Goldman Sachs, in a note this week, via India Dispatch: There are various reasons that explains this: (i) A large part of the global education spend goes towards formal education (schools, colleges and universities), which are typically either run by governments or are not-for-profit institutions;
(ii) It is difficult to replicate education quality at scale in our view, since most teachers would have a different pedagogy, and thus standardization is harder to achieve vs that in other internet categories;
(iii) Education is fragmented – it includes various fields (schools, undergrad courses, medicine, engg, management, etc.), each with their own curriculum, and the same being vastly different across countries globally; this makes scalability difficult beyond a few certain specializations and regions.
Additionally, we believe the ability for online education to capture a sizable value share of supplemental education is limited since the perceived value of offline, including that from community, in-person engagement and doubt solving, rigour, etc., is typically higher.
However, we note that before China’s double reduction policy in 2021, TAL and EDU had market caps of up to US$50 bn; these companies were mostly domestic focused and on the K-12 tutoring segment, which has large volumes. Similarly in India, Byju’s reached a peak valuation of US$20 bn+ (link; again, focused on K-12), before issues around governance etc. impacted the business.
We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.
At $749, the unlocked 128GB Google Pixel 10 Pro has slipped into its most affordable price yet on Amazon, down from its original $999 launch price, according to online price trackers. Stack the promo code PIXEL10, and you shave off another $100, which reframes the phone entirely. At full price, it felt like a confident but pricey Android flagship. At this discount, it becomes a far easier recommendation for people who value polish over raw benchmark wins.
This is still very much a Pixel. The design is clean and familiar, with a flat display, matte glass back, and a solid in-hand feel that avoids the slippery problem of earlier generations. It runs Android as Google intends it to run, with fast updates and a long support window that continues to be one of Pixel’s quiet advantages.
Most of what the Pixel 10 Pro does well will feel familiar to longtime Pixel fans. The triple-camera system is reliably consistent, with low-light photos that stay natural instead of heavily processed, portraits that rarely miss, and everyday shots that come out right without needing a second attempt. The display gets a modest but useful bump in brightness, which helps outdoors, and Qi2 wireless charging brings faster, more reliable cable-free top-ups. Google’s new Pixelsnap magnetic system fits cleanly into daily routines, snapping chargers and accessories into place much like Apple’s MagSafe, without feeling forced. Lifehacker’s Associate Tech Editor Michelle Ehrhardt notes in her review that this kind of convenience adds up, especially for car mounts or desk charging. One real trade-off is the move to eSIM-only support. That is fine if your carrier supports it, but it can be limiting if you switch networks often or rely on physical SIMs while traveling.
Where the Pixel 10 Pro feels less settled is in performance expectations. The Tensor G5 chip is smooth for everyday use, but it does not match Snapdragon rivals in demanding third-party apps, as noted in this PCMag review. Google’s latest AI features sound impressive, but many feel optional rather than essential, not exactly reasons to upgrade. Battery life is solid but unremarkable. It lasts a full day with moderate use, though heavier users may need a top-up before night. All told, this discounted Pixel 10 Pro is not aimed at spec chasers. It is a balanced, camera-first Android phone with thoughtful refinements, now priced in a way that finally makes sense.
One of the interesting open-source projects to come about this year was Wayback as an X11 compatibility layer using Wayland. Wayback could be used by default on Alpine Linux next year among other distributions. For ending out 2025 development, Wayback 0.3 is now available…
The House of Representatives cleared the way for a massive overhaul of the federal environmental review process last Thursday, despite last-minute changes that led clean energy groups and moderate Democrats to pull their support.
The Standardizing Permitting and Expediting Economic Development Act, or SPEED Act, overcame opposition from environmentalists and many Democrats who oppose the bill’s sweeping changes to a bedrock environmental law.
The bill, introduced by Rep. Bruce Westerman (R-Ark.) and backed by Rep. Jared Golden (D-Maine), passed the House Thursday in a 221-196 vote, in which 11 Democrats joined Republican lawmakers to back the reform effort. It now heads to the Senate, where it has critics and proponents on both sides of the aisle, making its prospects uncertain.
I’ve always wondered what it would be like to own a plug-in hybrid, and recently, fate handed me that opportunity. On a recent trip to Vancouver, I rented a 2024 Toyota Prius Prime for nearly two months — the ideal scenario to try out North America’s most popular PHEV.
My experience with the Prius Prime
Previously, the words “Prius” and “sexy” were rarely used in the same sentence. However, I think the wedge-shaped Prius Prime introduced for 2023is much sexier than its frumpy predecessors. The sleek shape also pierces the wind better to improve efficiency. It’s lower to the ground than before, though, which can make entry tough for taller or older people.
I was comfortable in the Prius Prime once seated, even though the materials and options aren’t quite as luxurious as other PHEVs sold in the US. On two 10-hour highway drives up to northern Canada I never felt sore (or cold) in the well-bolstered, heated seats. However, visibility wasn’t the greatest due to the low seating position and thick front pillars that occasionally blocked my view of traffic.
Steve Dent for Engadget
With its wraparound dash and 8-inch touchscreen, the interior is reasonably high-tech but not to the standard of some EVs I’ve tried recently. It came with wireless CarPlay and Android Auto support that gave me seamless streaming entertainment on long highway stretches. The driver safety features (lanekeeping, adaptive cruise, automatic braking and more) also boosted my confidence in Vancouver’s gnarly traffic. The Prius Prime doesn’t offer true one-pedal implementation, but it has a mode that’s close to that.
With a two-liter 150 HP gas motor and 161 HP electric motor (net 220 HP combined), the 2024 (fifth generation) Prius Prime has a whopping 100 more horsepower than the previous model. The electric drivetrain is supplied by a 13.6kWh battery (10.9kWH usable) that takes four hours to charge at 240 volts, or double that with 120-volt household electricity. That means you can juice it fully overnight, but it doesn’t have DC fast-charging for speedy power-ups on longer trips. The EPA electric range is 44 miles, 19 more than the fourth-gen Prius.
It accelerated surprisingly well (from 0 to 60mph in 6.7 seconds) and was agile, but had a fair amount of body roll since it’s not designed for the race track. Still, considering the Prius’s reputation as a staid hippie econobox, the new model was downright sporty. I enjoyed driving in the all-electric mode much more than the hybrid mode, though — it was quieter and smoother, with lower noise levels and vibration.
So, how far was I able to drive on that electric motor alone? On the highway at about 65 MPH, I eked out 30 miles and just over 40 miles in the city. On one trip, I drove from the city center to a suburb 30 miles away and made it there and partially back on a full charge. On another voyage, I was able to drive back and forth between the east- and west-most points of Vancouver (13 miles) — a typical commute for many city-dwellers — with about a quarter charge to spare.
Steve Dent for Engadget
With consistent charging, my fuel consumption over a two week period (averaging 25 miles per day) was about a quarter of a tank or around $7.50. In terms of electricity, I used nearly 70.5kWh during that time at $.085/kWh, for a total of $6. All told, I spent $13.50 for gas and electricity over 350 miles of mixed driving, so the Prius Prime was clearly cheap to operate.
For longer trips, it’s still as inexpensive as it gets for a gasoline-powered vehicle, thanks to the efficient ICE motor and hybrid system that’s among the best in the industry. With a full battery charge and tank, I set out on a 547-mile drive and travelled 470 miles before stopping for gas, with a quarter tank to spare. That fill-up cost around $25.
The true savings and the problem with plug-in hybrids
The timing for my test of this car was ideal. In October, I spotted a European study concluding that PHEVs aren’t as economical as expected over a large sample size. Engadget’s article about that stirred up some passion among owners and potential buyers, so I wanted to compare my experience with points in the study.
First, let’s see if a PHEV is worth the extra money compared to a regular hybrid. My calculations are for the average US buyer and don’t take state or federal clean air rebates into account.
When I chose to rent a “compact” car, Avis assigned me a mid-range Prius Prime XSE — a model that lists at $37,320 but typically sells for $34,590, according to Edmunds. That suited me well as it only lacked a few features of the high-end XSE Premium, notably the larger 13.2-inch infotainment display and solar roof option. A fully equipped 2026 XSE Premium model with those features costs $41,665.
Steve Dent for Engadget
Since Toyota also makes a regular hybrid Prius, that vehicle offers an ideal comparison. The equivalent Prius XTE model has a list price of $31,995 in a similar configuration, making it $5,325 cheaper than the Prius Prime XSE.
The average US driver covers 13,662 miles per year and gasoline currently has a median $3 per gallon price. Over that distance, a non-PHEV Prius driver could expect to burn 273 gallons at 50 MPG (EPA combined) in a year, spending $819 on fuel.
A Prius Prime driver, on the other hand, would use 70 to 85 percent less fuel by current EPA or WLPT estimates. If we generously take the high end of those numbers at 85 percent, that cuts fuel costs to $160. That would require using about 2,500 kWh of electricity, though, so at an average US price of $0.18/kWh, that amounts to $450, for a total of $610 (gas plus electricity). That means you’d save just $209 in a year, or $2,090.00 over 10 years — not enough to justify the extra price. (Fuel and electricity prices, usage and other factors vary by region and can have a big impact on those figures.)
It could be even worse than that, according to a European automotive thinktank called Transport & Environment (T&E). After gathering real-world OBFCM data from 800,000 vehicles, they determined that PHEVs only run in all-electric mode 27 percent of the time, rather than 84 percent as estimated by Europe’s WLPT standard. As a result, plug-in hybrids in Europe emit five times more emissions and cost users €500 ($586) more per year than previously thought. Those figures are likely similar in the US.
Steve Dent for Engadget
How could regulators be so wrong about this key data? The first, obvious reason is that they underestimated how often people charge their vehicles. With their relatively short range, plug-in hybrids often need a full charge to get through the day in electric-only mode — but many people aren’t doing that.
Why? One reason may be a lack of easy charger access away from home. I found them to be difficult to find and use, often requiring a sign-up or app rather than just letting me tap a credit card (I’m looking at you, ChargePoint, Flo and Switch Energy). It can also be more expensive than just buying gasoline, since many companies charge triple or more the market rate for electricity. Another factor is that drivers of company or fleet PHEVs charge their vehicles less often than private owners.
There’s one additional and especially pernicious reason: The ICE engine often kicks in when PHEVs are supposedly running in all-electric mode, particularly with heavier sedans or SUVs. That’s because the electric motors alone aren’t powerful enough for maneuvers like passing.
Larger batteries can boost all-electric usage, but only to a point. Beyond 45 miles of range, emissions actually increase. The reason is simple: “Long-range PHEVs are the heaviest in the dataset, averaging 28 percent more mass and 33 percent more engine power than the group just below,” T&E wrote.
Steve Dent for Engadget
Overall, I enjoyed my time with the Prius Prime and found it to be fun, practical and cheap to drive. It’s the most economical PHEV because it has excellent electric range and enough power that the ICE engine rarely needs to kick in. At the same time, it offers the highest EPA mileage rating of any non-EV sold in North America. If I were in the market for a new vehicle, the Prius Prime would be high on my list.
However, I also learned that PHEVs aren’t reducing emissions or saving buyers as much as regulators and manufacturers have promised. Governments are to blame for much of that, as they overestimated all-electric use in PHEVs and failed to support the charging infrastructure needed to make them practical.
Responsibility also falls to automakers and buyers. Consumers want SUVs, but manufacturers aren’t making the electric motors in PHEVs powerful enough to run all the time in EV mode or offering fast DC charging. At the same time, drivers are failing to charge their vehicles consistently. Until those issues are solved, in my experience plug-in hybrids are a poor substitute for EVs in terms of emissions and a less economical choice than regular hybrids.
This article originally appeared on Engadget at https://www.engadget.com/transportation/toyotas-prius-prime-saved-me-gas-money-but-probably-not-the-environment-133027378.html?src=rss
Digital game codes remain one of the simplest ways to add new titles to a Nintendo library. The Nintendo Switch 2 handles code redemption in a familiar way, building on the same process used on the original Switch. You can redeem a download code directly from the eShop on the console, or through a browser on a phone or desktop. Both routes link the purchase to the same Nintendo Account so the redeemed title appears on any compatible device tied to that account. The process is quick once you know where to look. Here is how to redeem game codes on Switch 2 through the console and online.
Things to know before you start
A Nintendo Account is required and the Switch 2 must be connected to the internet for on-device redemption. Most codes contain sixteen characters made up of letters and numbers. The characters are displayed on a physical card, on a receipt or in a confirmation email from a retailer. The characters should be entered exactly as supplied. Each code can only be redeemed once and becomes linked to the Nintendo Account used during the process.
How to redeem a game code on the Switch 2 via the eShop
Redeeming a code through the Switch 2 itself is the most direct approach because the download begins as soon as the code is accepted.
Start on the Switch 2 home screen and open the Nintendo eShop icon. The console will prompt you to choose the user profile that should receive the content. Each user profile has a separate eShop purchase history, so pick the correct one before moving on.
Once the eShop loads, look for the Redeem Code option in the main menu. On the Switch 2, this appears in the left navigation panel alongside the store’s standard browsing categories. Select Redeem Code to open a screen with an input field and the on-screen keyboard.
Enter the sixteen-character code carefully. Codes may contain similar-looking characters particularly when numbers and letters appear in the same sequence. After entering the full set of characters select OK to move forward. The system checks the code and shows a confirmation page with the game or content tied to it.
If the code is valid, select Redeem to finish the process. The game will appear on the home screen as soon as the redemption is complete. A progress bar shows the download status until the installation finishes. If the game includes a Switch and Switch 2 version, the system will prompt you to choose which version to download.
If the game does not begin downloading head to the All Software view or the Download Management screen to confirm the status. The redeemed title should be visible with either a queued or active download.
How to redeem a Switch 2 code online
Nintendo also supports online redemption through its digital store, which is useful if you are away from the console or prefer to type the code on a full keyboard. The redeemed title will still appear in your account’s purchase history and can be downloaded later from the Switch 2.
Open a browser on a phone, tablet or computer and go to the Nintendo Account login page. Sign in using the account linked to your Switch 2. Once logged in navigate to the code redemption page for the Nintendo eShop. This page contains a text box where you can paste or type the code.
Enter the code exactly as provided and select Next. A confirmation screen appears with the game details. Select Redeem to finalize the action. The content is now tied to your account and ready to download on your console.
If the console is in sleep mode and has auto downloads enabled, the game may start downloading immediately. Otherwise open the eShop on the Switch 2 later and check the Download Management section for your queued title.
What to expect after redeeming a code
Every redeemed game becomes part of your Nintendo Account’s digital library. This applies across devices tied to the same account, which means that games purchased on a Switch or a browser will still appear when you sign in on a Switch 2. If the game supports cloud backup through Nintendo Switch Online, save data will sync once the game is installed.
Some games offer separate versions for the original Switch and Switch 2. When this applies, you’ll be asked to pick which version you want to download. The choice does not usually affect your ownership but it determines which software build is installed on the device.
If a code fails, Nintendo advises checking that the characters are entered correctly and confirming that you are using a download code rather than a control number, which often appears on the same card. Codes for DLC downloads require the base game on the destination console. Region restrictions also apply so the code must match the region of your Nintendo Account’s eShop.
Why Switch 2 uses the same redemption system as the original Switch
The Switch 2 keeps the familiar eShop structure that users have relied on since the original console. Keeping this process consistent simplifies the transition for families with multiple consoles and reduces the friction of carrying existing purchases forward. Nintendo Accounts continue to serve as the backbone for digital ownership which makes it easy to claim a code on one device and access the game on another.
The updated interface on the Switch 2 adds clarity to navigation but code redemption still works the same way on both platforms. This continuity ensures that retailers can distribute standard sixteen-character download codes and users can rely on simple steps to activate them.
This article originally appeared on Engadget at https://www.engadget.com/gaming/nintendo/how-to-redeem-game-codes-on-the-nintendo-switch-2-130059847.html?src=rss
Editor’s note: Warning: Although we’ve done our best to avoid spoiling anything major, please note this list does include a few specific references to several of the listed shows that some might consider spoiler-y.
This was a pretty good year for television, with established favorites sharing space on our list with some intriguing new shows. Streaming platforms reigned supreme, with Netflix and Apple TV dominating our list with seven and five selections each. Genre-wise, we’ve got a bit of everything: period dramas (The Gilded Age, Outrageous), superheroes (Daredevil: Born Again), mysteries (Ludwig, Poker Face, Dept. Q), political thrillers (The Diplomats, Slow Horses), science fiction (Andor, Severance, Alien: Earth), broody fantasy (The Sandman), and even an unconventional nature documentary (Underdogs).
As always, we’re opting for an unranked list, with the exception of our “year’s best” selection at the very end, so you might look over the variety of genres and options and possibly add surprises to your eventual watchlist. We invite you to head to the comments and add your own favorite TV shows released in 2025.
Ahead of CES 2026, Samsung has unveiled a number of impressive gaming monitors that flaunt 3D and very high refresh rates. The headliner is the 32-inch Odyssey 3D, “the world’s first 6K display with glasses free 3D,” according to Samsung. Along with the high resolution, it offers a 165Hz refresh rate (boosted to 330Hz through Dual Mode), along with a 1ms GtG response time. Optimized 3D effects that enhance terrain, distance and object separation “beyond 2D” will work with featured games like Stellar Blade and The First Berserker: Khazan,” according to Samsung.
Another model, the 27-inch Odyssey G6, brings refresh rates to a new level. It’s the world’s first 1,040 Hz monitor through Dual Mode and offers native QHD support up to 600Hz “to help players track targets and see fine details during high-speed movement,” Samsung claims. It can provide instant performance boosts and is AMD FreeSync and NVIDIA G-Sync compatible.
Samsung
If it’s OLED you want, Samsung has that covered as well with the 32-inch Odyssey OLED G8. It pairs a 4K QD-OLED panel with a 240Hz refresh rate, while offering glare-free viewing and VESA DisplayHDR True Black 500 certification, though brightness is limited to 300 nits. It’s DisplayPort 2.1 and HDMI 2.1 compatible with up to 80Gbps bandwidth, while supporting both AMD FreeSync Premium Pro and NVIDIA G-Sync.
Finally, Samsung introduced two other new high-resolution Odyssey G8 models as well. The first is the 32-inch G80HS model with 6K resolution at up to 165Hz and the other is the 5K, 27-inch Odyssey G8 with native support up to 180Hz and Dual Mode boosting that to 360Hz in QHD. There’s no word on pricing yet for any of these models, but we’ll likely learn that early next year.
This article originally appeared on Engadget at https://www.engadget.com/computing/accessories/samsungs-latest-odyssey-gaming-monitor-has-a-32-inch-6k-screen-with-glasses-free-3d-130051748.html?src=rss
Today, we’re releasing the second of three special “full map” Rebel Routes. These routes cover every road on their respective maps, and I’m releasing them in the hope that Zwift will adopt them as new fondo routes. (It’s been way too long since we’ve had new fondo routes, and there are so many great new roads that fondos could use!)
But first, once again, credit where credit is due: Zwifter Ole Eichhorn gets the kudos for putting in the work to come up with this route and even test ride it. Finding the shortest route covering all roads on a complex map is known as “the postman problem”, and Ole wrote a program to solve it.
This New York route would make a great new Bambino Fondo. (The first route we released, The Full Makuri, would make a great Medio Fondo.) The third route we’ll release, The Full Watopia, will make an epic Gran Fondo route. Stay tuned for that!
About Rebel Routes
“Rebel Routes” are Zwift rides not available on Zwift’s routes list, thus requiring manual navigation.
The reward for your rebel ride? Exploring a new route, knowing you’ve gone where few Zwifters have gone before. And a Strava segment rank in the tens or hundreds instead of the thousands! Rebel Routes are also included as a separate category on our Veloviewer Route Hunter leaderboard.
Route Description
A complete turn-by-turn tour of this route would take much too much space. So I’ll summarize it instead.
The route begins in Central Park, which is fitting, since this was the first set of roads on Zwift’s New York map in October 2018. (For the shortest possible route, start by picking “The Highline” from Zwift’s route picker, which starts you on the glass roads.)
You’ll hit the NY KOM early, come down to Central Park’s paved roads, head back up to the glass roads to cover a few final bits, then return to ground level to finish out Central Park.
Then it’s time to head south on the new roads launched in October 2025. You’ll travel through the first of three subways, then around Zwift’s reimagined Times Square. Then you’ll enter the brick-lined City Hall Station subway before climbing up and out to cross the East River via the Manhattan Bridge.
After descending from the bridge, you’ll loop around and cross back over the river via the Brooklyn Bridge, before coming across once more on the Manhattan Bridge.
Whew. That’s a lot of bridges. But we’re in Brooklyn now, and almost finished!
Head through our final subway, then up to Grand Army Plaza before entering Prospect Park. You’ll ride a figure 8 of the park, then come out and ride around Grand Army Plaza to cover the last little bit of missed pavement on the Plaza’s roundabout.
At this point, you’ve covered all of NYC’s roads. But we like to end these routes at a defined waypoint, so travel back into Prospect Park and turn left to ride through the KOM banner. That’s the official finish line for the Strava segment.
New York actually already has a “full map” route. Well, almost. Fuhgeddaboudit covers every bit of road on the map… except for Prospect Park’s Central Drive. My hunch is this is actually a bug in the route design, and the intention was to cover every road. They just missed one turn near the end of the route!
Besides the Central Drive problem, Fuhgeddaboudit has another issue: at 79km, it’s much longer than it needs to be if your goal is to ride every bit of road in the shortest ride possible.
Rebel Route Suggestions
Got an idea for a great Rebel Route? Share it below and maybe we’ll publish a post about it!
An anonymous reader quotes a report from the Associated Press: The State Department announced Tuesday it was barring five Europeans it accused of leading efforts to pressure U.S. tech firms to censor or suppress American viewpoints. The Europeans, characterized by Secretary of State Marco Rubio as “radical” activists and “weaponized” nongovernmental organizations, fell afoul of a new visa policy announced in May to restrict the entry of foreigners deemed responsible for censorship of protected speech in the United States. “For far too long, ideologues in Europe have led organized efforts to coerce American platforms to punish American viewpoints they oppose,” Rubio posted on X. “The Trump Administration will no longer tolerate these egregious acts of extraterritorial censorship.”
The five Europeans were identified by Sarah Rogers, the under secretary of state for public diplomacy, in a series of posts on social media. […] The five Europeans named by Rogers are: Imran Ahmed, chief executive of the Centre for Countering Digital Hate; Josephine Ballon and Anna-Lena von Hodenberg, leaders of HateAid, a German organization; Clare Melford, who runs the Global Disinformation Index; and former EU Commissioner Thierry Breton, who was responsible for digital affairs. Rogers in her post on X called Breton, a French business executive and former finance minister, the “mastermind” behind the EU’s Digital Services Act, which imposes a set of strict requirements designed to keep internet users safe online. This includes flagging harmful or illegal content like hate speech. She referred to Breton warning Musk of a possible “amplification of harmful content” by broadcasting his livestream interview with Trump in August 2024 when he was running for president.
For decades, the technology industry has chased a specific dream: the ability to beam ourselves into a room anywhere in the world, not just as a disembodied voice or a flat image, but as a presence that feels real. We are currently at a fascinating, albeit messy, crossroads in that journey. The pandemic forced the world into a massive beta test for remote collaboration, and while we survived, we didn’t exactly thrive. The push to return to the office (RTO) is, in many ways, an indictment of our current tools. We have video, but we don’t have connection.
This brings us to the emerging world of Metaverse conferencing, with platforms like Arthur One Space leading the charge. They are pushing the boundaries of what is possible, yet they highlight exactly how far we still have to go. The picture above showcases this, it shows my meeting on Arthur compared to the resolution needed to make people feel like they are truly in another world (excuse the use of the Lost in Space Robot, I just thought that’d be fun).
From the World’s Fair to the Webcam: A History of Disconnect
The concept of video calling isn’t new; it’s older than the internet itself. The dream was formally introduced to the public at the 1964 New York World’s Fair where AT&T unveiled the Picturephone. It was a marvel of mid-century engineering, promising a Jetson-like future where distance would no longer matter.
However, the Picturephone was a commercial disaster. When it launched in the early 1970s, the service cost a staggering $160 a month (roughly $1,200 in today’s dollars) for a mere 30 minutes of call time. The equipment was bulky, the picture was grainy, and—crucially—nobody else had one. It was a network without a network effect.
For the next forty years, video conferencing remained the domain of high-end corporate boardrooms. Systems from Polycom and Cisco offered “Telepresence,” but they cost tens of thousands of dollars and required dedicated rooms. It wasn’t until the rise of Skype, and later FaceTime, that video calling became democratic. Yet, even as the technology became free and ubiquitous, it remained a secondary mode of communication—useful for grandmothers seeing grandchildren, but rarely the primary engine of business.
The Pandemic Stress Test and the Engagement Void
Then came 2020. Virtually overnight, the entire white-collar workforce was sent home. Platforms like Zoom, Microsoft Teams and Webex became our virtual offices. The technology worked—the bandwidth held up and business continued—but the psychological cracks appeared almost immediately.
“Zoom fatigue” became a recognized phenomenon. We learned that staring at a grid of faces, including our own, for eight hours a day is cognitively exhausting. More importantly, we discovered that video conferencing is excellent for transactions (sharing information) but terrible for interaction (building culture).
This failure is directly linked to the current tension over RTO mandates. CEOs are demanding workers return not necessarily because they hate remote work, but because they sense a degradation in company culture, loyalty and spontaneous innovation. Video calls eliminate the “water cooler” moments; they have no spatial reality. You are either “on” in a meeting or “off” alone in your house. There is no sense of sitting with your team.
Arthur Technologies: Walking the Line of Reality
This is the gap that Metaverse platforms like Arthur One are trying to bridge. Unlike the cartoonish, floating torsos of early Horizon Workrooms, Arthur is aiming for a professional, enterprise-grade realism. They have been one of the most aggressive innovators in trying to make Virtual Reality (VR) feel like a valid workspace rather than a video game.
Arthur Technologies was one of the first to tackle the “floating ghost” problem by introducing legs to avatars, grounding users in the virtual space. They realized early on that to be taken seriously in a boardroom, you cannot look like a Mii character. Arthur’s implementation allows for custom faces generated from a standard 2D photo. You upload a selfie, and the system wraps that texture onto a 3D head.
The platform also includes “Audio Zones,” a brilliant feature that mimics real life. In a standard Zoom call, only one person can speak at a time. In Arthur One, you can turn to the person next to you and have a private side conversation without interrupting the speaker, thanks to spatial audio. The tool also integrates Mixed Reality (MR) features, allowing you to bring your physical keyboard and desk into the VR environment, solving the “blindness” issue of wearing a headset.
The Uncanny Valley and Current Limitations
However, Arthur’s ambition also highlights the current limits of the technology. By pushing for photorealism, Arthur runs headfirst into the Uncanny Valley—that eerie dip in emotional response when a robot or avatar looks almost human but not quite.
While Arthur’s avatars are a massive step up from cartoons, they can still feel stiff. The eyes may not track perfectly, or the micro-expressions we rely on to judge intent (a slight furrow of the brow, a smirk) are missing or delayed. When you look at a cartoon avatar, your brain accepts it as an abstraction. When you look at a photorealistic avatar of your boss that doesn’t blink quite right, your primitive brain flags it as a “corpse” or “imposter.” It is unsettling.
Furthermore, the hardware remains a barrier. Even with lighter headsets like the Meta Quest 3 or the Apple Vision Pro, strapping a computer to your face for an hour-long meeting is physically taxing. The resolution, while improving, still makes reading fine text in a virtual spreadsheet more difficult than on a 4K monitor. Arthur has improved “pass-through” technology to help with this, but we are still limited by the physics of current optics and battery life.
The Road to 2030: What the Future Holds
For platforms like Arthur to truly replace the office, they must evolve past the “good enough” stage. By the end of this decade, we should expect a convergence of technologies that will finally fulfill the promise AT&T made in 1964.
AI-Driven Micro-Expressions: Future avatars won’t just map your face; they will use AI to infer and render micro-expressions in real-time, perhaps using internal headset cameras (eye and face tracking) that are standard in all devices. This will kill the “dead eye” stare and bridge the uncanny valley.
Holographic Form Factors: The headsets will likely slim down to the size of thick glasses, or we may move toward “Light Field” displays that project 3D images without the need for isolation from the real world.
Haptics: We need to feel the handshake. While distant, simple haptic gloves or wristbands could provide the tactile feedback necessary to make a virtual agreement feel binding.
Wrapping Up
We are in the “brick phone” era of the Metaverse. Just as the Gordon Gekko-sized cell phones of the 1980s were laughable but necessary precursors to the iPhone, platforms like Arthur One are the necessary pioneers of immersive work. They are clunky and occasionally creepy, but they offer a glimpse of a future where geography truly doesn’t matter.
Until the technology creates an experience that is better and more comfortable than a physical meeting—rather than just a digital approximation of one—the pressure to return to the office will remain. But if companies like Arthur Technologies continue to iterate, the commute of 2030 might just be putting on a pair of glasses.
The Trump administration has issued travel bans that prohibit five European tech researchers, including one former EU Commissioner, from entering the United States. “For far too long, ideologues in Europe have led organized efforts to coerce American platforms to punish American viewpoints they oppose. The Trump Administration will no longer tolerate these egregious acts of extraterritorial censorship,” said US Secretary of State Marco Rubio.
That official is Thierry Breton, the former European Commissioner for Internal Markets and Digital Services, who Sarah Rogers called the “mastermind” of the Digital Services Act. Rogers, the Under Secretary of State, said Breton threatened Elon Musk about ongoing formal proceedings for X’s noncompliance with “illegal content” and “disinformation” under the DSA just before his meeting with President Trump. The administration has also banned Imran Ahmed from the Center for Countering Digital Hate (CCDH), who apparently called for American anti-vaxxers to be deplatformed. One of those people is Robert F. Kennedy, Jr., who’s now the US Secretary of Health.
Clare Melford from the UK-based Global Disinformation Index has also been banned. Her group monitors online platforms for hate speech. Anna-Lena von Hodenberg and Josephine Ballon from German organization HateAid, have been banned, as well. HateAid flags hate speech online for the EU under DSA rules.
As The New York Times notes, these travel bans emphasize the administration’s close relationship with internet and tech companies, which would benefit from having DSA rules loosened or abolished. The Global Disinformation Index called the travel bans “an authoritarian attack on free speech and an egregious act of government censorship.” Meanwhile, von Hodenberg and Ballon said the bans mark a new escalation. “The US government is clearly questioning European sovereignty,” they said.
This article originally appeared on Engadget at https://www.engadget.com/big-tech/us-bans-former-eu-commissioner-and-others-over-social-media-rules-121804097.html?src=rss
AI coding agents from OpenAI, Anthropic, and Google can now work on software projects for hours at a time, writing complete apps, running tests, and fixing bugs with human supervision. But these tools are not magic and can complicate rather than simplify a software project. Understanding how they work under the hood can help developers know when (and if) to use them, while avoiding common pitfalls.
We’ll start with the basics: At the core of every AI coding agent is a technology called a large language model (LLM), which is a type of neural network trained on vast amounts of text data, including lots of programming code. It’s a pattern-matching machine that uses a prompt to “extract” compressed statistical representations of data it saw during training and provide a plausible continuation of that pattern as an output. In this extraction, an LLM can interpolate across domains and concepts, resulting in some useful logical inferences when done well and confabulation errors when done poorly.
These base models are then further refined through techniques like fine-tuning on curated examples and reinforcement learning from human feedback (RLHF), which shape the model to follow instructions, use tools, and produce more useful outputs.
The next kernel cycle that will be known as either Linux 6.20 or Linux 7.0 depending upon how Linus Torvalds handles the versioning for this next x.20 milestone. More than likely it will be Linux 7.0 given his historical versioning scheme, but whatever the case, ahead of this next kernel cycle some initialization changes for the CXL subsystem are building up…
As the old year is replaced by the new, here’s Jack Wallen’s wish list for Linux and open source in 2026 — from the desktop to the developers who keep it alive.
One of the features being worked on for a while with the read-only EROFS file-system is page cache sharing. Besides EROFS being popular on some mobile/embedded devices, this open-source read-only file-system has been quite popular for container usage and there this page cache sharing functionality can provide some significant reductions in RAM usage…
Longtime Slashdot reader MrFreak shares a public radio interview from 2015 discussing artificial intelligence as inference over abstract inputs, along with scaling limits, automation, and governance models, where for-profit engines are constrained by nonprofit oversight: Recorded months before OpenAI was founded, the conversation treats intelligence as math plus incentives rather than something mystical, touching on architectural bottlenecks, why “reasoning” may not simply emerge from brute force, labor displacement, and institutional design for advanced AI systems. Many of the themes align closely with current debates around large language models and AI governance.
The recording was revisited following recent remarks by Sergey Brin at Stanford, where he acknowledged that despite Google’s early work on Transformers, institutional hesitation and incentive structures limited how aggressively the technology was pursued. The interview provides an earlier, first-principles perspective on how abstraction, scaling, and organizational design might interact once AI systems begin to compound.
