[Most Recent Entries] [Calendar View]
Wednesday, December 13th, 2023
| Time | Event |
| 12:02a | Microsoft Targets Nuclear To Power AI Operations According to the Wall Street Journal, Microsoft wants to use nuclear energy to power its artificial intelligence operations. And in order to help cut the red tape required to make that happen, Microsoft plans to use AI. From a report: A Microsoft team has spent months building an AI trained on nuclear regulations and licensing requirements to help the tech giant fill out all the applications it needs to build its own power plants. This typically takes years and millions, but Microsoft is urgently looking for more power to bring next-generation AI to life. That's because the larger the model and the more capable it becomes, the more power it requires. Microsoft today reflects the sensibilities of its founder, Bill Gates, in that the company believes in carbon-neutral energy sources -- and, like Gates who himself invests in nuclear power innovation, the company seems to see more potential in nuclear than other renewable sources of energy. "If we're going to do that carbon-free, we're going to need all the tools in the tool kit," Michelle Patron, Microsoft's senior director of sustainability policy, told the Journal. 
Read more of this story at Slashdot. |
| 12:25a | Apple's New iPhone Security Setting Keeps Thieves Out of Your Digital Accounts According to the Wall Street Journal, Apple is including new Stolen Device Protection in iOS 17.3 that requires authentication through Face ID or Touch ID to perform certain actions. The Verge reports: The new feature appears to come in response to the concerns raised in previous reports by The Wall Street Journal describing how thieves watch their victims type in their iPhone passcodes and then steal their devices. This gives thieves access to a trove of personal and financial information stored on the device, allowing them to lock victims out of their iCloud accounts and spend thousands of dollars using saved payment information. If you opt in to the feature, you would have to verify your identity with face or fingerprint biometrics when doing things like viewing your saved passwords in iCloud Keychain, applying for a new Apple Card, factory resetting your device, using saved payment methods in Safari, and turning off Lost Mode. This way, thieves wouldn't be able to steal your information even if they have your phone and the passcode. For even more sensitive actions, like changing your Apple ID password, changing your iPhone passcode, or turning off Find My, the new Stolen Device Protection feature adds an additional hurdle if the device is somewhere other than locations you often frequent, like at home or in the office. It requires you to not only verify your identity with Face ID or Touch ID but also wait one hour and then repeat the authentication process again.
Read more of this story at Slashdot. |
| 12:45a | Arena Group Fires CEO In Wake of Sports Illustrated AI Articles Scandal Last month, Futurism noticed Sports Illustrated was publishing AI-generated articles under fake author biographies. Although the magazine removed the articles in question and released a statement blaming the issue on a contractor, it wasn't enough to quell the widespread backlash. Today, Sports Illustrated's publisher, Arena Group, fired the sports media outlet's CEO. The Hill reports: The Arena Group, which publishes Sports Illustrated, said Monday that its board of directors had moved to fire CEO Ross Levinsohn. The board took the action to "improve the operational efficiency and revenue of the company," the company said. Manoj Bhargava will serve as Arena Group's interim chief executive officer effective this week. Levinsohn was the CEO of Arena Group since 2019 and previously was the publisher of the Los Angeles Times.
Read more of this story at Slashdot. |
| 1:25a | Netflix Releases Viewing Numbers For 18,000 Titles For First Time For the first time, Netflix has released a comprehensive report of what people watched on the platform over a six month period. It includes hours viewed for every title, the premiere date for any Netflix show and movie, and whether a title was available globally. From the Hollywood Reporter: The list includes worldwide viewing for more than 18,000 movies and seasons of TV (18,214, to be exact) between January and June. Those 18,214 titles all had at least 50,000 hours of viewing over those six months, encompassing about 99 percent of all viewing on Netflix, vp strategy and analysis Lauren Smith told reporters during a presentation of the data on Tuesday. It is the deepest dive into viewing that Netflix (or any other streamer) has ever made public. Among the highlights: The Night Agent was the biggest title on Netflix in the first half of 2023, racking up 812.1 million hours of viewing. Season two of Ginny & Georgia was second at 665.1 million hours, followed by Korean drama The Glory (622.8 million hours). Wednesday ranked fourth at 507.7 million hours of viewing, despite being released in November 2022. The company is using total hours viewed in this report as a way to measure engagement by its users rather than the "view" formula (total viewing hours divided by running time) it employs to compare titles in its weekly top 10 lists. Original series and movies dominate the top of the chart, but Smith said the split between original and licensed titles was more even: About 55 percent of viewing was for originals and 45 percent was for licensed shows and films. Suits, which dominated the Nielsen U.S. streaming charts for much of the summer and fall, had a combined 599 million hours of viewing worldwide on Netflix across all nine seasons. The show's first season ranked highest, coming in 67th place with 129.1 million hours. At the other end, a little more than 20 percent of the titles on Netflix's list (3,813 in all) had very little viewing. The company rounded them to 100,000 hours but they would fall between 50,000 and 149,999 hours -- barely a drop in the streamer's more than 100 billion total hours of viewing for the six months. The full "What We Watched: A Netflix Engagement Report" can be downloaded here.
Read more of this story at Slashdot. |
| 2:02a | Meet Ashley, the World's First AI-Powered Political Campaign Caller An artificial intelligence campaign volunteer named Ashley is being used to call thousands of Pennsylvania voters on behalf of Democrat Shamaine Daniels, "ushering in a new era of political campaigning in which candidates use technology to engage with voters in ways increasingly difficult to track," reports Reuters. From the report: Like a seasoned campaign volunteer, Ashley analyzes voters' profiles to tailor conversations around their key issues. Unlike a human, Ashley always shows up for the job, has perfect recall of all of Daniels' positions, and does not feel dejected when she's hung up on. "This is going to scale fast," said 30-year-old Ilya Mouzykantskii, the London-based CEO of Civox, the company behind Ashley. "We intend to be making tens of thousands of calls a day by the end of the year and into the six digits pretty soon. This is coming for the 2024 election and it's coming in a very big way. ... The future is now." For Daniels, the tool levels the playing field: as the underdog, she is now armed with another way to understand voters better, reach out in different languages (Ashley is fluent in over 20), and conduct many more "high bandwidth" conversations. Mouzykantskii said he is fully aware of the potential downsides, and does not intend to take any venture capital funding which might entice him to prioritize profits over ethics. Mouzykantskii and his co-founder Adam Reis, former computer science students at Stanford and Columbia Universities respectively, declined to disclose the exact generative AI models they are using. They will only say they use over 20 different AI models, some proprietary and some open-source. Thanks to the latest generative AI technologies, Reis was able to build the product almost entirely on his own, whereas several years ago it would have taken a team of 50 engineers several years to do so, he said. The report notes that there are "few legal guardrails" regulating this particular use of AI. "No rules directly apply to what Civox is doing. Federal Trade Commission regulations ban telemarketers from making robocalls to people on the Do Not Call Registry, but the list does not apply to political calls -- and Civox's activity, with its 'personalized' messages, does not qualify as robocalling."
Read more of this story at Slashdot. |
| 3:30a | Solar and Wind To Top Coal Power In US For First Time In 2024 An anonymous reader quotes a report from REVE News: The U.S. Energy Information Administration (EIA) expects, for the first year on record, combined electricity generation from wind and solar to surpass generation from coal in 2024. EIA expects solar generation in 2024 to increase 39% (228 kilowatthours) from 2023, driven by continued increases in solar capacity. "Renewables, particularly solar photovoltaics, are growing rapidly and making large contributions to electricity generation," DeCarolis said. EIA expects natural gas prices to be $2.77 per million British thermal units this winter, about 23% lower than previously forecast. The winter season is off to a warmer-than-expected start, so U.S. households are consuming less natural gas for heat than expected. The lower natural gas consumption is also contributing to rising U.S. natural gas inventories, which typically results in lower prices. "We're seeing record domestic natural gas production paired with lower-than-expected natural gas demand, and we expect that is going to push prices lower this winter season," DeCarolis said. EIA will publish its next STEO on January 9, 2024, including the agency's first forecasts for the energy sector through 2025. The full report is available on the EIA website.
Read more of this story at Slashdot. |
| 7:00a | Scientists Discover Nuclear Fission Amongst the Stars For the first time, scientists have discovered nuclear fission occurring amongst the stars, supporting the idea that neutron stars create "superheavy" elements when they collide, which then break down via nuclear fission to birth rare elements. Space.com reports: Nuclear fission is basically the opposite of nuclear fusion. While nuclear fusion refers to the smashing of lighter elements to create heavier elements, nuclear fission is a process that sees energy released when heavy elements split apart to create lighter elements. Nuclear fission is pretty well known, too. It's actually the basis of energy-generating nuclear power plants here on Earth -- however, it had not been seen occurring amongst the stars before now. The team of researchers led by North Carolina State University scientist Ian Roederer searched data concerning a wide range of elements in stars to discover the first evidence that nuclear fission could therefore be acting when neutron stars merge. These findings could help solve the mystery of where the universe's heavy elements come from. Scientists know that nuclear fusion is not just the primary source of energy for stars, but also the force that forges a variety of elements, the "heaviest" being iron. The evidence of nuclear fission discovered by [Matthew Mumpower, research co-author and a scientist at Los Alamos National Laboratory] and the team comes in the form of a correlation between "light precision metals," like silver, and "rare earth nuclei," like europium, showing in some stars. When one of these groups of elements goes up, the corresponding elements in the other group also increases, the scientists saw. The team's research also indicates that elements with atomic masses -- counts of protons and neutrons in an atomic nucleus -- greater than 260 may exist around neutron star smashes, even if this existence is brief. This is much heavier than many of the elements at the "heavy end" of the periodic table. "The only plausible way this can arise among different stars is if there is a consistent process operating during the formation of the heavy elements," Mumpower said. "This is incredibly profound and is the first evidence of fission operating in the cosmos, confirming a theory we proposed several years ago." "As we've acquired more observations, the cosmos is saying, 'hey, there's a signature here, and it can only come from fission.'" The research was published in the journal Science.
Read more of this story at Slashdot. |
| 10:00a | OpenAI's Nonprofit Arm Showed Revenue of $45,000 Last Year Despite being valued at $86 billion by private investors, OpenAI reported $44,485 in revenue in 2022, almost entirely from investment income. CNBC reports: That's from the nonprofit parent's 990 filing with the Internal Revenue Service, a form that has to be filled out by organizations wishing to maintain their tax-exempt status. Federal standards don't require audited financial statements from nonprofits. In its home state of California, OpenAI was able to avoid submitting audited financials for 2022 because the foundation's stated revenue was below the $2 million reporting threshold. The last time OpenAI filed with the state was 2017, when revenue was $33.2 million, or more than 700 times what the foundation reported for 2022. For all its talk of openness, OpenAI's financials remain a black box. Created as a nonprofit in 2015, OpenAI launched a so-called capped-profit entity in 2019, enabling it to raise billions of dollars in outside funding and attain attributes of a tech startup, such as the ability to hand out equity to employees. The for-profit side of the house went on to develop ChatGPT, the chatbot that took the world by storm late last year and kicked off the generative AI boom. [...] Thad Calabrese, a professor of public and nonprofit financial management at New York University, said OpenAI's current status is confusing, and is unlike anything he has seen in the nonprofit world. He said OpenAI could give up its nonprofit status, and he cited the Blue Cross Blue Shield Association, which in 1994 allowed associated nonprofit medical insurance plans to switch into for-profit entities. "There's no real need to have the nonprofit," Calabrese said. "If you want to be a startup, be a startup." Regarding OpenAI's reporting with the IRS, he said "fundamentally you can't really get a holistic sense of these organizations when you don't have consolidated financial statements."
Read more of this story at Slashdot. |
| 1:00p | Study Shows 38% of Java Apps Still Affected By Log4Shell An anonymous reader quotes a report from The Register: Two years after the Log4Shell vulnerability in the open source Java-based Log4j logging utility was disclosed, circa one in four applications are dependent on outdated libraries, leaving them open to exploitation. Research from security shop Veracode revealed that the vast majority of vulnerable apps may never have updated the Log4j library after it was implemented by developers as 32 percent were running pre-2015 EOL versions. Prior investigations from Veracode also showed that 79 percent of all developers never update third-party libraries after first introducing them into projects, and given that Log4j2 -- the specific version of Log4j affected by the vulnerability -- dates back to 2014, this could explain the large proportion of unpatched apps. A far smaller minority are running versions that were vulnerable at the time of the Log4j vulnerability's disclosure in December 2021. Only 2.8 percent are still using versions 2.0-beta9 through 2.15.0 -- post-EOL versions that remain exposed to Log4Shell, the industry-coined moniker of the vulnerability's exploit. Some 3.8 percent are still running version 2.17, a post-patch version of the Java logger that's not exposed to Log4Shell attacks, but is vulnerable to a separate remote code execution (RCE) bug (CVE-2021-44832). The researchers believe this illustrates a minority of developers that acted quickly when the vulnerability was first disclosed, as was the advice at the time, had returned to older habits of leaving libraries untouched. Altogether, just shy of 35 percent remain vulnerable to Log4Shell, and nearly 40 percent are vulnerable to RCE flaws. The EOL versions of Log4j are also vulnerable to three additional critical bugs announced by Apache, bringing the total to seven high and critical-rated issues. "At a surface level, the numbers above show that the massive effort to remediate the Log4Shell vulnerability was effective in mitigating risk of exploitation of the zero-day vulnerability. That should not be surprising," said Chris Eng, chief research officer at Veracode. "The bigger story at the two-year anniversary, however, is that there is still room for improvement when it comes to open source software security. If Log4Shell was another example in a long series of wake-up calls to adopt more stringent open source security practices, the fact that more than one in three applications currently run vulnerable versions of Log4j shows there is more work to do. "The major takeaway here is that organizations may not be aware of how much open source security risk they are exposed to and how to mitigate it."
Read more of this story at Slashdot. |
| 2:00p | UK Proposes Capping Some Visa, Mastercard Fees Regulators in the UK are weighing a cap on some of the fees that Visa and Mastercard charge local merchants for each card transaction, seeking to rein in charges that have risen fivefold since Brexit. From a report: After a monthslong review, the UK's Payment Systems Regulator said it's concerned that the payment giants have no effective competition, especially when it comes to the interchange fees they charge UK merchants when a consumer carrying a card issued by a bank in the European Economic Area makes an online purchase. For now, the PSR is proposing to restore those fees to the pre-Brexit levels of 0.3% of a purchase price for credit cards and 0.2% for debit cards. For credit cards, those fees have risen in recent years to as high as 1.5% and the PSR estimated that the increases cost UK businesses as much as $250 million last year. The two companies have been under fire from a bevy of regulators and lawmakers around the world for the fees they charge. While it usually amounts to just pennies per purchase, the fees do add up: US merchants spent a record $160.7 billion on swipe fees last year, up 16.7% from 2021, according to the Nilson Report, an industry publication.
Read more of this story at Slashdot. |
| 2:40p | COP28 Nations Agree for First Time To Transition Away From Fossil Fuels More than 190 governments at the United Nations climate conference approved an agreement Wednesday calling for the world to transition away from fossil fuels, an accord that bridged differences between big energy-producing nations and countries that want to completely phase out coal, oil and natural gas. From a report: The deal, the result of all-night talks, calls for "transitioning away from fossil fuels in energy systems, in a just, orderly and equitable manner." It says the shift to clean energy for the global economy should accelerate this decade with the aim of net-zero greenhouse gas emissions by 2050. Scientists say that is crucial to fulfilling the Paris accord, the landmark climate agreement that calls for governments to attempt to limit global warming to 1.5 degrees Celsius above preindustrial temperatures. The deal marks the first time a U.N. climate agreement has called for governments to cut back on all fossil fuels.
Read more of this story at Slashdot. |
| 3:20p | Every Homeopathic Eye Drop Should Be Pulled Off the Market, FDA Says An anonymous reader shares a report: This year has been marked by many terrifying things, but perhaps the most surprising of the 2023 horrors was ... eye drops. The seemingly innocuous teeny squeeze bottle made for alarming headlines numerous times during our current revolution around the sun, with lengthy lists of recalls, startling factory inspections, and ghastly reports of people developing near-untreatable bacterial infections, losing their eyes and vision, and dying. Recapping this unexpected threat to health, the Food and Drug Administration on Tuesday released an advisory titled "What You Should Know about Eye Drops" in hopes of keeping the dangers of this year from leaking into the next. Among the notable points from the regulator was this stark pronouncement: No one should ever use any homeopathic ophthalmic products, and every single such product should be pulled off the market. The point is unexpected, given that none of the high-profile infections and recalls this year involved homeopathic products. But, it should be welcomed by any advocates of evidence-based medicine.
Read more of this story at Slashdot. |
| 4:00p | Google Debuts Imagen 2 With Text and Logo Generation Google's making the second generation of Imagen, its AI model that can create and edit images given a text prompt, more widely available -- at least to Google Cloud customers using Vertex AI who've been approved for access. From a report: But the company isn't disclosing which data it used to train the new model -- nor introducing a way for creators who might've inadvertently contributed to the data set to opt out or apply for compensation. Called Imagen 2, Google's enhanced model -- which was quietly launched in preview at the tech giant's I/O conference in May -- was developed using technology from Google DeepMind, Google's flagship AI lab. Compared to the first-gen Imagen, it's "significantly" improved in terms of image quality, Google claims (the company bizarrely refused to share image samples prior to this morning), and introduces new capabilities including the ability to render text and logos. "If you want to create images with a text overlay -- for example, advertising -- you can do that," Google Cloud CEO Thomas Kurian said during a press briefing on Tuesday.
Read more of this story at Slashdot. |
| 4:45p | Apple Set to Be Hit by EU Antitrust Order in App Store Fight With Spotify Apple is set to be hit by a ban on its App Store rules that govern music-streaming rivals and a potential hefty fine in the European Union's latest attempt to limit the power of Big Tech. From a report: EU regulators are putting the finishing touches to a decision that would prohibit Apple's practice of blocking music services from pushing their users away from the App Store to alternative subscription options, according to people familiar with the investigation. The decision is slated for early next year, they added. As part of the upcoming decision, Apple runs the risk of a potential fine of as much as 10% of its annual sales -- although EU penalties seldom reach that level and orders for companies to change their business models can be more hard-hitting.
Read more of this story at Slashdot. |
| 5:20p | Your Smart TV Knows What You're Watching An anonymous reader shares a report: If you bought a new smart TV during any of the holiday sales, there's likely to be an uninvited guest watching along with you. The most popular smart TVs sold today use automatic content recognition (ACR), a kind of ad surveillance technology that collects data on everything you view and sends it to a proprietary database to identify what you're watching and serve you highly targeted ads. The software is largely hidden from view, and it's complicated to opt out. Many consumers aren't aware of ACR, let alone that it's active on their shiny new TVs. If that's you, and you'd like to turn it off, we're going to show you how. First, a quick primer on the tech: ACR identifies what's displayed on your television, including content served through a cable TV box, streaming service, or game console, by continuously grabbing screenshots and comparing them to a massive database of media and advertisements. Think of it as a Shazam-like service constantly running in the background while your TV is on. These TVs can capture and identify 7,200 images per hour, or approximately two every second. The data is then used for content recommendations and ad targeting, which is a huge business; advertisers spent an estimated $18.6 billion on smart TV ads in 2022, according to market research firm eMarketer. For anyone who'd rather not have ACR looking over their shoulder while they watch, we've put together a guide to turning it off on three of the most popular smart TV software platforms in use last year. Depending on the platform, turning off ACR took us between 10 and 37 clicks.
Read more of this story at Slashdot. |
| 6:05p | Apple Now Requires a Judge's Consent To Hand Over Push Notification Data Apple has said it now requires a judge's order to hand over information about its customers' push notification to law enforcement, putting the iPhone maker's policy in line with rival Google and raising the hurdle officials must clear to get app data about users. From a report: The new policy was not formally announced but appeared sometime over the past few days on Apple's publicly available law enforcement guidelines. It follows the revelation from Oregon Senator Ron Wyden that officials were requesting such data from Apple as well as from Google, the unit of Alphabet that makes the operating system for Android phones. Apps of all kinds rely on push notifications to alert smartphone users to incoming messages, breaking news, and other updates. These are the audible "dings" or visual indicators users get when they receive an email or their sports team wins a game. What users often do not realize is that almost all such notifications travel over Google and Apple's servers. In a letter first disclosed by Reuters last week, Wyden said the practice gave the two companies unique insight into traffic flowing from those apps to users, putting them "in a unique position to facilitate government surveillance of how users are using particular apps."
Read more of this story at Slashdot. |
| 6:40p | AMD Says Ryzen Threadripper 7000 Overclocking Triggers Hidden Fuse, Warranty Unaffected Overclocking AMD's Ryzen Threadripper 7000 series blows a fuse, indicating modification. However, AMD has told Tom's Hardware that this does not automatically invalidate the warranty of these top-tier workstation CPUs. From the report: "Threadripper 7000 Series processors do contain a fuse that is blown when overclocking is enabled. To be clear, blowing this fuse does not void your warranty. Statements that enabling an overclocking/overvolting feature will 'void' the processor warranty are not correct. Per AMD's standard Terms of Sale, the warranty excludes any damage that results from overclocking/overvolting the processor. However, other unrelated issues could still qualify for warranty repair/replacement," an AMD representative told Tom's Hardware. In summation, overclocking your Ryzen Threadripper Pro 7000 or non-Pro processor will not void the warranty -- only damages directly resulting from overclocking will. As always, AMD isn't against overclocking. If it was, the chipmaker wouldn't advertise overclocking support as one of the features of the WRX90 and TRX50 platforms. Only OEM systems lack overclocking support.
Read more of this story at Slashdot. |
| 7:20p | Polish Hackers Repaired Trains the Manufacturer Artificially Bricked. Now The Train Company Is Threatening Them Hackers unbricked a train in Poland that had been deliberately disabled by its manufacturer. Now the manufacturer is threatening legal action against the hackers despite evidence it sabotaged the trains. From a report: The manufacturer is also now demanding that the repaired trains immediately be removed from service because they have been "hacked," and thus might now be unsafe, a claim they also cannot substantiate. The situation is a heavy machinery example of something that happens across most categories of electronics, from phones, laptops, health devices, and wearables to tractors and, apparently, trains. In this case, NEWAG, the manufacturer of the Impuls family of trains, put code in the train's control systems that prevented them from running if a GPS tracker detected that it spent a certain number of days in an independent repair company's maintenance center, and also prevented it from running if certain components had been replaced without a manufacturer-approved serial number. This anti-repair mechanism is called "parts pairing," and is a common frustration for farmers who want to repair their John Deere tractors without authorization from the company. It's also used by Apple to prevent independent repair of iPhones.
Read more of this story at Slashdot. |
| 9:00p | Which AI Model Provides the 'Best' Answers? An anonymous reader quotes a report from Ars Technica: For those looking for a more rigorous way of comparing various models, the folks over at the Large Model Systems Organization (LMSys) have set up Chatbot Arena, a platform for generating Elo-style rankings for LLMs based on a crowdsourced blind-testing website. Chatbot Arena users can enter any prompt they can think of into the site's form to see side-by-side responses from two randomly selected models. The identity of each model is initially hidden, and results are voided if the model reveals its identity in the response itself. The user then gets to pick which model provided what they judge to be the "better" result, with additional options for a "tie" or "both are bad." Only after providing a pairwise ranking does the user get to see which models they were judging, though a separate "side-by-side" section of the site lets users pick two specific models to compare (without the ability to contribute a vote on the result). Since its public launch back in May, LMSys says it has gathered over 130,000 blind pairwise ratings across 45 different models (as of early December). Those numbers seem poised to increase quickly after a recent positive review from OpenAI's Andrej Karpathy that has already led to what LMSys describes as "a super stress test" for its servers. Chatbot Arena's thousands of pairwise ratings are crunched through a Bradley-Terry model, which uses random sampling to generate an Elo-style rating estimating which model is most likely to win in direct competition against any other. Interested parties can also dig into the raw data of tens of thousands of human prompt/response ratings for themselves or examine more detailed statistics, such as direct pairwise win rates between models and confidence interval ranges for those Elo estimates. Chatbot Arena's latest public leaderboard update shows a few proprietary models easily beating out a wide range of open-source alternatives. OpenAI's ChatGPT-4 Turbo leads the pack by a wide margin, with only an older GPT-4 model ("0314," which was discontinued in June) coming anywhere close on the ratings scale. But even months-old, defunct versions of GPT-3.5 Turbo outrank the highest-rated open-source models available in Chatbot Arena's testbed. Anthropic's proprietary Claude models also feature highly in Chatbot Arena's top rankings. Oddly enough, though, the site's blind human testing tends to rank the older Claude-1 slightly higher than the subsequent releases of Claude-2.0 and Claude-2.1. Among the tested non-proprietary models, the Llama-based Tulu 2 and 01.ai's Yi get rankings that are comparable to some older GPT-3.5 implementations. Past that, there's a slow but steady decline until you get to models like Dolly and StableLM at the bottom of the pack (amid older versions of many models that have more recent, higher-ranking updates on Chatbot Arena's charts).
Read more of this story at Slashdot. |
| 9:40p | OpenAI, Axel Springer Strike Unprecedented Deal To Offer News In ChatGPT OpenAI has struck a deal with Politico parent company Axel Springer, allowing ChatGPT to summarize news stories from Politico and Business Insider. CNBC reports: Once the OpenAI-Axel Springer deal goes into effect, when a user asks ChatGPT a question, it will respond with summaries of news articles from media outlets such as Politico, Business Insider, Bild and Welt. The chatbot will also include articles that would otherwise be limited to subscribers of those outlets, according to a release, and the answers will include "attribution and links to the full articles for transparency." The partnership follows a deal that OpenAI struck with the Associated Press in July, allowing it to license the AP's news archive for training data. As part of the agreement, Axel Springer will provide content from its media brands as training data for OpenAI's large language models, such as GPT-4, the AI model that helps power ChatGPT. The News Media Alliance, a trade group representing more than 2,200 publishers, released research in October suggesting that data sets used to train popular AI models rely "significantly" more on publisher content, outweighing it by a factor ranging from over five to almost 100, compared to generic web content.
Read more of this story at Slashdot. |
| 10:20p | GM Says It's Ditching Apple CarPlay, Android Auto For Your Safety Earlier this year, General Motors announced plans to phase out Apple CarPlay and Android Auto, shifting instead to built-in infotainment systems developed with Google. Now, the company has explained why it made that decision to MotorTrend: Tim Babbitt, GM's head of product for infotainment, gave MT a better explanation at a press event for the new Chevrolet Blazer EV, the flagship vehicle in the no CarPlay or Android Auto strategy. According to him, there's an important factor that didn't make it into the fact sheet: safety. Specifically, he cited driver distraction caused by cell phone usage behind the wheel. According to Babbitt, CarPlay and Android Auto have stability issues that manifest themselves as bad connections, poor rendering, slow responses, and dropped connections. And when CarPlay and Android Auto have issues, drivers pick up their phones again, taking their eyes off the road and totally defeating the purpose of these phone-mirroring programs. Solving those issues can sometimes be beyond the control of the automaker. You can start to see GM's frustration. Babbitt's thesis is that if drivers were to do everything through the vehicle's built-in systems, they'd be less likely to pick up their phones and therefore less distracted and safer behind the wheel. He admits, though, GM hasn't tested this thesis in the lab or real world yet but believes it has potential, if customers go for it. The issues Babbitt cited with CarPlay and Android Auto seem like they'd be mostly linked to using those programs wirelessly, and while he says that's true, just plugging the phone into a USB data port doesn't solve all the problems. Babbitt says even when using a physical connection, Android phones are prone to compatibility issues between the vehicle and all the various phone manufacturers running Android. iPhones, meanwhile, suffer from backwards compatibility issues that cause older iPhone models to have trouble running CarPlay consistently. He points to J.D. Power data that shows issues with CarPlay and Android Auto are common owner complaints, and that customers tend to blame the automaker rather than the phone manufacturer or phone software. In that way, eliminating CarPlay and Android Auto potentially relieves GM of a key customer complaint dragging down their perceived quality scores. After MotorTrend's story was published, GM issued the following statement: "We wanted to reach out to clarify that comments about GM's position on phone projection were misrepresented and to reinforce our valued partnerships with Apple and Google and each company's commitment to driver safety. GM's embedded infotainment strategy is driven by the benefits of having a system that allows for greater integration with the larger GM ecosystem and vehicles."
Read more of this story at Slashdot. |
| 11:00p | European Union Lawmakers Agree To New Rules That Bolster Gig Worker Rights An anonymous reader quotes a report from TechCrunch: Some two years of talking about gig worker rights later and European Union lawmakers have finally reached a deal on the final shape of the Platform Worker Directive. [...] The Commission presented its original plan to reform labor laws to boost protections for platform workers back in December 2021, setting out a presumption of employment for workers in a bid to flip the odds on gig economy exploitation. But the proposal proved contentious, with heavy industry lobbying from tech platforms such as Uber pushing for gig workers to be carved out of Europe's employment protections. There were also divisions between Member States over how much worker protection vs platform shielding they were prepared to commit to. But after a final trilogue, lasting more than 12 hours, a provisional agreement has been clinched. The deal that's been provisionally agreed means a presumption of an employment relationship between a gig worker and a platform will be triggered when two out of a list of five "indicators of control or direction are present," as the parliament's press release puts it. "This list can be expanded by Member States. The presumption can be triggered by the worker, by their representatives, and by the competent authorities on their own initiative. This presumption can be rebutted if the platform proves that the contractual relationship is not an employment relationship," it adds. The agreement also contains transparency provisions that will require platforms to provide information to individuals performing platform work (and to their representatives) about how the algorithms that manage them work; and how their behavior affects decisions taken by automated systems. [...] The provisionally agreed new rules will also ban platforms from taking "certain important decisions," such as dismissals or decisions to suspend an account, without human oversight. Per the parliament, the agreed text also ensures "more human oversight on the decisions of systems that directly affect the persons performing platform work"; and obliges platforms to "assess the impact of decisions taken or supported by automated monitoring and decision-making systems on working conditions, health and safety and fundamental rights". So conducting data protection impact assessments looks set to be a hard requirement for complying with the new law. Another prohibition that's been agreed is a ban on platforms from processing certain types of personal data of workers, including personal beliefs, private exchanges with colleagues, or when a worker is not at work -- with the Directive billed as beefing up data protection rights for platform workers. Other provisions in the provisional deal include a requirement for platforms to share information on self-employed workers in their employ with competent national authorities and representatives of those performing platform work, such as trade unions. Measures to prevent platforms from circumventing the rules by using intermediaries has also been agreement -- a practice that's stepped up considerably in Spain since the country introduced its own labor reform, back in 2021, with the aim of forcing platforms to hire delivery workers. Some key details of exactly what's been agreed remain under wraps -- and full visibility and analysis of the ramifications will likely have to wait for a consolidated text to emerge in the coming weeks/months. [...] The final text still needs to be voted on by the Council and Parliament before it can be adopted as pan-EU law. What implementation period has been agreed also isn't yet clear. But today's political deal signals the train has now left the station.
Read more of this story at Slashdot. |
| 11:40p | Xbox Cloud Gaming Is Now Available On Meta's Quest VR Headsets A beta version of the Xbox Cloud Gaming app is now available for the Meta Quest headsets, allowing you to stream hundreds of Xbox games with an Xbox Game Pass Ultimate subscription. The Verge reports: The beta app is available from the Meta Quest Store, and you'll simply need to pair a supported Bluetooth controller to start playing. You can use an Xbox controller (that supports Bluetooth), a PS4 one, or even Nintendo's Switch Pro controllers. Support for PS5 controllers is "coming in the future," according to Meta. There are a variety of display sizes for an immersive VR environment to stream Xbox games in or even an Xbox-themed virtual space on the latest Quest 3 and Pro headsets that takes advantage of full-color passthrough.
Read more of this story at Slashdot. |
| << Previous Day |
2023/12/13 [Calendar] |
Next Day >> |