeepSeek's artificial intelligence (AI) service, saying that as the product is from China it represents a security concern. Democratically-governed Taiwan has long been wary of Chinese tech given Beijing's sovereignty claims over the island and its military and political threats against the government in Taipei. In a statement, Taiwan's Ministry of Digital Affairs said that government departments are not allowed to use DeepSeek's AI service to "prevent information security risks". "DeepSeek's AI service is a Chinese product, and its operation involves cross-border transmission and information leakage and other information security concerns, and is a product that jeopardises the country's information security," the ministry said.

Read more of this story at Slashdot.

Apple has filed an emergency motion [PDF] for a stay in the Google antitrust trial, warning that it faces "clear and substantial irreparable harm" if barred from participating in the case's remedies phase. The motion, filed on January 30, 2025, comes after Judge Amit Mehta denied Apple's request for limited intervention earlier in the week. Apple -- which makes more than $20 billion a year from Google to use the Android-maker's search engine on Safari -- argues that the U.S. Department of Justice's (DOJ) proposed remedy -- which includes a prohibition on "any contract between Google and Apple in which there would be anything exchanged of value" --would prevent it from negotiating agreements that benefit millions of users. Without the ability to fully participate, Apple contends it will be left as a "mere spectator" while the government pursues restrictions that directly impact its business interests. The company asserts that intervention is necessary to develop evidence, participate in discovery, and cross-examine witnesses regarding its market role and incentives. Apple also seeks access to trial records while its appeal is pending, including witness lists, depositions, and discovery materials, to ensure it can respond effectively if granted party status.

Read more of this story at Slashdot.

Nearly 100 journalists and other members of civil society using WhatsApp, the popular messaging app owned by Meta, were targeted by spyware owned by Paragon, an Israeli maker of hacking software, the company alleged today. From a report: The journalists and other civil society members were being alerted of a possible breach of their devices, with WhatsApp telling the Guardian it had "high confidence" that the users in question had been targeted and "possibly compromised." The company declined to disclose where the journalists and members of civil society were based, including whether they were based in the US. The company said it had sent Paragon a "cease and desist" letter and that it was exploring its legal options. WhatsApp said the alleged attacks had been disrupted in December and that it was not clear how long the targets may have been under threat.

Read more of this story at Slashdot.

Microsoft is charging business customers a $400 premium for Surface devices equipped with Intel's latest Core Ultra processors compared to models using Qualcomm's Arm-based chips, the company has disclosed. The Intel-powered Surface Pro tablet and Surface Laptop, starting at $1,499, come with a second-generation Core Ultra 5 processor featuring eight cores, 16GB of memory and 256GB storage. Comparable Qualcomm-based models begin at $1,099. The new Intel devices will be available to business customers from February 18, though versions with cellular connectivity will launch later. Consumer Surface devices will only be offered with Qualcomm processors. Microsoft also unveiled a USB 4 Dock supporting dual 4K displays and the Surface Hub 3, a conference room computer available in 50-inch or 85-inch touchscreen versions.

Read more of this story at Slashdot.

An anonymous reader quotes a report from MIT Technology Review: A Canadian startup called Xanadu has built a new quantum computer it says can be easily scaled up to achieve the computational power needed to tackle scientific challenges ranging from drug discovery to more energy-efficient machine learning. Aurora is a "photonic" quantum computer, which means it crunches numbers using photonic qubits -- information encoded in light. In practice, this means combining and recombining laser beams on multiple chips using lenses, fibers, and other optics according to an algorithm. Xanadu's computer is designed in such a way that the answer to an algorithm it executes corresponds to the final number of photons in each laser beam. This approach differs from one used by Google and IBM, which involves encoding information in properties of superconducting circuits. Aurora has a modular design that consists of four similar units, each installed in a standard server rack that is slightly taller and wider than the average human. To make a useful quantum computer, "you copy and paste a thousand of these things and network them together," says Christian Weedbrook, the CEO and founder of the company. Ultimately, Xanadu envisions a quantum computer as a specialized data center, consisting of rows upon rows of these servers. This contrasts with the industry's earlier conception of a specialized chip within a supercomputer, much like a GPU. [...] Xanadu's 12 qubits may seem like a paltry number next to IBM's 1,121, but Tiwari says this doesn't mean that quantum computers based on photonics are running behind. In his opinion, the number of qubits reflects the amount of investment more than it does the technology's promise. [...] Xanadu's next goal is to improve the quality of the photons in the computer, which will ease the error correction requirements. "When you send lasers through a medium, whether it's free space, chips, or fiber optics, not all the information makes it from the start to the finish," he says. "So you're actually losing light and therefore losing information." The company is working to reduce this loss, which means fewer errors in the first place. Xanadu aims to build a quantum data center, with thousands of servers containing a million qubits, in 2029. The company published its work on chip design optimization and fabrication in the journal Nature.

Read more of this story at Slashdot.

During the company's fourth-quarter earnings call Thursday, Intel co-CEO Michelle Johnston Holthaus announced that Intel has decided to cancel its Falcon Shores AI chip. Instead, it'll opt to use it as an internal test chip while shifting focus to Jaguar Shores for AI data center solutions. TechCrunch reports: "AI data center ... is an attractive market for us," Holthaus said during the call. "[B]ut I am not happy with where we are today. We're not yet participating in the cloud-based AI data center market in a meaningful way ... One of the immediate actions I have taken is to simplify our roadmap and concentrate our resources." The focus instead will be on Jaguar Shores, which Holthaus called Intel's opportunity to "develop a system-level solution at rack scale ... to address the AI data center more broadly." Holthaus tempered expectations for Falcon Shores last month, when she implied that it was an "iterative" step over the company's previous dedicated AI data center chip, Gaudi 3. "One of the things that we've learned from Gaudi is, it's not enough to just deliver the silicon," Holthaus said during Thursday's earnings call. "Falcon Shores will help us in that process of working on the system, networking, memory -- all those component[s]. But what customers really want is that full-scale rack solution, and so we're able to get to that with Jaguar Shores." "As I think about our AI opportunity, my focus is on the problems our customers are trying to solve, most notably the need to lower the cost and increase the efficiency of compute," Holthaus said. "As such, a one-size-fits-all approach will not work, and I can see clear opportunities to leverage our core assets in new ways to drive the most compelling total cost of ownership across the continuum."

Read more of this story at Slashdot.

A bipartisan group of legislators has reintroduced the AM Radio for Every Vehicle Act of 2025, aiming to mandate AM radio in all new vehicles at no additional cost. Adweek reports: The AM Radio for Every Vehicle Act was first introduced in May 2023. It continued to take on new co-sponsors through the fall of 2024. It was reintroduced as the AM Radio for Every Vehicle Act of 2025 with 62 cosponsors. Upper Midwest senators showing support for the bill include Sens. John Hoeven and Kevin Cramer R-North Dakota, and Sens. Amy Klobuchar and Tina Smith, DFL- Minnesota. If enacted, the bill would require the Department of Transportation to issue a rule requiring new vehicles to maintain access to broadcast AM radio at no additional cost to the consumer and provide small vehicle manufacturers at least four years after the date DOT issues the rule to comply. The act also requires automakers to inform consumers, during the period before the rule takes effect, that the vehicles do not maintain access to broadcast AM radio. "With 82 million Americans tuning in each month, AM radio delivers more than just emergency alerts," says the National Association of Broadcasters in a news release. "It connects communities through hyper-local content, including news, weather and diverse cultural programming," according to a news release from the National Association of Broadcasters."

Read more of this story at Slashdot.

Decreasingly hungry thrillseeker Weaponized Fun has second thoughts about the risk to which they're willing to expose their palate. "In addition to Budget Bytes mailing list not knowing who I am, I'm not sure they know what they're making. I'm having a hard time telling whether 'New Recipe 1' sounds more enticing than 'New Recipe 3.' I sure hope they remembered the ingredients."

 

An anonymous reader frets that "The Guardian claims an article is *more* than 7 years old (it's not, as of today, January 26)" Date math is hard.

 

"Oh snap!" cried The Beast in Black I feel like we've seen several errors like this from Firefox recently: problems with 0 and -1 as sentinel values.

 

Faithful contributor Michael R. doubled up on the FB follies this week; here's one. Says Michael "Those hard tech interviews at META really draw in the best talent."

 

Finally for this week, a confused Stewart found an increasingly rare type of classic Error'd. "Trying to figure out how to ignore as instructed, when there is no ignore option. Do I just ignore it?" For completeness, the options should also include Abort

 

[Advertisement] ProGet’s got you covered with security and access controls on your NuGet feeds. Learn more.

At an all hands meeting inside Meta Thursday, the company's co-founder and chief executive Mark Zuckerberg said he was increasingly careful about what he says internally at Meta. From a report: "Everything I say leaks. And it sucks, right?," Zuckerberg said. Meta made changes to the question-and-answer section of the company all hands meeting because of the leaks, Zuckerberg said, according to meeting audio obtained by 404 Media. "I want to be able to be able to talk about stuff openly, but I am also trying to like, well, we're trying to build stuff and create value in the world, not destroy value by talking about stuff that inevitably leaks," he said. So rather than take direct questions, the company used a "poll" system, where questions asked beforehand were voted on so that "main themes" of questions were addressed. "There are a bunch of things that I think are value-destroying for me to talk about, so I'm not going to talk about those. But I think it'll be good. You all can give us feedback later," he added. "Maybe it's just the nature of running a company at scale, but it's a little bit of a bummer."

Read more of this story at Slashdot.

An anonymous reader quotes a report from Ars Technica: The Video Game History Foundation has officially opened up digital access to a large portion of its massive archives today, offering fans and researchers unprecedented access to information and ephemera surrounding the past 50 years of the game industry. Today's launch of the VGHF Library comprises more than 30,000 indexed and curated files, including high-quality artwork, promotional material, and searchable full-text archives over 1,500 video game magazine issues. This initial dump of digital materials also contains never-before-seen game development and production archival material stored by the VGHF, such as over 100 hours of raw production files from the creation of the Myst series or Sonic the Hedgehog concept art and design files contributed by artist Tom Payne. In a blog post and accompanying launch video, VGHF head librarian Phil Salvador explains how today's launch is the culmination of a dream the organization has had since its launch in 2017. But it's also just the start of an ongoing process to digitize the VGHF's mountains of unprocessed physical material into a cataloged digital form, so people can access it "without having to fly to California." The VGHF doesn't require any special credentials or even a free account to access its archives, a fact that might be contributing to overloaded servers on this launch day. Despite those server issues, amateur researchers online are already sharing crucial library-derived information about the history of describing games as "immersive" or that one time Garfield ranked games in GamePro, for instance. Unfortunately, digital libraries cannot offer direct, playable access to retail video games due to DMCA restrictions, notes Ars. However, organizations like the VGHF "continue to challenge those copyright rules every three years," raising hope for future access.

Read more of this story at Slashdot.

Italy's data protection agency has blocked the Chinese AI chatbot DeekSeek after its developers failed to disclose how it collects user data or whether it is stored on Chinese servers. Reuters reports: DeepSeek could not be accessed on Wednesday in Apple or Google app stores in Italy, the day after the authority, known also as the Garante, requested information on its use of personal data. In particular, it wanted to know what personal data is collected, from which sources, for what purposes, on what legal basis and whether it is stored in China. The authority's decision -- aimed at protecting Italian users' data -- came after the Chinese companies that supply chatbot service to DeepSeek provided information that "was considered to totally insufficient," the authority said in a note on its website. The Garante added that the decision had "immediate effect" and that it had also opened an investigation. Thanks to new submitter axettone for sharing the news.

Read more of this story at Slashdot.

Apple reported a record-breaking first quarter of 2025 with $124.3 billion in revenue and $36.3 billion in profit, or $2.40 per diluted share, driven by strong growth in its services business. That's "compared to revenue of $119.6 billion and net quarterly profit of $33.9 billion, or $2.18 per diluted share, in the year-ago quarter," notes MacRumors. From the report: Apple set all-time records during the quarter for total revenue, earnings per share, and services revenue. Total revenue was up 4 percent year-over-year, while earnings per share rose by 10 percent. Services, Mac, and iPad revenue figures were all up significantly year-over-year, while iPhone and Wearables saw small declines. Gross margin for the quarter was 46.9 percent, compared to 45.9 percent in the year-ago quarter. Apple also declared a quarterly dividend payment of $0.25 per share, payable on February 13 to shareholders of record as of February 10. "Today Apple is reporting our best quarter ever, with revenue of $124.3 billion, up 4 percent from a year ago," said Tim Cook, Apple's CEO. "We were thrilled to bring customers our best-ever lineup of products and services during the holiday season. Through the power of Apple silicon, we're unlocking new possibilities for our users with Apple Intelligence, which makes apps and experiences even better and more personal. And we're excited that Apple Intelligence will be available in even more languages this April."

Read more of this story at Slashdot.

An anonymous reader quotes a report from 404 Media: Datasets aggregated on data.gov, the largest repository of U.S. government open data on the internet, are being deleted, according to the website's own information. Since Donald Trump was inaugurated as president, more than 2,000 datasets have disappeared from the database. As people in the Data Hoarding and archiving communities have pointed out, on January 21, there were 307,854 datasets on data.gov. As of Thursday, there are 305,564 datasets. Many of the deletions happened immediately after Trump was inaugurated, according to snapshots of the website saved on the Internet Archive's Wayback Machine. Harvard University researcher Jack Cushman has been taking snapshots of Data.gov's datasets both before and after the inauguration, and has worked to create a full archive of the data. "Some of [the entries link to] actual data," Cushman told 404 Media. "And some of them link to a landing page [where the data is hosted]. And the question is -- when things are disappearing, is it the data it points to that is gone? Or is it just the index to it that's gone?" For example, "National Coral Reef Monitoring Program: Water Temperature Data from Subsurface Temperature Recorders (STRs) deployed at coral reef sites in the Hawaiian Archipelago from 2005 to 2019," a NOAA dataset, can no longer be found on data.gov but can be found on one of NOAA's websites by Googling the title. "Stetson Flower Garden Banks Benthic_Covage Monitoring 1993-2018 -- OBIS Event," another NOAA dataset, can no longer be found on data.gov and also appears to have been deleted from the internet. "Three Dimensional Thermal Model of Newberry Volcano, Oregon," a Department of Energy resource, is no longer available via the Department of Energy but can be found backed up on third-party websites. [...] Data.gov serves as an aggregator of datasets and research across the entire government, meaning it isn't a single database. This makes it slightly harder to archive than any individual database, according to Mark Phillips, a University of Northern Texas researcher who works on the End of Term Web Archive, a project that archives as much as possible from government websites before a new administration takes over. "Some of this falls into the 'We don't know what we don't know,'" Phillips told 404 Media. "It is very challenging to know exactly what, where, how often it changes, and what is new, gone, or going to move. Saving content from an aggregator like data.gov is a bit more challenging for the End of Term work because often the data is only identified and registered as a metadata record with data.gov but the actual data could live on another website, a state .gov, a university website, cloud provider like Amazon or Microsoft or any other location. This makes the crawling even more difficult." Phillips said that, for this round of archiving (which the team does every administration change), the project has been crawling government websites since January 2024, and that they have been doing "large-scale crawls with help from our partners at the Internet Archive, Common Crawl, and the University of North Texas. We've worked to collect 100s of terabytes of web content, which includes datasets from domains like data.gov." [...] It is absolutely true that the Trump administration is deleting government data and research and is making it harder to access. But determining what is gone, where it went, whether it's been preserved somewhere, and why it was taken down is a process that is time intensive and going to take a while. "One thing that is clear to me about datasets coming down from data.gov is that when we rely on one place for collecting, hosting, and making available these datasets, we will always have an issue with data disappearing," Phillips said. "Historically the federal government would distribute information to libraries across the country to provide greater access and also a safeguard against loss. That isn't done in the same way for this government data."

Read more of this story at Slashdot.

A proposed class-action lawsuit accuses Amazon of secretly tracking consumers' movements through their cellphones via its Amazon Ads SDK embedded in third-party apps, allegedly collecting sensitive geolocation data without consent. The complaint, filed by a California resident in a San Francisco federal court, claims Amazon violated state laws on unauthorized computer access in the process. Reuters reports: This allegedly enabled Amazon to collect an enormous amount of timestamped geolocation data about where consumers live, work, shop and visit, revealing sensitive information such as religious affiliations, sexual orientations and health concerns. "Amazon has effectively fingerprinted consumers and has correlated a vast amount of personal information about them entirely without consumers' knowledge and consent," the complaint said. The complaint was filed by Felix Kolotinsky of San Mateo, California, who said Amazon collected his personal information through the "Speedtest by Ookla" app on his phone. He said Amazon's conduct violated California's penal law and a state law against unauthorized computer access, and seeks unspecified damages for millions of Californians.

Read more of this story at Slashdot.