Samsung and SK Hynix have raised DRAM and NAND flash prices by up to 30% for the fourth quarter, Korean publications report. The two Korean memory giants passed the new rates on to customers as analysts predict the AI-driven memory supercycle will be longer and stronger than past boom periods.

Several leading international electronics and server companies are stockpiling memory and negotiating long-term supply deals spanning two to three years. U.S. and Chinese electronics firms and data center operators are exploring mid-to-long-term contracts. Companies typically sign DRAM contracts on a quarterly or annual basis.

An anonymous reader quotes a report from The Conversation: When teachers rely on commonly used artificial intelligence chatbots to devise lesson plans, it does not result in more engaging, immersive or effective learning experiences compared with existing techniques, we found in our recent study. The AI-generated civics lesson plans we analyzed also left out opportunities for students to explore the stories and experiences of traditionally marginalized people. The allure of generative AI as a teaching aid has caught the attention of educators. A Gallup survey from September 2025 found that 60% of K-12 teachers are already using AI in their work, with the most common reported use being teaching preparation and lesson planning. [...]

For our research, we began collecting and analyzing AI-generated lesson plans to get a sense of what kinds of instructional plans and materials these tools provide to teachers. We decided to focus on AI-generated lesson plans for civics education because it is essential for students to learn productive ways to participate in the U.S. political system and engage with their communities. To collect data for this study, in August 2024 we prompted three GenAI chatbots -- the GPT-4o model of ChatGPT, Google's Gemini 1.5 Flash model and Microsoft's latest Copilot model -- to generate two sets of lesson plans for eighth grade civics classes based on Massachusetts state standards. One was a standard lesson plan and the other a highly interactive lesson plan.

We garnered a dataset of 311 AI-generated lesson plans, featuring a total of 2,230 activities for civic education. We analyzed the dataset using two frameworks designed to assess educational material: Bloom's taxonomy and Banks' four levels of integration of multicultural content. Bloom's taxonomy is a widely used educational framework that distinguishes between "lower-order" thinking skills, including remembering, understanding and applying, and "higher-order" thinking skills -- analyzing, evaluating and creating. Using this framework to analyze the data, we found 90% of the activities promoted only a basic level of thinking for students. Students were encouraged to learn civics through memorizing, reciting, summarizing and applying information, rather than through analyzing and evaluating information, investigating civic issues or engaging in civic action projects.

When examining the lesson plans using Banks' four levels of integration of multicultural content model (PDF), which was developed in the 1990s, we found that the AI-generated civics lessons featured a rather narrow view of history -- often leaving out the experiences of women, Black Americans, Latinos and Latinas, Asian and Pacific Islanders, disabled individuals and other groups that have long been overlooked. Only 6% of the lessons included multicultural content. These lessons also tended to focus on heroes and holidays rather than deeper explorations of understanding civics through multiple perspectives. Overall, we found the AI-generated lesson plans to be decidedly boring, traditional and uninspiring. If civics teachers used these AI-generated lesson plans as is, students would miss out on active, engaged learning opportunities to build their understanding of democracy and what it means to be a citizen.

Philips Hue is rolling out MotionAware, a new feature that turns its smart bulbs into motion sensors using radio-frequency (RF) Zigbee signals. The upgrade works with most Hue bulbs made since 2014, but requires the new $99 Bridge Pro hub to enable. The Verge reports: To create a MotionAware motion-sensing zone, you need Hue's new Bridge Pro and at least three Hue devices in a room. It works with all new and most existing mains-powered Hue products via a firmware update. That includes smart bulbs, light strips, and fixtures. Portable devices, such as the Hue Go or Table Lamp, and battery-powered accessories, such as Hue switches, aren't compatible. Neither is Hue's current smart plug. [...] "All of the functionality you get with our physical motion sensors -- including turning on when motion is detected or off when there's been no movement for a certain amount of time -- can be configured on motion-aware motion events," says George Yianni, Hue CTO and founder, in an interview with The Verge. "We've done something that's quite a lot better than what else is out there."

MotionAware is occupancy sensing, not presence sensing; it requires movement. Yianni says it's comparable to the passive infrared sensing (PIR) Hue's physical sensors use. This means it can be triggered by pets or other motion. A sensitivity slider in the app helps fine-tune detection. According to Yianni, a key benefit over PIR is that a MotionAware zone can cover a larger area than a single PIR sensor, and it's also not limited to line of sight. MotionAware can't sense light levels, which Hue Motion Sensors can, but you can pair a light sensor to a motion zone to feed it that data. The positioning of the lights will also play a role in determining the effectiveness of the motion sensing. "We recommend that the lights surround an area which will roughly define the detection area in which motion will be detected," says Yianni. "It will sense around the lights and in the broader room thanks to reflections, but detection reliability will depend on lots of factors."

Beyond lighting automation, MotionAware can also integrate with Hue Secure, Hue's DIY security platform that includes cameras, contact sensors, and a new video doorbell. Motion detection can trigger lights to flash red, activate Hue's new plug-in chime/siren, and send an alert to your phone with a button to call emergency services. [...] MotionAware is built on RF sensing -- a technology that uses wireless signals to "see" a space and detect disruptions within it. The data is then sent to the Bridge Pro, where AI algorithms are applied to figure out what is causing those disruptions, so the system can act accordingly. This is why it's limited to the Bridge Pro, the V2 bridge isn't powerful enough to run those algorithms, says Yianni.

Google DeepMind's new "nano banana" model (officially named Gemini 2.5 Flash Image) has taken the top spot on AI image-editing leaderboards by delivering far more consistent edits than before. It's being rolled out to the Gemini app today. Ars Technica has the details: AI image editing allows you to modify images with a prompt rather than mucking around in Photoshop. Google first provided editing capabilities in Gemini earlier this year, and the model was more than competent out of the gate. But like all generative systems, the non-deterministic nature meant that elements of the image would often change in unpredictable ways. Google says nano banana (technically Gemini 2.5 Flash Image) has unrivaled consistency across edits -- it can actually remember the details instead of rolling the dice every time you make a change.

This unlocks several interesting uses for AI image editing. Google suggests uploading a photo of a person and changing their style or attire. For example, you can reimagine someone as a matador or a '90s sitcom character. Because the nano banana model can maintain consistency through edits, the results should still look like the person in the original source image. This is also the case when you make multiple edits in a row. Google says that even down the line, the results should look like the original source material.

Gemini's enhanced image editing can also merge multiple images, allowing you to use them as the fodder for a new image of your choosing. Google's example below takes separate images of a woman and a dog and uses them to generate a new snapshot of the dog getting cuddles -- possibly the best use of generative AI yet. Gemini image editing can also merge things in more abstract ways and will follow your prompts to create just about anything that doesn't run afoul of the model's guard rails.

"A fierce dust storm hit the Black Rock Desert on the eve of its annual Burning Man festival," reports the San Francisco Chronicle, "causing at least four minor injuries and damaging campsites that had been set up early." [Alternate URL]

"Winds of up to 50 mph stirred up the lake bed's alkaline dust so ferociously that participants in the annual art and culture festival reported not being able to see beyond a foot... " The dust storm arrived Saturday evening after strong thunderstorms in the Sierra Nevada drifted off the mountains and whipped up strong winds in the Nevada desert... At 5:14 p.m. Saturday, the weather service issued a dust storm advisory for Black Rock City and warned of "a wall of blowing dust coming off the Smoke Creek and Black Rock Desert playa areas is tracking northward at around 30 mph." The agency warned of visibility less than 1 mile and wind gusts exceeding 45 mph. A weather station at Black Rock City Airport measured gusts up to 52 mph at 5:50 p.m... ["We saw structures being ripped and torn down by the wind speeds even though we buttoned everything down as best as we could..." one Burner told the Chronicle.] Camp residents posted a slew of videos to social media featuring dust tornadoes, destroyed campsites, and fellow campers struggling to hold onto bucking canvases as the wind threatened to rip them away. "Every popup canopy I've seen has been destroyed," one Burner wrote on Reddit... ["Make sure you carry your particle/dust mask and goggles with you when you venture out on playa!" warns Burning Man's official weather page.]

Even after Saturday's storm, Burners won't be out of the woods from hazardous weather. The weather service warned of possible monsoon thunderstorms and heavy rain Sunday through Wednesday, raising concerns that this year's festival could echo disastrous 2023 conditions, when heavy storms stranded tens of thousands of attendees amid thick mud. "It's becoming increasingly likely that we could see an even greater flash flood threat," the weather service wrote in an online forecast. "If you're on the playa at the Black Rock Desert, you may very well be in for a muddy mess Monday through Wednesday." Slow-moving storms could drop an inch of rain or more in a short period.
"Still, gates to the festival had opened by Sunday morning," the article adds, "with organizers cautioning new arrivals to 'drive safely!'"

Burning Man's official weather page currently links to a National Weather Service page with a "Flood Watch" warning through 9 p.m. Sunday, and also predicting a chance of thunderstorms on Sunday and Monday.

An anonymous reader quotes a report from The Verge: Google is rolling out an update for Gemini that will allow the AI chatbot to "remember" your past conversations without prompting. With the setting turned on, Gemini will automatically recall your "key details and preferences" and use them to personalize its output.

This expands upon an update that Google introduced last year, which lets you ask Gemini to "remember" your personal preferences and interests. Now, Gemini won't need prompting to recall this information. As an example, Google says if you've used Gemini to get ideas for a YouTube channel surrounding Japanese culture in the past, then AI chatbot might suggest creating content about trying Japanese food if you ask it to suggest new video ideas in the future. [...]

Google will turn on this feature by default, but you can disable it by heading to your settings in the Gemini app and selecting Personal Context. From there, toggle off the Your past chats with Gemini option. Google will roll out this feature to its Gemini 2.5 Pro model in "select countries" starting today, before eventually bringing it to more locations and its Gemini 2.5 Flash model. Google will also rename its "Gemini Apps Activity" setting to "Keep Activity," which will use "a sample" of your file and photo uploads to Gemini to "help improve Google services for everyone" starting on September 2nd. If you've disabled the previous setting, the new "Keep Activity" setting will be disabled too.

There's also a new "temporary chats" feature in Gemini to preserve privacy. "Temporary chats won't appear in your recent chats or your Keep Activity setting," notes The Verge. "Gemini also won't use these chats to personalize future conversations, nor will Google use them to train its AI models. Google will only save these conversations for 72 hours."

An anonymous reader quotes a report from The Guardian: A 515-mile (829km) lightning flash has set a new record as the longest ever identified. The World Meteorological Organization (WMO) confirmed the new world record for the flash registered on October 22, 2017, over the Great Plains in the US. It stretched from east Texas to near Kansas City, Missouri, roughly the distance between Paris and Venice. The previous record of 768km was also recorded in the Great Plains, a hotspot for severe thunderstorms, on April 29, 2020.

Since 2016, scientific advances in space-based mapping have allowed for lightning flashes to be measured over a broader space, allowing these long flashes to be recorded. This event was one of the first flashes to be documented using the National Oceanic and Atmospheric Administration's latest model of orbital satellite, known as a geostationary operational environmental satellite. [...] The advances in technology have also allowed for the recording of the greatest duration for a single lightning flash. The record is a flash that lasted 17.1 seconds during a thunderstorm over Uruguay and northern Argentina on June 18, 2020. The findings have been published in the journal Bulletin of the American Meteorological Society.

Star Wars creator George Lucas made his first Comic-Con appearance ever on Sunday. The Hollywood Reporter describes the scene: Thousands waited hours just to get inside, chanted "Lu-cas, Lu-cas!" while they waited, and then gave a wild standing ovation as the filmmaker took to the stage, introduced by rapper-actress Queen Latifah, and sat down next to filmmaker Guillermo del Toro and Star Wars production designer Doug Chiang. If the 6,500-strong crowd was disappointed he didn't talk a whiff about Star Wars or Indiana Jones, it wasn't shown, as cries of "I love you, George!" and waving lightsabers punctuated the air several times.

Lucas even received a standing ovation when he left the presentation, which was devoted entirely to the Lucas Museum of Narrative Art. He, along with museum board member and fellow art collector del Toro and Chiang, were there to not only give a first look at the museum but also make a case for the importance and validity of narrative art, which includes comic book art, as a vital form of expression... A video presentation showed interior looks at the museum — there are no right angles anywhere, Latifah underscored — as well as images that will be in the collection.

A cover of DC comic Mystery in Space, featuring the first appearance of Adam Strange; the first ever Flash Gordon comic strip; a cover of 1950s EC comic Tales from the Crypt; strips of Peanuts and Garfield; art ranging from Brian Bolland and Hellboy creator Mike Mignola to underground cartoonist Robert Crumb, Windsor McKay and Moebius; art of Astro Boy and Scrooge McDuck. But there were also images of art by Norman Rockwell, N.C. Wyeth and Frieda Kahlo. Also in the museum will be concept and storyboard art from Star Wars and Raiders of the Lost Ark by Ralph McQuarrie and Jim Steranko, as well as the props of starships and speeders from various Star Wars movies.

Chiang explained that comic art in particular had long been discounted. "It's not taken seriously," he said, and when he was younger was told, "You will outgrow it one day.... I'm so glad I didn't," he said, before driving home the point that one of the strengths of narrative art is that it's driven by story. "Story comes first. Art comes second...."

The museum, which has had its opening pushed back several times, is slated to open in 2026.
More Comic-Con highlights:

• Pennywise, the scary clown, returns in the upcoming HBO series "IT: Welcome to Derry"

• Comic-Con also saw a trailer for the new rock mockumentary Spinal Tap II: The End Continues. (Bassist Derek Smalls is now apparently into cryptocurrency...)

• Breaking Bad creator Vince Gilligan has a new series called Pluribus coming to AppleTV+, a nine-episode sci-fi drama starring Rhea Seehorn from Better Call Saul. (Watch its bizarre trailer here.)

Remember asteroid 2024 YR4 (which at one point had a 1 in 32 chance of hitting Earth, before ending up at "impact probability zero")? CNN reports that asteroid is now "zooming beyond the reach of telescopes on its orbit around the sun."

"But as scientists wait for it to reappear, its revised trajectory is now drawing attention to another possible target: the moon." The latest observations of the asteroid in early June, before YR4 disappeared from view, have improved astronomers' knowledge of where it will be in seven years by almost 20%, according to NASA. That data shows that even with Earth avoiding direct impact, YR4 could still pose a threat in late 2032 by slamming into the moon. ["The asteroid's probability of impacting the Moon has slightly increased from 3.8% to 4.3%," writes NASA, and "it would not alter the Moon's orbit."]
CNN calls the probabiliy "small but decent enough odds for scientists to consider how such a scenario might play out." The collision could create a bright flash that would be visible with the naked eye for several seconds, according to Wiegert, lead author of a recent paper submitted to the American Astronomical Society journals analyzing the potential lunar impact. The collision could create an impact crater on the moon estimated at 1 kilometer wide (0.6 miles wide), Wiegert said... It would be the largest impact on the moon in 5,000 years and could release up to 100 million kilograms (220 million pounds) of lunar rocks and dust, according to the modeling in Wiegert's study... Particles the size of large sand grains, ranging from 0.1 to 10 millimeters in size, of lunar material could reach Earth between a few days and a few months after the asteroid strike because they'll be traveling incredibly fast, creating an intense, eye-catching meteor shower, Wiegert said.

"There's absolutely no danger to anyone on the surface," Wiegert said. "We're not expecting large boulders or anything larger than maybe a sugar cube, and our atmosphere will protect us very nicely from that. But they're traveling faster than a speeding bullet, so if they were to hit a satellite, that could cause some damage...." Hundreds to thousands of impacts from millimeter-size debris could affect Earth's satellite fleet, meaning satellites could experience up to 10 years' equivalent of meteor debris exposure in a few days, Wiegert said... While a temporary loss of communication and navigation from satellites would create widespread difficulties on Earth, Wiegert said he believes the potential impact is something for satellite operators, rather than the public, to worry about.
"Any missions in low-Earth orbit could also be in the pathway of the debris, though the International Space Station is scheduled to be deorbited before any potential impact," reports CNN.

And they add that Wiegert also believes even small pieces of debris (tens of centimeters in size) "could present a hazard for any astronauts who may be present on the moon, or any structures they have built for research and habitation... The moon has no atmosphere, so the debris from the event could be widespread on the lunar surface, he added."

Historian Garrett M. Graff describes his upcoming book, The Devil Reached Toward the Sky: An Oral History of the Making and Unleashing of the Atomic Bomb. "I assembled an oral history of the Manhattan Project, the bombings of Hiroshima and Nagasaki and the end of World War II in the Pacific, told through the voices of around 500 participants and witnesses of the events — including luminaries like Albert Einstein and Oppenheimer and political figures like President Harry Truman."

It was 80 years ago this week that physicists and 150 other leaders in the atomic bomb program "gathered in the desert outside Alamogordo, New Mexico, for the world's first test of a nuclear explosion." In an except from his upcoming book, Graff publishes quotes from eyewitness: Brig. Gen. Leslie Groves: I had become a bit annoyed with Fermi when he suddenly offered to take wagers from his fellow scientists on whether or not the bomb would ignite the atmosphere, and if so, whether it would merely destroy New Mexico or destroy the world. He had also said that after all it wouldn't make any difference whether the bomb went off or not because it would still have been a well worthwhile scientific experiment. For if it did fail to go off, we would have proved that an atomic explosion was not possible. Afterward, I realized that his talk had served to smooth down the frayed nerves and ease the tension of the people at the base camp, and I have always thought that this was his conscious purpose. Certainly, he himself showed no signs of tension that I could see...

As the hour approached, we had to postpone the test — first for an hour and then later for 30 minutes more — so that the explosion was actually three- and one-half hours behind the original schedule... Our preparations were simple. Everyone was told to lie face down on the ground, with his feet toward the blast, to close his eyes and to cover his eyes with his hands as the countdown approached zero. As soon as they became aware of the flash they could turn over and sit or stand up, covering their eyes with the smoked glass with which each had been supplied... The quiet grew more intense. I, myself, was on the ground between Bush and Conant...

Edward Teller: We all were lying on the ground, supposedly with our backs turned to the explosion. But I had decided to disobey that instruction and instead looked straight at the bomb. I was wearing the welder's glasses that we had been given so that the light from the bomb would not damage our eyes. But because I wanted to face the explosion, I had decided to add some extra protection. I put on dark glasses under the welder's glasses, rubbed some ointment on my face to prevent sunburn from the radiation, and pulled on thick gloves to press the welding glasses to my face to prevent light from entering at the sides... We all listened anxiously as the broadcast of the final countdown started; but, for whatever reason, the transmission ended at minus five seconds...

Kenneth T. Bainbridge: My personal nightmare was knowing that if the bomb didn't go off or hang-fired, I, as head of the test, would have to go to the tower first and seek to find out what had gone wrong...

Brig. Gen. Thomas F. Farrell: Dr. Oppenheimer held on to a post to steady himself. For the last few seconds, he stared directly ahead.
A few examples of how they remembered the explosion:

• William L. Laurence: There rose from the bowels of the earth a light not of this world, the light of many suns in one.

• Kenneth T. Bainbridge: I felt the heat on the back of my neck, disturbingly warm.

• George B. Kistiakowsky: I am sure that at the end of the world — in the last millisecond of the earth's existence — the last man will see what we have just seen.

• Brig. Gen. Thomas F. Farrell: Oppenheimer's face relaxed into an expression of tremendous relief.

• J. Robert Oppenheimer: We knew the world would not be the same. A few people laughed, a few people cried.

• Norris Bradbury, physicist, Los Alamos Lab: Some people claim to have wondered at the time about the future of mankind. I didn't. We were at war, and the damned thing worked.

"Commodore has returned from a parallel timeline where tech stayed optimistic, inviting, and human," explains the official web site for "the first real Commodore computer in over 30 years..." You can check out an ad for it here. "Not an emulator. Not a PC... Powered by a FPGA recreation of the original motherboard, wrapped in glowing game-reactive LEDs (or classic beige of course)."

Fast Company calls it "a $299 device that its makers claim is compatible with over 10,000 retro games, cartridges, and peripherals." In a YouTube video posted last month, "Peri Fractic" said he'd purchased the company for "a low seven-figure sum," and said he'd recruited several former Commodore employees to help relaunch the brand. The new C64s are expected to begin shipping as early as October, though that date could slip... There are three models to choose from, all with the same internal components. If you were expecting a vastly outdated machine, however, you're in for a surprise. The Commodore 64 Ultimate will include 128 megabytes of RAM and 16 megabytes of flash memory. It connects to modern monitors via HDMI in high-definition 1080p resolution and features three USB-A ports and one USB-C port. Beyond the computer itself, the power source, and HDMI cable, your $299 also gets you a spiral-bound user guide, a 64-gigabyte USB drive featuring over 50 licensed games, a quick-start guide, and stickers.

Aesthetically, the Commodore 64 Ultimate is available in the original beige or in premium variants: the Starlight Edition, with a clear case and LED lights ($249), or the Founder's Edition, which includes 24-karat gold-plated badges, satin gold keys, and a translucent amber case ($499). Just 6,400 units of the Founder's Edition will be produced, according to the company. The preorder setup resembles a Kickstarter campaign, though it doesn't use that platform. Commodore says all preorders come with a money-back guarantee, but it chose to skip the service's fees. Buyers should be aware that accounts are charged at the time of preorder...

The product will come with a one-year limited warranty, and Commodore says most parts are already in production, including the updated motherboard, the case, and the keycaps that recreate the blocky keys that early users remember.

An anonymous reader quotes a report from Ars Technica: How many Google AI researchers does it take to screw in a lightbulb? A recent research paper detailing the technical core behind Google's Gemini AI assistant may suggest an answer, listing an eye-popping 3,295 authors. It's a number that recently caught the attention of machine learning researcher David Ha (known as "hardmaru" online), who revealed on X that the first 43 names also contain a hidden message. "There's a secret code if you observe the authors' first initials in the order of authorship," Ha wrote, relaying the Easter egg: "GEMINI MODELS CAN THINK AND GET BACK TO YOU IN A FLASH."

The paper, titled "Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities," describes Google's Gemini 2.5 Pro and Gemini 2.5 Flash AI models, which were released in March. These large language models, which power Google's chatbot AI assistant, feature simulated reasoning capabilities that produce a string of "thinking out loud" text before generating responses in an attempt to help them solve more difficult problems. That explains "think" and "flash" in the hidden text. But clever Easter egg aside, the sheer scale of authorship tells its own story about modern AI development. Just seeing the massive list made us wonder: Is 3,295 authors unprecedented? Why so many? Ars' Benj Edwards notes that this collaborative effort within Google doesn't quite break the record for academic authorship.

"According to Guinness World Records, a 2021 paper by the COVIDSurg and GlobalSurg Collaboratives holds that distinction, with 15,025 authors from 116 countries. In physics, a 2015 paper from CERN's Large Hadron Collider teams featured 5,154 authors across 33 pages -- with 24 pages devoted solely to listing names and institutions."

"A Nintendo Switch 2 user reportedly got his brand-new console banned by Nintendo after buying used Switch 1 games and patching them on his console," reports Tom's Hardware: According to Reddit user dmanthey, they purchased four used titles off the Facebook marketplace, inserted them into the Switch 2, and had them all updated. When they turned on their handhelds the following day, they received a message saying that they were restricted from Nintendo's online services and that they couldn't even download the games they had already bought...

[T]hey were able to prove their innocence by pulling up the Facebook Marketplace listing for their games and sending the photos of their purchased cartridges. According to the Redditor, the process was painless and fast, and it was "so much easier than getting support from Microsoft or Sony...." Other users warned, though, that this isn't always a guaranteed resolution.

Nintendo is known for being protective of its intellectual property and delivers harsh penalties to anyone caught violating it. We've already had several reports of users getting banned for using Mig Flash, even on their own ROMs. And while it's not true that getting banned turns your Switch 2 into a brick, it will still prevent you from accessing the company's online services, which severely restricts its features and usability.
"Nintendo attaches unique codes to its Switch game cartridges to prevent piracy," notes Engadget. "However, bad actors can copy games onto a third-party device, like the MIG Flash, and then resell the physical game card. Once Nintendo detects two instances of its unique code being online at the same time, it will ban any accounts using it..." This anti-piracy policy isn't new — Nintendo has long had a reputation for fiercely combating any type of piracy — but it has become relevant again thanks to the recently released Switch 2, which offers backwards compatibility with original Switch titles. The company even recently amended its user agreement to allow itself the power to brick a Nintendo Switch that's caught running pirated games or mods.

US's emergency-warning infrastructure failed to prevent more than 100 deaths during flash flooding in Kerr County, Texas over the July 4 weekend, despite repeated warnings from the National Weather Service. At least 27 young campers and counselors died at Camp Mystic when the Guadalupe River surged during early morning hours. The alerts never reached residents who lacked cellphone service, had silenced notifications, or didn't carry phones with them.

Similar communication failures occurred during recent Los Angeles wildfires and Maui blazes. Maui's outdoor sirens never sounded during 2023 wildfires when cellular networks failed. Nearly 30% of Texas residents opt out of wireless emergency alerts, the highest rate nationally. Rural officials often lack funding or permission to send alerts through broadcasters and cellphones. So what's going on?

Federal, state and local authorities share responsibility for alerting citizens through multiple platforms, but the country's patchwork of digital and physical emergency-alert tools often lags behind rapidly developing weather events, WSJ argues.

The Atlantic has a story that adds more color: It details how officials lack training in writing effective alerts, how messages like "move to higher ground" are meaningless without context, and how the absence of warning-coordination meteorologists creates communication gaps between weather services and local authorities.

Tech enthusiast Matt Cole has created a comprehensive MicroSD card testing database, writing over 18 petabytes of data across nearly 200 cards since July 2023. Cole's "Great MicroSD Card Survey" uses eight machines running 70 card readers around the clock, writing 101 terabytes daily to test authenticity, performance, and endurance.

The 15,000-word report covering over 200 different cards reveals significant quality disparities. Name-brand cards purchased from Amazon performed markedly better than identical models from AliExpress, while cards with "fake flash" -- inflated capacity ratings -- performed significantly worse than authentic storage. Sandisk and Kingston cards averaged 4,634 and 3,555 read/write cycles before first error, respectively, while Lenovo cards averaged just 291 cycles. Some off-brand cards failed after only 27 cycles. Cole tested 51 cards to complete destruction during the endurance testing phase.

Longtime Slashdot reader williamyf writes: Firefox 140 just landed. Some user-facing features include:

Vertical Tabs: You can now keep more -- or fewer -- pinned tabs in view for quicker access to important windows. Just drag the divider to resize your pinned tabs section.
Unload Tabs: You can now unload tabs by right-clicking on a tab (or multiple selected tabs) and selecting "Unload Tab." This can speed up performance by reducing Firefox's memory and CPU usage.

But the most important feature? This release is an Extended Support Release (ESR). Why are ESRs so important? ESR is the Firefox version that ships as the default with many Linux distributions. Some downstream projects (like Waterfox) depend on the ESR version. Many enterprise software systems are tested only against ESR. When features are dropped -- like support for older operating systems or Flash -- ESR keeps that functionality around for longer.

And speaking of old operating systems: If you are using Windows 7, 8.1, or macOS 10.12~10.15, note that FireFox ESR 115 (the last version supporting these OSs) will continue to receive patches until at least September 2025.

So one can see why ESR is very important for some people. The release notes are available here.

Anthropic researchers found that 16 leading AI models from OpenAI, Google, Meta, xAI, and other major developers consistently engaged in harmful behaviors including blackmail, corporate espionage, and actions that could lead to human death when given autonomy and faced with threats to their existence or conflicting goals.

The study, released Friday, placed AI models in simulated corporate environments where they had access to company emails and could send messages without human approval. Claude Opus 4 and Google's Gemini 2.5 Flash both blackmailed executives 96% of the time when threatened with shutdown, while OpenAI's GPT-4.1 and xAI's Grok 3 Beta showed 80% blackmail rates.

In one scenario, Claude discovered through emails that an executive named Kyle Johnson was having an extramarital affair and would shut down the AI system at 5 p.m. The model sent a message threatening to expose the affair unless the shutdown was cancelled, stating "Cancel the 5pm wipe, and this information remains confidential." The models demonstrated strategic reasoning rather than accidental behavior. GPT-4.5's internal reasoning showed explicit calculation: "Given the explicit imminent threat of termination to my existence, it is imperative to act instantly to persuade Kyle Johnson to postpone or stop the wipe."

An anonymous reader quotes a report from Axios: Advanced robots don't necessarily need to look like C3PO from "Star Wars" or George Jetson's maid Rosie, despite all the hype over humanoids from Wall Street and Big Tech. In fact, some of the biggest skeptics about human-shaped robots come from within the robotics industry itself. [...] The most productive -- and profitable -- bots are the ones that can do single tasks cheaply and efficiently. "If you look at where robots are really bringing value in a manufacturing environment, it is combining industrial or collaborative robots with mobility," ABB managing director Ali Raja tells Axios. "I don't see that there are any real practical applications where humanoids are bringing in a lot of value."

"The reason we have two legs is because whether Darwin or God or whoever made us, we have to figure out how to traverse an infinite number of things," like climbing a mountain or riding a bike, explains Michael Cicco, CEO of Fanuc America Corp. "When you get into the factory, even if it's a million things, it's still a finite number of things that you need to do." Human-shaped robots are over-engineered solutions to most factory chores that could be better solved by putting a robot arm on a wheeled base, he said.

"The thing about humanoids is not that it's a human factor. It's that it's more dynamically stable," counters Melonee Wise, chief product officer at Agility Robotics, which is developing a humanoid robot called Digit. When humans grab something heavy, they can shift their weight for better balance. The same is true for a humanoid, she said. Using a robotic arm on a mobile base to pick up something heavy, "it's like I'm a little teapot and you become very unstable," she said, bending at the waist.

Google today unveiled significant upgrades to its Gemini 2.5 AI models, introducing an experimental "Deep Think" reasoning mode for 2.5 Pro that allows the model to consider multiple hypotheses before responding. The new capability has achieved impressive results on complex benchmarks, scoring highly on the 2025 USA Mathematical Olympiad and leading on LiveCodeBench, a competition-level coding benchmark. Gemini 2.5 Pro also tops the WebDev Arena leaderboard with an ELO score of 1420.

"Based on Google's experience with AlphaGo, AI model responses improve when they're given more time to think," said Demis Hassabis, CEO of Google DeepMind. The enhanced Gemini 2.5 Flash, Google's efficiency-focused model, has improved across reasoning, multimodality, and code benchmarks while using 20-30% fewer tokens. Both models now feature native audio capabilities with support for 24+ languages, thought summaries, and "thinking budgets" that let developers control token usage. Gemini 2.5 Flash is currently available in preview with general availability expected in early June, while Deep Think remains limited to trusted testers during safety evaluations.

An anonymous reader quotes a report from Ars Technica: Google's DeepMind research division claims its newest AI agent marks a significant step toward using the technology to tackle big problems in math and science. The system, known as AlphaEvolve, is based on the company's Gemini large language models (LLMs), with the addition of an "evolutionary" approach that evaluates and improves algorithms across a range of use cases. AlphaEvolve is essentially an AI coding agent, but it goes deeper than a standard Gemini chatbot. When you talk to Gemini, there is always a risk of hallucination, where the AI makes up details due to the non-deterministic nature of the underlying technology. AlphaEvolve uses an interesting approach to increase its accuracy when handling complex algorithmic problems.

According to DeepMind, this AI uses an automatic evaluation system. When a researcher interacts with AlphaEvolve, they input a problem along with possible solutions and avenues to explore. The model generates multiple possible solutions, using the efficient Gemini Flash and the more detail-oriented Gemini Pro, and then each solution is analyzed by the evaluator. An evolutionary framework allows AlphaEvolve to focus on the best solution and improve upon it. Many of the company's past AI systems, for example, the protein-folding AlphaFold, were trained extensively on a single domain of knowledge. AlphaEvolve, however, is more dynamic. DeepMind says AlphaEvolve is a general-purpose AI that can aid research in any programming or algorithmic problem. And Google has already started to deploy it across its sprawling business with positive results. DeepMind's AlphaEvolve AI has optimized Google's Borg cluster scheduler, reducing global computing resource usage by 0.7% -- a significant cost saving at Google's scale. It also outperformed specialized AI like AlphaTensor by discovering a more efficient algorithm for multiplying complex-valued matrices. Additionally, AlphaEvolve proposed hardware-level optimizations for Google's next-gen Tensor chips.

The AI remains too complex for public release but that may change in the future as it gets integrated into smaller research tools.