General Motors announced that it will cut more than 1,700 manufacturing jobs in reaction to changes in the electric vehicle market. "In response to slower near-term EV adoption and an evolving regulatory environment, General Motors is realigning EV capacity," the company said in a statement reported by CNBC. "Despite these changes, GM remains committed to our US manufacturing footprint, and we believe our investments and dedication to flexible operations will make GM more resilient and capable of leading through change."
The layoffs are primarily happening at a Michigan plant that builds GM's EVs and at an Ultium Cells battery cell plant in Ohio. The company is also "temporarily" laying off 700 at an Ultium Cells plant in Tennessee.
The regulatory issues in question are likely the $7,500 federal tax rebate that had previously been granted to EV purchases, which expired earlier this year under the "Big Beautiful Bill" that made things a lot less pretty for many environmentally-focused programs and industries. But GM had also said earlier this month that it would sunset much of its hydrogen fuel cell R&D in order to place more focus on batteries, charging tech and EVs, so announcing layoffs in those very areas is a rough move.
This article originally appeared on Engadget at https://www.engadget.com/gm-will-cut-more-than-1700-jobs-in-ev-and-battery-manufacturing-200814378.html?src=rss
We all know that space is the next major frontier that remains to be conquered by humans, sparked by Elon’s ambitions to colonize Mars. Previously, NASA and now SpaceX are spearheading the dreams of space tourism and ultimately moving beyond planet Earth’s nurture. These developments have triggered inspiration for countless concept vehicles that envision what the future could be like.
This future-forward concept overshadows most of them with its unique idea of connecting Earth and space travel without all the strings attached. The idea is to make a seamless transition from driving on the highways to traveling beyond the stratosphere within just a few hours. Meet the Polestar Kairos, a hypertravel vehicle that comes with a normal driving mode and a space transfer mode to dock with a spacecraft that’s destined for space.
The front section of the hypercar, made out of reinforced lightweight material, is stronger than the fuselage of an airplane for structural integrity. Transition from the car mode to the docked spacecraft version is seamless, indicating how we could travel in space in the coming decades. Sure, it sounds like science fiction, but we are talking of a time ten years down the line when traveling to Mars would be a usual affair. This concept is based on the Polestar’s DNA and design physics, thereby explaining the industrial look.
The compatible spacecraft will be capable of docking two Kairos hypercars, transporting them to the other habitable planet in the solar system. When docked with the spacecraft, the hypercar unit transforms into an ultra-low temperature sleep chamber. This chamber is loaded with amenities for life maintenance during extended space trips. Since space travel can have a toll on mental and physical conditioning, the module is equipped to handle that too. The spacecraft will essentially be a carrier for people who want their loved ride to stay with them even beyond the limits of Earth.
From a pure four-wheeler perspective, the Kairos concept arrives at a time when Polestar is sharpening its focus on electric mobility and design-led sustainability. While the brand’s recent concepts, such as the Precept, have already charted its future direction, Kairos brings new aesthetic and experiential ideas to the table, melding dynamic proportions, refined materials. The Hypercar is low, wide, and sculpted, with crisp surfaces that emphasize tension and purity. The roofline sweeps seamlessly into the rear haunches, giving the car an almost coupe-gravity despite a potentially four-door or high-performance nature. The front face carries Polestar’s signature front “Thor’s Hammer” lighting graphic while a stubby profile reinforces a performance posture.
Inside, the Kairos concept highlights the brand’s commitment to premium yet sustainable materials. Natural-fiber composites, recycled textiles, and minimalist interfaces dominate. The cabin layout puts the driver firmly in command: clean surfaces, clear lines, and a restrained palette convey calm focus rather than techno-clutter. The floating center console, integrated display architecture, and expansive glazing underline a sense of openness and precision.
A decade ago, when smart speakers with built-in voice assistants were the hot new thing, many imagined they'd quickly evolve into highly intelligent companions. (Think C-3PO or Star Trek's Data living inside a speaker.) That road has been much longer than expected, as virtual helpers like Alexa, Siri and Google Assistant seemed to sit in neutral for years.
But now that generative AI is here (for better or worse), smart speakers are finally scratching the surface of those expectations. Google's new version, Gemini for Home, is now available to try. Here's how.
First, keep in mind that the Gemini for Home voice assistant is in early access. This means Google is gathering feedback about its features, and — as with all generative AI — it's wise to assume it will make mistakes. If or when it does, you can send feedback to Google in the Google Home app or by saying, "Hey Google, send feedback."
Devices compatible with Gemini for Home voice assistant
Other models support everything but Gemini Live. That list includes the Google Nest Wifi point, Google Nest Hub (1st gen), Google Home Max, Google Home Mini (1st gen) and Google Home.
Another point is that once you dive in, your Google Assistant days will be over (on your speakers, anyway). That's because Google says that, once you upgrade to Gemini for Home, your compatible devices can't downgrade to Assistant. That shouldn't be a problem, but it's worth keeping in mind before you take the plunge.
How to sign up for early Gemini access
Once you've confirmed that your speaker(s) are at least partially compatible, head to the Google Home app on a mobile device. There, tap your profile picture (or initials) on the top right. Then tap Home Settings > Early Access. Congratulations: You've put in your request.
The bad news is you may have to wait a bit to confirm your entry into the beta program. Once you're in, you'll see a notification from the Google Home app that reads, "Introducing Gemini for Home." Select that, and follow the prompts. (If you accidentally dismiss the notification, you'll see the setup banner under Home settings in the Google Home app.)
Cherlynn Low for Engadget
At that point, all compatible speakers in your home will be upgraded to Google's more intelligent AI assistant. You can now throw more advanced questions at it, similar to what you'd ask text-based chatbots like ChatGPT. Except this one sits on a shelf, ready to field your verbal requests at any moment.
Keep in mind that Gemini Live requires a Google Home Premium subscription. The standard version costs $10 per month or $100 per year. Meanwhile, the advanced tier doubles that: $20 per month or $200 per year. At least for now, the only difference between the two (for these purposes) is that the pricier plan supports a camera history search feature. Both premium tiers unlock access to Gemini Live. So, if that's all you need, you can save money and get standard.
Google offers plenty of examples to get started. You can get quick answers to facts, like "Hey Google, who are the top five scoring players in basketball history?" (FYI: James, Abdul-Jabbar, Malone, Bryant and Jordan.) You can also ask Gemini Live to have a chat about ingredients for people with dietary needs. Or, ask it to explain complex topics (like how Wi-Fi works) in simple terms. It may not be at C-3PO level yet, but it's certainly moving in that direction.
This article originally appeared on Engadget at https://www.engadget.com/ai/early-access-for-gemini-home-voice-assistant-is-now-available-heres-how-to-get-it-191250927.html?src=rss
Sony just announced the PS Plus Monthly games for November and there are some standouts. All of these titles will be available to download on November 4 until December 1.
First up, there's the cat sim Stray. We praised the third-person cyberpunk adventure in our initial review, in which we lauded the game's unique mechanics and puzzles. After all, you play as a cat in a dystopian and futuristic city. That's a one-of-a-kind premise, which is one reason it became such a big hit. It'll be available for both PS4 and PS5. It's also worth noting that this was a day-one release for PS Plus Extra and Premium back in 2022.
EA Sports WRC 24 is a popular racing sim developed by Codemasters, who made the Dirt series, the Grid franchise and plenty of official F1 games. This new one lets players participate in the FIA World Rally Championship (WRC.) Racers get access to a bunch of WRC, WRC2 and Junior WRC cars from the 2024 season, complete with current drivers and teams. This one is only for PS5.
Totally Accurate Battle Simulator is certainly a battle similar, but not an accurate one. The game lets you design all kinds of bizarre battle scenarios with "the wobbliest physics system ever created." This is the best game ever made that lets you battle a wooly mammoth with balloon arrows, thus forcing that beast into the sky. It'll be available for both PS4 and PS5.
As always, a crop of new games means that some pre-existing titles are heading to that farm upstate. You have until November 3 to download Alan Wake 2, Goat Simulator 3 and Cocoon to their game library. Dang. Last month was stacked.
This article originally appeared on Engadget at https://www.engadget.com/gaming/playstation/stray-leads-novembers-lineup-of-ps-plus-games-180339318.html?src=rss
Character.AI will no longer permit teenagers to interact with its chatbots, as AI companies face increasing pressure to better safeguard younger users from harm. In a statement, the company confirmed that it is removing the ability for users under 18 to engage in any open-ended chats with AI on its platform, which refers to back-and-forth conversations between a user and a chatbot.
The changes come into effect on November 25, and until that date, Character.AI will presents users with a new under-18 experience. It'll encourage its users to use chatbots for creative purposes that might include, for example, creating videos or streams, as opposed to seeking companionship. To manage the transition, under-18s can now only interact with bots for up to two hours per day, a time limit the company says it will reduce in the lead-up to the late November deadline.
Character.AI is also introducing a new age assurance tool it has developed internally, which it says will "ensure users receive the right experience for their age." Along with these new protections for younger users, the company has founded an "AI Safety Lab" that it hopes will allow other companies, researchers and academics to share insights and work collaboratively on improving AI safety measures.
Character.AI said it has listened to concerns from regulators, industry experts and concerned parents and responded with the new measures. They come after The Federal Trade Commission (FTC) recently launched a formal inquiry into AI companies that offer users access to chatbots as companions, with Character.AI named as one of seven companies that had been asked to participate. Meta, OpenAI and Snap were also included.
Both Meta AI and Character AI also faced scrutiny from Texas Attorney General Ken Paxton in the summer, who said chatbots on both platforms can "present themselves as professional therapeutic tools" without the requisite qualifications. Seemingly to put an end to such controversy, Character.AI CEO Karandeep Anand told TechCrunchthat the company’s new strategic direction will see it pivot from AI companion to a "role-playing platform" focused on creation rather than mere engagement-farming conversation.
The dangers of young people relying on AI chatbots for guidance has been the subject of extensive reporting in recent months. Last week, the family of Adam Raine, who claim that ChatGPT enabled their 16-year-old son to take his own life, filed an amended lawsuit against OpenAI for allegedly weakening its self-harm safeguards in the lead-up to his death.
This article originally appeared on Engadget at https://www.engadget.com/ai/characterai-to-ban-teens-from-talking-to-its-chatbots-180027641.html?src=rss
California-based AI and robotics company 1X is now accepting pre-orders for its humanoid robot NEO, which was designed to automate everyday chores and to offer personalized assistance. Users will be able to control NEO and have it accomplish tasks around the house with the click of a button or a verbal command. It will come with the ability to do basic tasks autonomously when it starts shipping next year, including opening doors, fetching items and turning the lights on or off. However, if early adopters want NEO to be capable of more specific or complex tasks, they'd have to be comfortable with the idea of a human teleoperator controlling the robot remotely and seeing inside their homes.
In an interview with The Wall Street Journal's Joanna Stern, 1X CEO Bernt Børnich explained that the AI neural network running the machine still needs to learn from more real-world experiences. Børnich said that anybody who buys NEO for delivery next year will have to agree that a human operator will be seeing inside their houses through the robot's camera. It's necessary to be able to teach the machines and gather training data so it can eventually perform tasks autonomously. "If we don't have your data, we can't make the product better," he said.
Børnich admitted that much of the work will be done by teleoperators in the beginning. Owners will have access to an app where they can schedule when the teleoperator can take over NEO and where they can specify the task they want the machine to do. He said 1X is putting control in the hands of the owner to respect people's privacy as much as possible. The company can blur people so that the remote operator doesn't see them, and owners can designate no-go zones in their homes that the operator cannot go to. Teleoperators also cannot take control of NEO without the owner's approval. Of course, there's always potential security breaches to think of — Børnich at least assured that NEO has several layers of security to prevent it from hurting people.
If you want to see more of the robot in action, the WSJ video is definitely worth a watch.
1X NEO is available in tan, gray and dark brown. It's now available for pre-order from the company's website with a deposit of $200. Those who want early access to it can get it for $20,000, but it will also be available as a subscription service of $499 a month.
This article originally appeared on Engadget at https://www.engadget.com/ai/1x-neo-is-a-20000-home-robot-that-will-learn-chores-via-teleoperation-040252200.html?src=rss
Most docking stations are either a tangle of cables or a bland box you hide behind your monitor out of shame and embarrassment. For creative pros and MacBook power users, finding a dock that keeps up with demanding workflows while actually looking good on your desk is a rare win that deserves celebrating. Most people settle for function over form, accepting ugly tech as the unavoidable price of modern productivity.
The iVANKY FusionDock Max 2 changes the game with a design that’s as visually striking as it is functionally capable. With Thunderbolt 5, native triple-display support for Apple Silicon Macs, and a chassis built for both performance and aesthetics, it’s a dock you’ll want to show off front and center on your desk, not hide away beneath it in shame.
The FusionDock Max 2 is machined from a single block of aluminum using a 2,000-ton extrusion press, giving it a dense, premium feel that most plastic or stamped-metal docks simply can’t match. The matte dark finish, copper-alloy midframe for efficient heat dissipation, and bold orange accent lines create a modern, professional look that stands out in any workspace without screaming for attention.
The floating chassis design and visible fan vents aren’t just practical engineering choices—they’re deliberate aesthetic statements that celebrate the dock’s industrial elegance and functional beauty. The elevated base allows air to flow naturally underneath, while the integrated fan remains whisper-quiet during operation. The orange accents frame the port array, turning functional connectivity into visual interest that enhances rather than detracts.
With Thunderbolt 5 and a dedicated DisplayPort-Alt chip working together seamlessly, the FusionDock Max 2 is the first dock to offer native triple-display support for Apple Silicon MacBooks without adapters, dongles, or driver headaches. Video editors, developers, and designers can finally spread their work across three external monitors while keeping their MacBook lid closed for a cleaner, more organized setup.
The dock packs 23 thoughtfully chosen ports selected specifically for professional workflows rather than just chasing numbers. Five 10Gbps USB-C ports, four 10Gbps USB-A ports, dual HDMI 2.1, DisplayPort 1.4, 2.5G Ethernet, SD 4.0 and microSD 4.0 card readers with 312MB per second transfer speeds, optical audio, and a security lock slot cover virtually every professional need imaginable.
Thunderbolt 5 delivers up to 120Gbps bandwidth—triple that of Thunderbolt 4—enabling smooth 6K displays, lightning-fast file transfers, and zero-compromise workflows for creative professionals. The dock powers your MacBook Pro at a steady 140W even with every single port busy and drawing power, while a separate 30W USB-C port handles phones and tablets without stealing power from the host machine.
Active cooling with a copper midframe and floating chassis keeps the dock up to 20 degrees Celsius cooler than typical Thunderbolt 5 docks during sustained full-load operation. This thermal management prevents flickering displays, dropped connections, and performance throttling that plague lesser docks when you push them hard during intensive video rendering, large file transfers, or multi-monitor coding sessions.
The FusionDock Max 2 is designed to be seen, appreciated, and admired rather than hidden from view. Its industrial elegance, compact footprint, and thoughtful cable management make it a natural centerpiece for creative studios and home offices that value both performance and presentation equally in their workspace design.
YouTube is rolling out some updates aimed at making visuals look better on its TV apps, including automatic AI video upscaling. To begin with, YouTube plans to upscale videos that were uploaded in under 1080p to an HD resolution. It aims to support 4K upscaling in the future.
The platform will still retain the original files and video resolution options. Creators can opt-out of the AI upscaling feature, which it's calling Super Resolution, as well as automated audio adjustments. The platform says it's also working with select creators to test larger video uploads to allow for higher-quality footage.
Creators will soon be able to upload thumbnails in higher quality too. YouTube will increase the thumbnail file limit from 2MB to 50MB.
As for viewers, it seems like YouTube saw those annoying automatic previews Netflix has had for the last decade and decided to copy that playbook. "Viewers will be able to see and flip through their favorite YouTube channels with immersive previews on the homepage, enhancing content discovery and engagement," Kurt Wilms, the senior director of product management for YouTube on TV, wrote in a blog post.
Elsewhere, YouTube has added a contextual search function to its TV apps. When you search for something from a creator's channel page, videos from said channel will appear first in the results. That's a smart, logical idea.
With people watching YouTube on TV more than on mobile these days, it only makes sense for the platform to invest in improving the experience there.
This article originally appeared on Engadget at https://www.engadget.com/entertainment/youtube/youtube-adds-automatic-ai-upscaling-for-low-res-videos-170342511.html?src=rss
It looks like a hundred thousand people fell for a fake NVIDIA livestream featuring an AI-generated version of CEO Jensen Huang, as reported by PC Gamer. Perhaps the scariest part is that the fake stream ran at the same time as an actual NVIDIA event and dwarfed the live viewership numbers.
The actual keynote speech of NVIDIA's GPU Technology Conference (GTC) garnered around 20,000 live views, while the fake stream maxed out at 100,000 live views. Even weirder? The fake Huang was talking about some crazy stuff mostly involving bogus crypto investments.
The deepfake spoke of "a crypto mass adoption event that ties directly into NVIDIA's mission to accelerate human progress." The avatar urged viewers to scan a QR code to, uh, send in cryptocurrencies. It's unclear if any of the 100,000 viewers fell for this obvious scam that asked people to send the world's richest company money to "accelerate human progress."
The fake video has since been deleted. I haven't been able to check it out to see just how real it looked and, thereby, how it was able to lure in 100,000 viewers. Before we all start screaming into the night about how reality doesn't matter anymore, there are a few things worth considering.
First of all, we don't know anything about the 100,000 accounts that were watching the fake stream. We don't know where they're from or even if the accounts were tied to real people. It's also worth noting that the real stream has since garnered 200,000 views, despite just having 20,000 live viewers. We don't know what kind of promotional tools the fakers used to draw in people or how long people stayed once they clicked.
Finally, there's Huang himself. The man has hosted four GTC conferences just this year, and dozens upon dozens in previous years. There is an absolute abundance of footage of him standing on a stage and talking about stuff. That's a whole lot of video for bad actors to use as AI training data.
Also, the real stream wasn't exactly a barn burner. The most notable news we got was the announcement of a partnership with Uber to promote autonomous driving. This wasn't an event to hype up new graphics cards or anything like that. The company revealed some government contracts, if that's your bag.
This article originally appeared on Engadget at https://www.engadget.com/ai/more-people-watched-a-fake-nvidia-livestream-than-the-real-thing-170035672.html?src=rss
You know those moments when something so simple makes you stop and stare? That’s exactly what Bo Zhang is doing with his latest collection, and honestly, it’s kind of mind-blowing. Lightware is a series of transparent glass trays that don’t just sit there looking pretty (though they do that too). They actually play with light and shadow to create this ever-changing visual experience that transforms depending on where you stand, how the light hits them, and what surface they’re resting on.
If you caught wind of Zhang’s previous work, the Ripples collection, you already know he’s got a thing for messing with our perception. But where Ripples played with color-shifting optical illusions through hexagonal folded forms, Lightware takes a completely different approach. This time, it’s all about harnessing the science of polarizing film and glass to create what I can only describe as shadow choreography.
The genius here is in the subtlety. Zhang, who founded his studio Desz Office in New York back in 2019, has always been about creating pieces that interact with people on both visual and emotional levels. With Lightware, he’s tapping into something we usually take for granted: the way light moves through transparent objects. By layering polarizing film within the glass structure, these trays generate dynamic geometric patterns that shift and dance as you move around them or as the light source changes throughout the day.
Picture this: you set one of these trays on your coffee table in the morning. The sunlight streaming through your window casts one pattern. By afternoon, with the light coming from a different angle, it’s transformed into something completely new. Move it to another surface, and the shadows beneath it create yet another composition. It’s like having functional art that never gets boring because it literally never looks the same twice.
What’s really cool is that Zhang isn’t just making pretty objects for the sake of it. His design philosophy revolves around the idea that good artwork should have a soul, not be a cold entity. These trays are practical (you can actually use them to hold your keys, jewelry, or whatever), but they’re also conversation starters, meditation pieces, and honestly just really satisfying things to look at.
The recognition is rolling in too. Lightware recently won the NYCxDesign Awards 2025 Honoree in the accessories category, adding to Zhang’s growing list of accolades that includes being named one of UNESCO’s “100 Future Artists” and getting featured in Design Miami’s DMBX project. The design world has definitely been paying attention to what he’s doing, with features in publications like Vogue, Interior Design magazine, and Design Milk.
What I love about this collection is how it bridges the gap between tech-savvy innovation and accessible design. You don’t need to understand the science of polarization to appreciate these trays (though if you’re curious, polarizing film works by filtering light waves in specific directions, creating those mesmerizing patterns). You just need to be someone who appreciates when everyday objects transcend their basic function and become something more.
Since we live now with a lot of digital stimulation, there’s something refreshing about analog magic. Lightware offers this quiet, contemplative kind of wonder. It reminds us that transformation doesn’t always have to be loud or flashy. Sometimes the most captivating changes happen slowly, subtly, right in front of us, as naturally as light moving across a room.
Zhang’s work with Lightware proves that innovation in design isn’t always about reinventing the wheel. Sometimes it’s about looking at something as fundamental as a tray and asking, “What if this could do more?” The answer, in this case, is a collection that turns functional objects into portals of perpetual surprise, where shadows become art and transparency reveals hidden complexity.
