ArsTechnica
DirecTV screensavers will show AI-generated ads with your face in 2026
As if DirecTV doesn't have enough trouble keeping customers, the satellite TV provider's streaming devices will show AI-generated screensaver ads next year, according to an announcement today from partnering ads company Glance.
People who use either of DirecTV’s two Gemini streaming devices will start seeing the ads “in early 2026,” per the announcement. DirecTV’s Gemini Air is an Android TV-powered USB device that people can plug into a TV for access to live TV channels, as well as streaming apps. Gemini Air doesn’t require a DirecTV satellite connection, and DirecTV gives all of its Internet customers the device. DirecTV first started selling Gemini devices in 2023, when it launched a separate Gemini set-top box that connects through DirecTV satellite setups.
DirecTV made an agreement with Glance to show AI-generated content and ads on Gemini devices' screensavers. Currently, Gemini devices show Google wallpapers as screensavers, which are on by default. When the new screensavers launch, Glance's AI content will show if the TV is idle for 10 minutes, The Verge reported.
Google will let Gemini schedule meetings for you in Gmail
Meetings can be a real drain on productivity, but a new Gmail feature might at least cut down on the time you spend scheduling them. The company has announced "Help Me Schedule" is coming to Gmail, leveraging Gemini AI to recognize when you want to schedule a meeting and offering possible meeting times for the email recipient to choose.
The new meeting feature is reminiscent of Magic Cue on Google's latest Pixel phones. As you type emails, Gmail will be able to recognize when you are planning a meeting. A Help Me Schedule button will appear in the toolbar. Upon clicking, Google's AI will swing into action and find possible meeting times that match the context of your message and are available in your calendar.
When you engage with Help me schedule, the AI generates an in-line meeting widget for your message. The recipient can select the time that works for them, and that's it—the meeting is scheduled for both parties. What about meetings with more than one invitee? Google says the feature won't support groups at launch.
OpenAI unveils “wellness” council; suicide prevention expert not included
Ever since a lawsuit accused ChatGPT of becoming a teen's "suicide coach," OpenAI has been scrambling to make its chatbot safer. Today, the AI firm unveiled the experts it hired to help make ChatGPT a healthier option for all users.
In a press release, OpenAI explained its Expert Council on Wellness and AI started taking form after OpenAI began informally consulting with experts on parental controls earlier this year. Now it's been formalized, bringing together eight "leading researchers and experts with decades of experience studying how technology affects our emotions, motivation, and mental health" to help steer ChatGPT updates.
One priority was finding "several council members with backgrounds in understanding how to build technology that supports healthy youth development," OpenAI said, "because teens use ChatGPT differently than adults."
Nvidia sells tiny new computer that puts big AI on your desktop
On Tuesday, Nvidia announced it will begin taking orders for the DGX Spark, a $4,000 desktop AI computer that wraps one petaflop of computing performance and 128GB of unified memory into a form factor small enough to sit on a desk. Its biggest selling point is likely its large integrated memory that can run larger AI models than consumer GPUs.
Nvidia will begin taking orders for the DGX Spark on Wednesday, October 15, through its website, with systems also available from manufacturing partners and select US retail stores.
The DGX Spark, which Nvidia previewed as "Project DIGITS" in January and formally named in May, represents Nvidia's attempt to create a new category of desktop computer workstation specifically for AI development.
GM’s EV push will cost it $1.6 billion in Q3 with end of the tax credit
The prospects of continued electric vehicle adoption in the US are in an odd place. As promised, the Trump administration and its congressional Republican allies killed off as many of the clean energy and EV incentives as they could after taking power in January. Ironically, though, the end of the clean vehicle tax credit on September 30 actually spurred the sales of EVs, as customers rushed to dealerships to take advantage of the soon-to-disappear $7,500 credit.
Predictions for EV sales going forward aren't so rosy, and automakers are reacting by adjusting their product portfolio plans. Today, General Motors revealed that will result in a $1.6 billion hit to its balance sheet when it reports its Q3 results late this month, according to its 8-K.
Q3 was a decent one for GM, with sales up 8 percent year on year and up 10 percent for the year to date. GM EV sales look even better: up 104 percent for the year to date compared to the first nine months of 2024, with nearly 145,000 electric Cadillacs, Chevrolets, and GMCs finding homes.
Windows 10 support “ends” today, but it’s just the first of many deaths
Today is the official end-of-support date for Microsoft's Windows 10. That doesn't mean these PCs will suddenly stop working, but if you don't take action, it does mean your PC has received its last regular security patches and that Microsoft is washing its hands of technical support.
This end-of-support date comes about a decade after the initial release of Windows 10, which is typical for most Windows versions. But it comes just four years after Windows 10 was replaced by Windows 11, a version with stricter system requirements that left many older-but-still-functional PCs with no officially supported upgrade path. As a result, Windows 10 still runs on roughly 40 percent of the world's Windows PCs (or around a third of US-based PCs), according to StatCounter data.
But this end-of-support date also isn't set in stone. Home users with Windows 10 PCs can enroll in Microsoft's Extended Security Updates (ESU) program, which extends the support timeline by another year. We've published directions for how to do this here—while you do need one of the Microsoft accounts that the company is always pushing, it's relatively trivial to enroll in the ESU program for free.
OpenAI wants to stop ChatGPT from validating users’ political views
"ChatGPT shouldn't have political bias in any direction."
That's OpenAI's stated goal in a new research paper released Thursday about measuring and reducing political bias in its AI models. The company says that "people use ChatGPT as a tool to learn and explore ideas" and argues "that only works if they trust ChatGPT to be objective."
But a closer reading of OpenAI's paper reveals something different from what the company's framing of objectivity suggests. The company never actually defines what it means by "bias." And its evaluation axes show that it's focused on stopping ChatGPT from several behaviors: acting like it has personal political opinions, amplifying users' emotional political language, and providing one-sided coverage of contested topics.
SpaceX finally got exactly what it needed from Starship V2
SpaceX closed a troubled but instructive chapter in its Starship rocket program Monday with a near-perfect test flight that carried the stainless steel spacecraft halfway around the world from South Texas to the Indian Ocean.
The rocket's 33 methane-fueled Raptor engines roared to life at 6:23 pm CDT (7:23 pm EDT; 23:23 UTC), throttling up to generate some 16.7 million pounds of thrust, by large measure more powerful than any rocket before Starship. Moments later, the 404-foot-tall (123.1-meter) rocket began a vertical climb away from SpaceX's test site in Starbase, Texas, near the US-Mexico border.
From then on, the rocket executed its flight plan like clockwork. This was arguably SpaceX's most successful Starship test flight to date. The only flight with a similar claim occurred one year ago Monday, when the company caught the rocket's Super Heavy booster back at the launch pad after soaring to the uppermost fringes of the atmosphere. But that flight didn't accomplish as much in space.
Hackers can steal 2FA codes and private messages from Android phones
Android devices are vulnerable to a new attack that can covertly steal two-factor authentication codes, location timelines, and other private data in less than 30 seconds.
The new attack, named Pixnapping by the team of academic researchers who devised it, requires a victim to first install a malicious app on an Android phone or tablet. The app, which requires no system permissions, can then effectively read data that any other installed app displays on the screen. Pixnapping has been demonstrated on Google Pixel phones and the Samsung Galaxy S25 phone and likely could be modified to work on other models with additional work. Google released mitigations last month, but the researchers said a modified version of the attack works even when the update is installed.
Like taking a screenshotPixnapping attacks begin with the malicious app invoking Android programming interfaces that cause the authenticator or other targeted apps to send sensitive information to the device screen. The malicious app then runs graphical operations on individual pixels of interest to the attacker. Pixnapping then exploits a side channel that allows the malicious app to map the pixels at those coordinates to letters, numbers, or shapes.
Measles outbreak in SC sends 150 unvaccinated kids into 21-day quarantine
Health officials in South Carolina are warning that the highly infectious measles virus is spreading undetected in communities in the northern part of the state, specifically Spartanburg and Greenville counties.
Last week, officials in Greenville identified an eighth measles case that is potentially linked to the outbreak. Seven outbreak cases had been confirmed since September 25 in neighboring Spartanburg, where transmission was identified in two schools: Fairforest Elementary and Global Academy, a public charter school.
Across those two schools, at least 153 unvaccinated children were exposed to the virus and have been put in a 21-day quarantine, during which they are barred from attending school, state officials said in a press conference. Twenty-one days is the maximum incubation period, spanning from when a person is exposed to when they would develop a rash if infected.
Google’s Photoshop-killer AI model is coming to search, Photos, and NotebookLM
Google began experimenting with conversational image editing earlier this year in the dev-focused AI studio, but the feature didn't remain experimental for long. Over the summer, Google rolled out the "Nano Banana" image-editing model in Gemini 2.5 Flash. You can use this feature to modify images with just a prompt, and now you don't even need to go to Gemini to use it. Google says Nano Banana is now coming to search, Google Photos, and NotebookLM.
The AI image editor is coming to search via Lens and AI Mode. For Lens, you can simply open the app (iOS and Android) and snap a photo to get started. When the rollout is complete, you'll see a "Create" button at the bottom, with a banana icon. Tap that to enter a prompt, telling the AI how you'd like the photo changed.
When you begin an edit in Lens, the Google app will display the results and offer the chance for follow-up edits in the AI Mode interface. Google is always looking for more ways to get people plugged into its conversational search bot, so there's also a separate way to access Nano Banana there. Simply select the "Create image" tool and enter your prompt to create an image. You can then continue the conversation to have Nano Banana change the image.
Starship’s elementary era ends today with mega-rocket’s 11th test flight
SpaceX is set to launch the 11th full-scale test flight of the company's Starship rocket Monday evening, with hopes of capping a tumultuous year with a successful one-hour voyage from South Texas to the Indian Ocean.
Liftoff of the Super Heavy booster with the Starship upper stage is scheduled for 6:15 pm CDT (7:15 pm EDT; 23:15 UTC). SpaceX has a 75-minute window to launch Monday. You can watch a livestream of the flight here.
SpaceX's control team, positioned a couple of miles away from the launch pad at Starbase, Texas, will oversee the loading of more than 10.5 million pounds of super-cold methane and liquid oxygen into the two-stage rocket beginning about an hour before liftoff. In the final minutes of the countdown, the world's largest rocket will undergo a steering check, and the launch director will give a final "go" for launch.
To shield kids, California hikes fake nude fines to $250K max
California is cracking down on AI technology deemed too harmful for kids, attacking two increasingly notorious child safety fronts: companion bots and deepfake pornography.
On Monday, Governor Gavin Newsom signed the first-ever US law regulating companion bots after several teen suicides sparked lawsuits.
Moving forward, California will require any companion bot platforms—including ChatGPT, Grok, Character.AI, and the like—to create and make public "protocols to identify and address users’ suicidal ideation or expressions of self-harm."
Apple’s streaming service gets harder to tell apart from its streaming app, box
Apple has lightly rebranded its video-on-demand streaming service. The Netflix rival that has brought us critically acclaimed shows and movies like Slow Horses and The Lost Bus has gone from Apple TV+ to Apple TV.
Apple announced the name change today in a press release that was primarily about the film F1: The Movie coming to its streaming service on December 12. Unlike previous announcements, however, today’s release referred to the streaming service as Apple TV, instead of Apple TV+. The announcement reads:
Apple TV+ is now simply Apple TV, with a vibrant new identity.
Apple didn’t specify how its streaming service’s “identity” has changed at all. As of this writing, accessing Apple’s streaming service via a browser or smart TV app still shows the original Apple TV+ branding.
Why Signal’s post-quantum makeover is an amazing engineering achievement
The encryption protecting communications against criminal and nation-state snooping is under threat. As private industry and governments get closer to building useful quantum computers, the algorithms protecting Bitcoin wallets, encrypted web visits, and other sensitive secrets will be useless. No one doubts the day will come, but as the now-common joke in cryptography circles observes, experts have been forecasting this cryptocalypse will arrive in the next 15 to 30 years for the past 30 years.
The uncertainty has created something of an existential dilemma: Should network architects spend the billions of dollars required to wean themselves off quantum-vulnerable algorithms now, or should they prioritize their limited security budgets fighting more immediate threats such as ransomware and espionage attacks? Given the expense and no clear deadline, it’s little wonder that less than half of all TLS connections made inside the Cloudflare network and only 18 percent of Fortune 500 networks support quantum-resistant TLS connections. It's all but certain that many fewer organizations still are supporting quantum-ready encryption in less prominent protocols.
Triumph of the cypherpunksOne exception to the industry-wide lethargy is the engineering team that designs the Signal Protocol, the open source engine that powers the world’s most robust and resilient form of end-to-end encryption for multiple private chat apps, most notably the Signal Messenger. Eleven days ago, the nonprofit entity that develops the protocol, Signal Messenger LLC, published a 5,900-word write-up describing its latest updates that bring Signal a significant step toward being fully quantum-resistant.
4chan fined $26K for refusing to assess risks under UK Online Safety Act
A battle over the United Kingdom's Online Safety Act (OSA) heated up Monday as UK regulator Ofcom fined the notorious image-hosting board 4chan about $26,000 for failing to provide a risk assessment detailing the potential harms of illegal content hosted on its forum.
In a press release provided to Ars, Ofcom said 4chan refused to respond to two requests for information that the regulator considered "routine." The first asked for the risk assessment and the second for 4chan's "qualifying worldwide revenue."
4chan was anticipating the Monday fine, noting in a lawsuit—which was jointly filed with the online trolling forum Kiwi Farms in August and seeks to permanently enjoin Ofcom from enforcing OSA—that Ofcom had made it clear that because 4chan ignored Ofcom's emails, the fine was coming.
Hans Koenigsmann, who investigated all of SpaceX’s rocket failures, is going to space
Hans Koenigsmann is one of SpaceX's earliest, longest-tenured, and most-revered employees.
When Elon Musk started the company in 2002, he was joined by two other "founding" employees, Tom Mueller in propulsion and Chris Thompson in structures. Koenigsmann was the next hire, brought on to develop avionics for the Falcon 1 rocket.
Koenigsmann remained at the company for two decades before leaving SpaceX in late 2021. During that time, he transitioned from avionics to lead mission assurance and safety while also spearheading every major failure investigation of the Falcon 9 rocket. He was a beloved leader and mentor for his employees within the company's demanding culture.
Blue Origin aims to land next New Glenn booster, then reuse it for Moon mission
There's a good bit riding on the second launch of Blue Origin's New Glenn rocket.
Most directly, the fate of a NASA science mission to study Mars's upper atmosphere hinges on a successful launch. The second flight of Blue Origin's heavy-lifter will send two NASA-funded satellites toward the red planet to study the processes that drove Mars's evolution from a warmer, wetter world to the cold, dry planet of today.
A successful launch would also nudge Blue Origin closer to winning certification from the Space Force to begin launching national security satellites.
Why iRobot’s founder won’t go within 10 feet of today’s walking robots
When a robotics pioneer who has spent decades building humanoid machines recommends that you stand at least nine feet away from any full-sized walking robot, you should probably listen.
"My advice to people is to not come closer than 3 meters to a full-size walking robot," Rodney Brooks writes in a technical essay titled "Why Today’s Humanoids Won’t Learn Dexterity" published on his blog last week. "Until someone comes up with a better version of a two-legged walking robot that is much safer to be near, and even in contact with, we will not see humanoid robots get certified to be deployed in zones that also have people in them."
Brooks, the MIT professor emeritus who co-founded iRobot (of Roomba fame) and Rethink Robotics, believes companies pouring billions into humanoid development are chasing an expensive fantasy. Among other problems yet to be addressed, he warns that today's bipedal humanoids are fundamentally unsafe for humans to be near when they walk due to the massive kinetic energy they generate while maintaining balance. That stored-up energy can cause severe injury if the robot falls or its limbs strike someone.
RFK Jr. drags feet on COVID-19 vaccine recommendations, delaying shots for kids
As respiratory virus season draws near, no federal health official has signed off on recommendations for this year's updated COVID-19 vaccines, leaving the federal vaccination schedule without an update and access to the shots in limbo for some low-income children.
According to reporting by Stat news earlier this week, two immunization program experts—who spoke on the condition of anonymity to avoid reprisal from the Trump administration—said that because there is no sign-off, states are not yet able to order COVID-19 shots for children who receive vaccines through the Vaccines for Children (VFC) program. The federal program provides vaccines to children who are Medicaid-eligible and under- or uninsured, which includes around half of all children in the US.
Typically, federal vaccination recommendations come about after a panel of expert advisors for the Centers for Disease Control and Prevention—the Advisory Committee on Immunization Practices (ACIP)—publicly reviews vaccine data and votes on recommendations for use. Then, the CDC director decides whether to adopt those recommendations. While directors don't always adopt ACIP's recommendations, they usually do—and often on the same day as the ACIP vote. After that, the recommendation becomes part of federal vaccine guidance, and insurance companies and federal programs are required to cover the recommended shots.