The last time I used a set of Noble Audio earbuds, the company managed to pack five drivers into the tiny housings. Now it’s back with a new model, the FoKus Prestige Encore, and despite only having four drivers this time, the company still took plenty of cues from its line of pro-grade in-ear monitors (IEMs). The biggest downside is that this new set is still plenty pricey, hitting your wallet for $699.

Inside, the FoKus Prestige Encore has one 8mm dynamic driver, two balanced armature drivers and one 6mm planar magnetic driver. According to Noble, each of these components have different duties. The dynamic drivers handle bass while the midrange goes to the Knowles balanced armature drivers. Lastly, the planar magnetic driver is tasked with the treble to “reveal microdetail.” The company says it also designed new “triple-vented” nozzles to make airflow efficient and the seal consistent.

Noble Audio

Rather than using plastic or some other affordable material, Noble opted for solid wood for both the earbud housings and the charging case. Since this is the real thing and not a printed plate, each set of the FoKus Prestige Encore will have its own unique grain pattern — no two sets will look the same. What’s more, Noble reduced the overall size of the buds, which it says offers “a secure, fatigue-free fit.”

The FoKus Prestige Encore is equipped with hybrid active noice cancellation (ANC) and a transparency mode. The earbuds also support aptX Lossless, aptX Adaptive and LDAC for higher quality listening. For calls, dual noise suppression mics won’t overly process your voice for the sake of reducing background distractions, which Noble says will keep you sounding clear and natural in all settings.

Noble promises up to seven hours of use on a charge with ANC enabled or up to 10 hours with noise canceling turned off. When you factor in the case, you’re looking at up to 35 hours of total listening time. A quick-charge features will give you up to two hours of playback. The FoKus Prestige Encore also supports Audiodo for personalized calibration for each user’s hearing, and those profiles are stored on the earbuds for use on any device.

The Fokus Prestige Encore is available starting today for $699 (£649/€799).

This article originally appeared on Engadget at https://www.engadget.com/audio/headphones/noble-audios-latest-earbuds-have-a-four-driver-setup-and-a-solid-wood-design-090000651.html?src=rss

A new study challenges the Nobel-winning theory that the universe’s expansion is accelerating, suggesting instead that it may be slowing down as dark energy weakens — potentially leading to a future “big crunch” where the cosmos collapses back in on itself. “Our study shows that the universe has already entered a phase of decelerated expansion at the present epoch and that dark energy evolves with time much more rapidly than previously thought,” said Prof Young-Wook Lee, of Yonsei University in South Korea, who led the work. “If these results are confirmed, it would mark a major paradigm shift in cosmology since the discovery of dark energy 27 years ago.” The Guardian reports: The latest work focuses on the reliability of observations of distant supernovae (exploding stars) that led to the discovery of dark energy, work that was awarded the 2011 Nobel prize in physics. […] By estimating the ages of 300 host galaxies using a different method, the team concluded that there are simply variations in the properties of stars in the early universe that mean they produce, on average, fainter supernovae. Correcting for this systematic bias still results in an expanding universe, but suggests that the expansion has slowed down and that dark energy is waning, the analysis concluded. If dark energy keeps decreasing to the point where it becomes negative, the universe is theoretically predicted to end in a big crunch. The findings are published in Monthly Notices of the Royal Astronomical Society.

Read more of this story at Slashdot.

Canon just barged into the mid-range full-frame hybrid camera arena with a new entry. The 32.5MP EOS R6 III is a dramatic improvement over the EOS R6 II thanks to the higher resolution, faster shooting speeds and better video specs. However, it comes at a relatively high price and lacks a stacked or partially stacked sensor, unlike other cameras in this category. 

The new 32.5MP sensor provides a big boost in resolution compared to the R6 II’s 24MP chip. At the same time, the R6 III offers higher shooting speeds of 40 fps in burst mode with the electronic shutter, or 12 fps with the mechanical or first-curtain shutter. It also supports 20 frames (a half second) of pre-shooting so you won’t miss key moments for wildlife or sports shooting. 

Canon

The R6 III uses Canon’s fast and dependable Dual Pixel AF system and comes with the company’s latest AI tracking algorithms. It can operate in fairly dim conditions down to -6 EV and should be more responsive with Canon’s latest Digic X processor, pending our review. It offers subject tracking for people, animals and vehicles, along with an Auto mode that will pick one of those automatically. It also borrowed the “Register People Priority” feature from the R5 II, which lets you keep focus locked on specific subjects that you’ve previously memorized. 

The biggest update for the R6 III, though, is with video. Thanks to the higher resolution sensor, it can now capture up to 7K 60 fps RAW light video, 7K 30 fps “open gate” video and 4K at up to 120 fps, all with Canon’s C-Log2 and C-Log3 on tap. There’s a dizzying array of other video formats available (12 pages worth in the specs), with HEVC S, AVC-S, RAW, RAW Light and others, at resolutions up to 7K. All the AF subject detection features are available (vehicles, animals and people), and Canon is typically among the best for video AF in terms of speed and accuracy. 

Canon

Nobody expects any radical design changes in a Canon camera (the company tried that with the EOS R and it really didn’t work), so the R6 III kept the last model’s tried and true form factor. That includes two adjustment dials on the top and one on the rear, along with a joystick, photo/video selector, mode dial and a good assortment of programmable buttons. 

The rear display flips out as you’d expect for vlogging, but it doesn’t tilt like Panasonic’s S1 II — so it may obstruct the mic or headphone jacks, and isn’t as useful for low-angle photo shooting. The viewfinder has 3.69 million dots of resolution as before, the same as Sony’s A7 IV but less than Panasonic’s slightly more expensive Lumix S1 II. One big change is the addition of a CFexpress card slot that allows RAW video capture and faster burst speeds, along with an SD UHS II slot. The battery is the same as the one for EOS R5 II and allows up to 390 shots (CIPA rating) with the viewfinder enabled. 

Canon

Other features include waveform monitoring that will be much appreciated by pros, plus new focus speed algorithms borrowed from Canon’s cinema cameras that offer “natural, professional” behavior, the company wrote. Inputs include 3.5mm mic and headphone jacks, along with high-speed USB-C and a full-sized HDMI port. In-body stabilization has been boosted slight from 8 to 8.5 stops, matching the latest Panasonic models. 

Along with the EOS R6 III, Canon introduced some interesting new glass. The RF45mm F1.2 STM lens brings very high speeds and shallow depth of field to a much smaller and lighter .76 pound (346 gram) form factor — less than half the weight of Canon’s RF 50mm f/1.2L USM lens. It’s also shockingly cheap for an f/1.2 lens at $470. 

The Canon EOS R6 III arrives later this month for $2,799 (body only) or $4,049 with the RF24-105 F4 L IS USM lens. That’s a bit more than the Nikon’s $2,500 Z6 III, which has a partially stacked but lower-resolution 24MP sensor. Panasonic’s S1 II also has a partially stacked 24MP but can shoot at up to 70 fps and costs $3,200. Finally, Sony’s A7 IV has a similar 33MP sensor but lacks the RAW video features of the R6 III. 

This article originally appeared on Engadget at https://www.engadget.com/cameras/canons-eos-r6-iii-arrives-with-7k-raw-video-and-40-fps-burst-speeds-060035923.html?src=rss

An anonymous reader quotes a report from MIT Technology Review: The US- and UK-based company Quantinuum today unveiled Helios, its third-generation quantum computer, which includes expanded computing power and error correction capability. Like all other existing quantum computers, Helios is not powerful enough to execute the industry’s dream money-making algorithms, such as those that would be useful for materials discovery or financial modeling. But Quantinuum’s machines, which use individual ions as qubits, could be easier to scale up than quantum computers that use superconducting circuits as qubits, such as Google’s and IBM’s. “Helios is an important proof point in our road map about how we’ll scale to larger physical systems,” says Jennifer Strabley, vice president at Quantinuum, which formed in 2021 from the merger of Honeywell Quantum Solutions and Cambridge Quantum. Honeywell remains Quantinuum’s majority owner.

Located at Quantinuum’s facility in Colorado, Helios comprises a myriad of components, including mirrors, lasers, and optical fiber. Its core is a thumbnail-size chip containing the barium ions that serve as the qubits, which perform the actual computing. Helios computes with 98 barium ions at a time; its predecessor, H2, used 56 ytterbium qubits. The barium ions are an upgrade, as they have proven easier to control than ytterbium. These components all sit within a chamber that is cooled to about 15 Kelvin (-432.67 ), on top of an optical table. Users can access the computer by logging in remotely over the cloud. […] Helios is noteworthy for its qubits’ precision, says Rajibul Islam, a physicist at the University of Waterloo in Canada, who is not affiliated with Quantinuum. The computer’s qubit error rates are low to begin with, which means it doesn’t need to devote as much of its hardware to error correction. Quantinuum had pairs of qubits interact in an operation known as entanglement and found that they behaved as expected 99.921% of the time. “To the best of my knowledge, no other platform is at this level,” says Islam.

[…] Besides increasing the number of qubits on its chip, another notable achievement for Quantinuum is that it demonstrated error correction “on the fly,” says David Hayes, the company’s director of computational theory and design, That’s a new capability for its machines. Nvidia GPUs were used to identify errors in the qubits in parallel. Hayes thinks that GPUs are more effective for error correction than chips known as FPGAs, also used in the industry. Quantinuum has used its computers to investigate the basic physics of magnetism and superconductivity. Earlier this year, it reported simulating a magnet on H2, Quantinuum’s predecessor, with the claim that it “rivals the best classical approaches in expanding our understanding of magnetism.” Along with announcing the introduction of Helios, the company has used the machine to simulate the behavior of electrons in a high-temperature superconductor. Quantinuum is expanding its Helios line with a new system in Minnesota. It’s also started developing its fourth-generation quantum computer, Sol, set for 2027 with 192 qubits. Then, a fifth-generation system, Apollo, is expected in 2029 with thousands of qubits and full fault tolerance.

Read more of this story at Slashdot.

A bungled October 18 heist that saw $102 million of crown jewels stolen from the Louvre in broad daylight has exposed years of lax security at the national art museum. From trivial passwords like ‘LOUVRE’ to decades-old, unsupported systems and easy rooftop access, the job was made surprisingly easy. PC Gamer reports: As Rogue cofounder and former Polygon arch-jester Cass Marshall notes on Bluesky, we owe a lot of videogame designers an apology. We’ve spent years dunking on the emptyheadedness of game characters leaving their crucial security codes and vault combinations in the open for anyone to read, all while the Louvre has been using the password “Louvre” for its video surveillance servers. That’s not an exaggeration. Confidential documents reviewed by Liberation detail a long history of Louvre security vulnerabilities, dating back to a 2014 cybersecurity audit performed by the French Cybersecurity Agency (ANSSI) at the museum’s request. ANSSI experts were able to infiltrate the Louvre’s security network to manipulate video surveillance and modify badge access.

“How did the experts manage to infiltrate the network? Primarily due to the weakness of certain passwords which the French National Cybersecurity Agency (ANSSI) politely describes as ‘trivial,'” writes Liberation’s Brice Le Borgne via machine translation. “Type ‘LOUVRE’ to access a server managing the museum’s video surveillance, or ‘THALES’ to access one of the software programs published by… Thales.” The museum sought another audit from France’s National Institute for Advanced Studies in Security and Justice in 2015. Concluded two years later, the audit’s 40 pages of recommendations described “serious shortcomings,” “poorly managed” visitor flow, rooftops that are easily accessible during construction work, and outdated and malfunctioning security systems. Later documents indicate that, in 2025, the Louvre was still using security software purchased in 2003 that is no longer supported by its developer, running on hardware using Windows Server 2003.

Read more of this story at Slashdot.

A new survey of over 300 US and UK gaming executives found that 72% view Steam as a monopoly. “Furthermore, 88% said that at least three-quarters of their revenue came from Steam, while 37% reported that the platform accounted for 90% of their total revenue,” adds Techspot. From the report: Atomik Research conducted the recent survey on behalf of Rokky, a company that helps game publishers minimize the impact of grey market key resellers on prices. In addition to opinions on Steam, developers also answered questions about the PC market’s biggest challenges.

The increasing popularity of free-to-play games such as Fortnite, DOTA 2, Counter-Strike 2, Call of Duty: Warzone, and Roblox topped the list of concerns for 40% of respondents. Approximately a third mentioned market saturation and discoverability, echoing data that suggests there aren’t enough players for the thousands of new titles released on Steam each year. A similar portion of survey respondents also expressed concerns regarding subscription services.

Read more of this story at Slashdot.

An anonymous reader quotes a report from the Associated Press: Google Maps is heading in a new direction with artificial intelligence sitting in the passenger’s seat. Fueled by Google’s Gemini AI technology, the world’s most popular navigation app will become a more conversational companion as part of a redesign announced Wednesday. The hands-free experience is meant to turn Google Maps into something more like an insightful passenger able to direct a driver to a destination while also providing nearby recommendations on places to eat, shop or sightsee, when asked for the advice. “No fumbling required — now you can just ask,” Google promised in a blog post about the app makeover.

The AI features are also supposed to enable Google Maps to be more precise by calling out landmarks to denote the place to make a turn instead of relying on distance notifications. AI chatbots, like Gemini and OpenAI’s ChatGPT, have sometimes lapsed into periods of making things up — known as “hallucinations” in tech speak — but Google is promising that built-in safeguards will prevent Maps from accidentally sending drivers down the wrong road. All the information that Gemini is drawing upon will be culled from the roughly 250 million places stored in Google Maps’ database of reviews accumulated during the past 20 years. Google Maps’ new AI capabilities will be rolling out to both Apple’s iPhone and Android mobile devices.

Read more of this story at Slashdot.

The University of Pennsylvania has a somewhat unusual distinction: It is the alma mater of two of the planet’s most polarizing figures, Elon Musk and Donald Trump. As the political power of both men rose over the last year, the US government began to pressure Penn, first by pulling its research funding and then by targeting the school for past actions related to a transgender swimmer.

After the “sticks” were deployed, a “carrot” was offered. Penn became one of just nine schools nationally to be offered a special “compact” with the federal government, which would give the feds broad control over the school and its speech in return for preferential access to federal funds. Penn declined to sign the deal. (Making the whole surreal situation stranger was the fact that one of Penn’s own wealthy boosters apparently helped the Trump administration write the compact.)

In other words, Penn has become an obvious target of the Trump administration; now it has been targeted by a hacker claiming to share Trump and Musk’s grievances over affirmative action and “wokeness.”

Read full article

Comments

A new bipartisan bill introduced by Senators Mark Warner and Josh Hawley would require companies and federal agencies to report quarterly on AI-related workforce changes, including layoffs, new hires, and retraining efforts. The data from the AI-Related Job Impacts Clarity Act (PDF) would then be compiled by the Department of Labor into a publicly available report.

“This bipartisan legislation will finally give us a clear picture of AI’s impact on the workforce,” Warner said in a statement. “Armed with this information, we can make sure AI drives opportunity instead of leaving workers behind.”

Read more of this story at Slashdot.

Anna’s Archive, a popular website for pirating books and articles, seems to be square in Google’s sights, according to copyright and digital rights publication TorrentFreak. The search giant is said to have blocked some 749 million Anna’s Archive URLs from showing up in search results, TorrentFreak found, after combing through a recent transparency report.

The removal wasn’t necessarily targeted, as Google regularly delists content at the request of copyright holders. At time of this writing, has taken down links to 15,125,359,564 pages since 2011. But this is the latest in an ongoing, AI-prompted saga that is seeing copyright holders crack down on so-called “shadow libraries,” and it already represents around 5% of Google’s overall takedowns.

Anna’s Archive is a platform for pirated e-books

Personally, I hadn’t heard of Anna’s Archive, which makes sense—it’s a newer player in the field. The platform popped up in 2022, shortly after its predecessor, Z-Library, had its domains seized by the U.S. Department of Justice. Since then, it’s been quietly operating on its own little corner of the internet, serving as an open-source search engine for literary works that links to free publicly available sources when they exist, and pirated uploads when they don’t. Like Z-Library, it’s been blocked by German ISPs and sued in the U.S., but remains operational.

You can think of it kind of like the Pirate Bay, but for literary works—but on a larger scale (impressive given how new it is). TorrentFreak notes that only 4.2 million Pirate Bay URLs have been taken off Google, which is paltry compared to Anna’s Archive’s numbers.

AI scraping could be a factor

That discrepancy could be due to more aggressive takedown filing from publishers and authors, as more than 1,000 separate users have issued takedown requests to date, according to the Google data. These include both individuals and larger names like Penguin Random House, and their diligence could be related to Anna’s Archives’ stance on AI, as the site has admitted that it has freely provided access to 30 LLM developers to train on its “illegal archive of books,” and still openly hosts freely accessible pages for others to access.

Where copyright holders and readers will go from here is still up in the air. It’s important to note that, despite all appearances to the contrary, Google does not own the internet. Removing a site from its search engine does not prevent users from visiting it directly, and all three Anna’s Archive domains—annas-archive.org, annas-archive.se, and annas-archive.li—remain live.

Additionally, Anna’s Archive does not host any pirated content itself, but simply provides users to links where they can find it. All of this puts it in a legal gray area, which, when backed by the site’s open-source nature and strong commitment to the ideal that “preserving and hosting these files is morally right,” means it’s likely to continue in some form or another for years.

Still, as companies like Meta are found to have used pirated content to train its AI models, it’s likely actions that Google’s will become more common, and other sites, or even legal entities, might follow suit. Plan accordingly. (And if, like me, you’ve been asking yourself “Who the heck is Anna?” the archive’s FAQ has an answer: “You are Anna.” It’s a nod at the anonymous uploaders who provide it with much of its material.)

Syphilis can be a tricky disease to diagnose—especially when a patient may not be sharing the whole story.

Doctors in Belgium met with a real head-scratcher when an 83-year-old married man came in with a rare form of secondary syphilis—the second of four stages of the sexually transmitted bacterial infection that has been called a “master of disguise.”

The man told doctors up front that he was in a monogamous 50-year-long marriage and had been sexually inactive in recent years following treatment for cancer. In a Clinical Problem-Solving report published today in the New England Journal of Medicine, doctors laid out the step-by-step tests and reasoning they used to get to the right diagnosis, which still didn’t answer all their questions.

Read full article

Comments

Google on Wednesday revealed five recent malware samples that were built using generative AI. The end results of each one were far below par with professional malware development, a finding that shows that vibe coding of malicious wares lags behind more traditional forms of development, which means it still has a long way to go before it poses a real-world threat.

One of the samples, for instance, tracked under the name PromptLock, was part of an academic study analyzing how effective the use of large language models can be “to autonomously plan, adapt, and execute the ransomware attack lifecycle.” The researchers, however, reported the malware had “clear limitations: it omits persistence, lateral movement, and advanced evasion tactics” and served as little more than a demonstration of the feasibility of AI for such purposes. Prior to the paper’s release, security firm ESET said it had discovered the sample and hailed it as “the first AI-powered ransomware.”

Don’t believe the hype

Like the other four samples Google analyzed—FruitShell, PromptFlux, PromptSteal, and QuietVault—PromptLock was easy to detect, even by less-sophisticated endpoint protections that rely on static signatures. All samples also employed previously seen methods in malware samples, making them easy to counteract. They also had no operational impact, meaning they didn’t require defenders to adopt new defenses.

Read full article

Comments

We may earn a commission from links on this page.

Meta has begun rolling out its latest firmware update for its Ray-Ban and Oakley smart glasses. The update, V19.2, is available this week, and improves Garmin integration and video recording for both first and second generation Ray-Ban Meta and Oakley Meta wearables.

Ray-Ban Meta – Wayfarer Large glasses with Meta AI, Audio, Photo, Video Compatibility – Polarized Gradient Graphite Lens (Refurbished) – Matte Black

$263.00
at Best Buy

$263.00
at Best Buy

The improvements to Garmin device integration lets Meta users sync Garmin stats to videos and photos, ask Meta AI for health info, and get real-time voice updates of fitness metrics while active. These features were previously exclusive to Oakley Meta Vanguard glasses, but now everyone has access.

As for the improvements to videos, the update introduces selectable stabilization modes (auto, low, medium, and high), the ability to record clips that are five minutes long (up from three minutes), and the option of recording 1080p video at 60 FPS—not bad for a camera you wear on your face.

Oakley | Meta HSTN – with Meta AI, Audio, Photo, Video Compatibility – Black, Prizm™ Black Polarized

$449.00
at Amazon

$449.00
at Amazon

How to update your Meta smart glasses

If you have Meta glasses set to “automatic update,” all you have to do is make sure your specs have power and are connected to your phone via Bluetooth and they’ll automatically update over the air. If you want to check for an update manually, follow these steps:

• Open the Meta AI app.

• Click the glasses icon at the top right.

• Select the Gear icon next to the glasses.

• Hit “Updates”

• Hit “Check” to see if your device is up-to-date.

Meta’s update schedule is impressive

Ray-Ban Metas have been my daily-wear glasses for over a year now and they’re my choice for best overall smart glasses on the market. One of the reasons I like them so much is that Meta provides regular software updates. The company makes fixes and adds features approximately every six to eight weeks, and they’re usually notable improvements.

August’s update added AI photo restyling and Google and Outlook calendar integration, and May’s update added Live AI and translation features. The pace of updates is on par with a platform instead of a gadget, making these smart glasses feel like they’re designed for the long haul, instead of a “buy one this year, then buy the new one next year” gadget.

A new Activate Consulting report reveals that 43% of Gen Z now prefer YouTube and TikTok over traditional TV or paid streaming. With global media revenues surging and traditional TV viewership collapsing, the average person now spends over 13 hours a day consuming content across platforms, effectively living a “32-hour day” through multitasking. Variety reports: Per the same survey, the popularity of “microdramas” — one of the latest trends on those platforms, consisting of 1-2 minute scripted episodes of an ongoing storyline — has been increasing at a fast rate with 28 million U.S. adults (52% aged 18-34) reportedly watching that new form of content.

Additional findings include projections for global internet and media revenue to increase by $388 billion by 2029, while average daily time spent streaming video will climb to 4 hours and 8 minutes as time spent watching traditional TV is set to collapse to just 1 hour and 17 minutes. Activate estimates that, as a result, streaming revenues (from ads and subscriptions) will grow 18-19% annually while traditional TV revenues will fall 4-6% year to year.

Read more of this story at Slashdot.

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

Ahead of Black Friday, Amazon has discounted several versions of the Google Pixel 10. Right now, you can get the Google Pixel 10 Pro XL with 256 GB of storage and 16 GB of RAM for $899 (originally $1,199), marking a record-low price according to price-tracking tools.

Google Pixel 10 Pro XL

$899.00
at Amazon

$1,199.00
Save $300.00

$899.00
at Amazon

$1,199.00
Save $300.00

Built with aluminum and Corning Gorilla Glass Victus 2, the phone has an IP68 rating, meaning it can withstand dust and be submerged to 1.5 meters for up to 30 minutes. Its 6.8″ Super Actua display reaches 3,300-nit peak brightness, making it easier to use in direct sunlight. Powered by the Google Tensor G5 chip, it integrates AI tools like the Magic Cue and Gemini Live, which help you brainstorm or point your camera at something you want more information about. 

The camera system offers 50 MP images and 8K video in low light. It has a 42 MP main lens, a  48 MP ultrawide lens, and a 48 MP telephoto with up to 100× digital zoom. The battery should last over 24 hours on a charge, and up to 100 hours with Extreme Battery Saver turned on. It takes between 77 and 90 minutes to fully charge and reaches around 70% in half an hour with a wired 45W charger. 

Google promises seven years of software support, which ensures lasting performance and security updates down the line. Altogether, the currently discounted Google Pixel 10 Pro XL is a great option for Android users who want to level up with a bigger, faster, and more powerful phone with impressive camera performance.

Hilarious co-op video game Overcooked is making the move to the real world. A24 has acquired the rights to the popular indie hit and Deadline reports that the company is working with Netflix to develop the concept into an unscripted competition reality TV series. I adore this idea, and with the right people behind it, I think it could be a masterwork of reality TV. According to the reports, this is the first time independent film and TV studio A24 will make a reality show, but Netflix has already had some success with food and cooking programs, such as with the absolute gem that is Nailed It! 

The often ridiculous antics of Overcooked are a wonderful match for that type of lighthearted competition show. In the game, one to four players work together to cook customers’ food orders as quickly as possible while the environment creates silly, unexpected obstacles. Unfortunately, there’s no way the Netflix legal team will green light challenges where contestants might fall into actual lava, and recreating the space travel levels would probably blow the budget. But I can see all sorts of Overcooked-style mayhem from sliding on icy floors or suddenly moving countertops. I’m also imagining a lot of dashing between inconveniently placed kitchen appliances and occasionally plunging the contestants into total darkness. There’s no timeline given in Deadline’s report, but I cannot wait for this to exist.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/co-op-game-overcooked-may-become-a-competition-reality-tv-show-on-netflix-222406920.html?src=rss

Snap and Perplexity AI have struck a $400 million deal that will bring the AI search engine directly to Snapchat sometime in “early 2026,” the two companies announced. With the partnership, Perplexity’s AI search engine will be a prominent part of Snapchat’s “chat” interface so users can “ask questions and get clear, conversational answers drawn from verifiable sources, all within Snapchat.”

The news was announced alongside the company’s third-quarter earnings. The company said that revenue from the deal — Perplexity is paying Snap $400 million for the integration — is “expected to begin contributing” to the company’s bottom line in 2026. In a letter to shareholders, CEO Evan Spiegel also hinted that Snap could pursue similar partnerships with other AI companies. “This collaboration makes AI-powered discovery native to Snapchat, enhances personalization, and positions Snap as a leading distribution channel for intelligent agents, laying the groundwork for a broader ecosystem of AI partners to reach our global community,” he wrote. 

Snap, like its peers, has been leaning into generative AI in recent years. The company has its own LLM-powered chatbot, called MyAI, which uses models from OpenAI and Google. Snap has also introduced AI-powered lenses and creation tools, which have helped boost its Snapchat+ subscription service.

This article originally appeared on Engadget at https://www.engadget.com/social-media/snap-and-perplexity-sign-400-million-deal-to-put-ai-search-directly-in-snapchat-221101734.html?src=rss

An anonymous reader quotes a report from the Guardian: Authorities in Denmark are urgently studying how to close an apparent security loophole in hundreds of Chinese-made electric buses that enables them to be remotely deactivated. The investigation comes after transport authorities in Norway, where the Yutong buses are also in service, found that the Chinese supplier had remote access for software updates and diagnostics to the vehicles’ control systems — which could be exploited to affect buses while in transit.

Amid concerns over potential security risks, the Norwegian public transport authority Ruter decided to test two electric buses in an isolated environment. Bernt Reitan Jenssen, Ruter’s chief executive, said: “The testing revealed risks that we are now taking measures against. National and local authorities have been informed and must assist with additional measures at a national level.” Their investigations found that remote deactivation could be prevented by removing the buses’ sim cards, but they decided against this because it would also disconnect the bus from other systems.

Ruter said it planned to bring in stricter security requirements for future procurements. Jenssen said it must act before the arrival of the next generation of buses, which could be even “more integrated and harder to secure.” Movia, Denmark’s largest public transport company, has 469 Chinese electric buses in operation — 262 of which were manufactured by Yutong. Jeppe Gaard, Movia’s chief operating officer, said he was made aware of the loophole last week. “This is not a Chinese bus problem,” he said. “It is a problem for all types of vehicles and devices with Chinese electronics built in.”

Read more of this story at Slashdot.

A familiar face will be helping Mayor-elect Zohran Mamdani set up his new administration before he takes office in 2026. Lina Khan, former Federal Trade Commission Chair under President Joe Biden, has been officially announced as one of Mamdani’s transition co-chairs, alongside Grace Bonilla, Maria Torres-Springer and Melanie Hartzog.

Mamdani’s platform is focused on affordability, with fighting corporate corruption a key way he hopes to lower prices for New Yorkers. Mamdani’s proposed policies include working to ban hidden fees and non-compete clauses, while funding challenges to utility company rate hikes. It’s not surprising that Khan and Mamdani would be aligned. As Chair, Khan is best known for trying to rebuild the FTC’s anti-monopolist backbone, but she was similarly interested in banning non-compete clauses and hidden junk fees. Khan has also publicly expressed her appreciation for the Mamdani campaign’s focus on small businesses in The New York Times Opinion section.

“I think what we saw last night was New Yorkers not just electing a new mayor, but clearly rejecting a politics where outsized corporate power and money too often end up dictating our politics,” Khan said at a press conference announcing her new role. “And a clear mandate for change, where New Yorkers can get ahead and where all workers and small businesses can thrive, not just get by.”

While Mamdani has served as a New York state assemblyman, his relative lack of experience has been used as a consistent criticism of his candidacy for mayor. Clearly, that didn’t matter to voters, but Mamdani’s chosen transition team members suggest he plans to surround himself with people who are experienced. In the case of Khan, that includes a transition co-chair who’s willing to be openly critical of corporate power. The Trump administration has effectively remade the FTC in its image, but there’s more than one place the influence of big businesses can be checked.

This article originally appeared on Engadget at https://www.engadget.com/general/former-ftc-chair-lina-khan-will-help-zohran-mamdani-build-his-new-administration-220304723.html?src=rss

This is Ze Frank’s ‘True Facts: The Science of the Praying Mantis’, an incredibly in-depth look at the lives of praying mantises, discussing everything you’ve ever wanted (and not wanted) to know about them. Not to brag about my own amateur entomology or anything, but most of this stuff I already knew because I used to be really into bugs. Truthfully, I still am, but I keep it on the down-low because I don’t want my girlfriend to find out I’ve been raising beetles in shoeboxes under the bed.