XTHE

Classic PC games are still loved for their stories, characters, and gameplay; however, as technology moves forward, their graphics can look outdated. This makes it harder for players to enjoy them as much as before.  

NVIDIA RTX Remix is a modern modding platform for RTX AI PCs designed to address outdated graphics in classic games. It updates these games with advanced path tracing, a technique that simulates realistic lighting by tracking the path of light rays through scenes. This enhancement lets fans enjoy more detailed visuals and allows new players to discover these improved classics.  

Today, we released a new RTX Remix update through the NVIDIA app. With this update, we are introducing a powerful logic system that makes it easier for modders to trigger interactive graphics effects in games based on many different events.  

Introducing NVIDIA RTX Remix Logic 

Historically, modifying a game’s graphics to respond to real-time game events required access to the game’s source code or its engine.  

The Logic System in Remix enables models to change visuals instantly, using 900 graphics settings triggered by events in more than 165 classic games.  

Without RTX Remix logic, if a modder created a snowy winter day using Remix’s path-traced volumetric system a method that simulates 3D elements like fog, clouds, and snow based on how light travels through them that weather effect would appear in every level, making environments feel less dynamic.  

Depth logic models can now set weather and fog to appear only when the player is outdoors, making changes more specific and realistic.  

Remix logic is flexible and supports over thirty types of game events. These include:  

• the player’s game camera view  

• world boundaries  

• object states  

• time flow  

• which keys the player presses  

Logic uses a simple, no-code, node-based interface to connect triggers to over 900 graphical changes. The system offers limitless options and can be expanded by the community.  

For example, in Half-Life 2: RTX, it can make a door that opens to two different versions of Ravenholm. Each version features unique lighting and weather. Every time the door opens, a new version appears.  

You can also craft magical items that change the world. For example, you can create a science fiction device. It makes the world appear to breathe when you hold it.  

Logic can detect how close the object is to the player and creates this effect by animating the movement of world materials when the device is held. Remix logic is not just for visual effects. It can also help create new gameplay systems in classic games. For example, we tested a night vision system in Half-Life 2 RTX. It turns on when the crossbow zooms in and uses post-processing to color the image green and black. (Credit to modder xoxor4d) for the idea.  

Systems created with Remix logic can even use multiple game events simultaneously. There’s a paranoia system that warns the player of danger when hidden enemies are nearby. The game detects when an enemy is near the player but out of view, triggering post-processing effects such as chromatic aberration and vignetting to simulate a heartbeat.  

Remix logic can even allow models to create bombastic set pieces similar to those in top AAA games. Here is an example we made by combining traditional source modding with remix logic.  

• Path-traced particles that turn increasingly chaotic over time.  

• Lights in the environment start to throb from unstable energy, then a bright flash blinds the player and transforms the skies into a nuclear hellscape.  

• Zombies approach the player and begin to disintegrate from the new radioactive conditions.  

You can download the RTX Remix Logic update from the NVIDIA app. Check the release notes for full details about all the new changes and features.   

The Remix Community Soars to Greater Heights 

In 2025, the Remix Community had a breakthrough year. Over 50 new Remix mods were released, and more than 20 community tools and plugins were made available through open source contributions. These tools help expand compatibility, improve artist workflows, and boost modder productivity.  

Some of our favorite plugins from this year are compatibility wrappers for Unreal Engine 1 and 2 for the utility engine and for Blender and Adobe Substance. Remix Community members also trend on generative AI models like PBR Fusion 3, which create game-ready materials from low-res textures.  

The new Remix Logic update also includes a brand-new look for the Remix Runtime Menu, designed by Community Developer xoxor4d. This open-source contribution expands customization of the Remix Runtime Menu, adding the ability to resize/re-theme and adjust its transparency. If you are a developer looking to work alongside us to make RTX Remix better, be sure to contribute to our official RTX Remix GitHub.  

On top of all the functionality expansions, several impressive mods have been released, including Unreal RTX, MST Watt 401, and Left 4 Dead 2 RTX by the Nostalgia Drive Team. In addition to these releases, we also got sneak peeks at mods in development, like Call of Duty Black Ops RTX by Skurtyskirts and Clive Baker’s Undying @PathtracedParadise on ModDB.  

With these innovations, Remix logic transforms each rendering feature into an interactive sandbox for experimentation and creativity. We’re excited to see what models will make with it.  

To create your own RTX mods, download NVIDIA RTX Remix from the home screen of the NVIDIA app. You can also get the Half-Life 2 RTX Demo and Portal with RTX from Steam. These show what’s possible with RTX Remix.  

If you want to find RTX Remix mods to download and play, check out Mod DB and the RTX Remix showcase Discord server.  

Visit GeForce.com regularly for updates on RTX Remix, new games, drivers, and software.

Source:  Out Now: NVIDIA RTX Remix Update Introduces New Logic System For Dynamic Graphics 

NVIDIA will host GTC, the world’s premier AI and accelerated computing conference, from March 16 to 19 in San Jose, California. Over 30,000 people, including developers, researchers, business leaders, and AI-focused companies from more than 190 countries, will come together to see how AI is becoming a key part of up-to-date infrastructure and driving a new industrial era.  

GTC is at the center of the AI industrial era, said Jensen Huang, founder and CEO of NVIDIA. AI is no longer just one breakthrough or application; it is now essential infrastructure. Every company will use it, and every nation will build it, from energy and chips to infrastructure models and applications. Every part of the stack is moving forward at the same time, and you will watch it all come to life at GTC.  

What to expect at GTC 2026 

Jensen Huang will deliver the keynote at the SAP Center on Monday, March 16, at 11 p.m. PT, outlining NVIDIA’s latest advancements across the AI Stack:  

• Accelerated Computing  

• AI Factories  

• Open Models  

• Agent Tech Systems  

• Physical AI  

The keynote will be live-streamed and available on demand at nvidia.com. No registration is required to watch online.  

GTC will showcase all five AI layers:  

• Energy Chips  

• Infrastructure  

• Models  

• Applications  

through keynotes, sessions, and demos. Attendees will learn how each layer’s unique partners, technologies, and skilled jobs contribute to a major infrastructure expansion.  

GTC will offer over 1,000 sessions featuring topics such as AI factories and workflows, large-scale inference, robotics, digital twins, scientific computing, quantum computing, and enterprise AI developments. These sessions will demonstrate how to build, scale, and optimize every part of the AI stack in practical settings. Jensen Huang will also lead a discussion about the latest advances in open frontier models and industry trends with leaders from organizations such as A16Z/A12 ALP coalition, Black Forest Labs, Cursor, and Reflection AI.  

The GTC Live keynote pre-game show will bring industry leaders together to discuss accelerated computing beyond AI and the underlying 5 layers driving large infrastructure projects. Virtual attendees can watch the pre-game show live online on Monday, March 16th at 8:00 AM PT. Featured speakers will include Perplexity CEO Arvind Srinivasan, LangChain CEO Harrison Chase, Skilled AI CEO Deepak Pathak, OpenEvidence CEO Daniel Nadler, and Mistral AI CEO Arthur Mensch.  

Developers and engineers can gain skills in AI-accelerated computing, networking, data science, and physical AI through nine workshops, over sixty hands-on labs, and on-site certification opportunities. Additional resources include self-paced courses, educator programs, and one-on-one training consultations to grow the workforce for expanding AI infrastructure.  

More than 240 NVIDIA in-sector startups will showcase technologies across physical AI, robotics, generative AI, and enterprise applications. Special startup and venture programs, AI Day with top venture firms, and in-ventures engagement sessions will connect founders and investors working across all parts of the AI industry.  

Developer research and tracks at GTC: developers can learn directly from NVIDIA engineers and partners and plug into the global ecosystem to build the next wave of AI applications. Technical deep dives, C-U-D sessions, and infrastructure workshops will explore large-scale model training in finance optimization and the deployment of AI services across cloud, edge, and software environments.  

Over 150 poster presentations will showcase the latest research from the global AI community, covering model innovation, robotics, systems architecture, and new applications.  

AI across industries and around the world. GTC will turn downtown San Jose into an AI campus spanning 10 venues. Cesar Chavez Park will host a day-and-night market featuring food, entertainment, and live programming. The all-in-podcast will host interviews from the show floor on Wednesday, March 18. Student and Community Day on Thursday, March 19, will open GTC to the wider community at a discounted rate.  

Virtual programming with customized legal regional content is also available through the online session catalog.  

Leaders from organizations across the AI industry, including cloud platforms, model labs, application developers, and robotics companies, will share how AI is boosting productivity and transforming industries.  

Participating organizations span leading companies in software, robotics, automotive research, and AI, including major technology firms, manufacturers, and research institutions.

Source: NVIDIA CEO Jensen Huang and Global Technology Leaders to Showcase Age of AI at GTC 2026 

Intel and Ericsson are partnering to drive AI-enabled innovation in 6G network infrastructure, combining their strengths in computing, connectivity, and cloud from the core to the edge.  

• The partnership aims to make 6G more open, efficient, and affordable for operators and the broader tech community.  

Their collaboration aims to enable a seamless transition to AI-powered 6G networks and solutions, leveraging joint expertise.  

The extended partnership was announced at Mobile World Congress Barcelona 2026, continuing their long-standing partnership. Their joint efforts focus on mobile connectivity, cloud, and computing for AI-driven RAM and core use cases, as well as on-platform security and accelerated cloud-native solutions.  

Lip-Bu Tan, CEO of Intel, says Intel’s ambition is to be the undisputed technology leader in unifying RAN, core, and edge AI to enable an effortless transition to AI-native 6G environments. Together with Ericsson, we will continue to demonstrate that the future of network connectivity is open, power-efficient, secure, and grounded in intelligent AI inference with future Ericsson silicon powered by Intel’s most advanced process nodes, ongoing multi-year research plans, and flexible AI infrastructure. We are well on our way to supplying the future performance efficiency and supply security that the world’s leading operators require.  

Borje Elholm, President and CEO of Ericsson, says 6G is not simply an iteration of mobile technology. It is the infrastructure that will distribute AI across devices, the edge, and the Cloud. Ericsson’s long history of network innovation and large-scale operator deployments enables us to lead practical integration across the value chain and move 6G from research into commercial reality.  

A Joint Commitment 

As 6G moves from research to use, the industry must work together and follow global standards to turn new ideas into infrastructure.  

The partnership will develop high-performance, power-efficient computing for AI-enabled and AI-supported networks.  

AI-Powered 6G will unite flexible networks, advanced computing, and real-time sensing, enabling faster, more efficient, and capable services as sensing and computing become closely linked across the network.  

Showcasing Collaboration Results 

Intel and Ericsson have achieved milestones in Cloud RAN 5G core and open infrastructure at MWG 2026. They will show more demos at the Ericsson Pavilion (Hall 2), Intel’s stand (Hall 3, stand 3E31), and other event spaces, highlighting their teamwork.  

About Ericsson. 

Ericsson’s high-performance networks connect billions of people every day. For 150 years, we’ve been pioneers in developing communication technology. We offer mobile communication and connectivity solutions for service providers and enterprises. Together with our customers and partners, we make the digital world of tomorrow a reality. www.ericsson.com 

About Intel 

Intel is an industry leader creating world-changing technology that enables global progress and enriches lives. Inspired by Moore’s Law, we continuously work to advance the design and manufacturing of semiconductors to help address our customers’ greatest challenges by embedding intelligence in the cloud, at the network edge, and in every kind of computing device. We unleash the potential of data to transform business and society for the better. To learn more about Intel’s innovations, go to newsroom.intel.com and intel.com.

Source: Ericsson and Intel collaborate to accelerate the path to commercial AI-native 6G 

In early 2026, Meta began changing how it moderates content, especially in the United States. The company is moving away from using third-party fact-checkers, a practice in place prior to 2026, and is shifting to a more open, user-driven approach from 2026 onwards. Some reports call this a U-turn: the new strategy lowers restrictions on political speech, relies more on user-moderated community networks, and addresses criticism that earlier models were politically biased.  

The Three-Hour Rule Context (International Versus US) 

Although there are questions about a 3-hour rule in the US, recent reports indicate that this rule is a new regulation in India, introduced in early 2025 and effective on February 20, 2026.  

India’s 3-hour rule: The Ministry of Electronics and IT mandated that social media platforms, including Meta (Facebook/Instagram), remove fake, unlawful, or deepfake AI content within 3 hours (down from 36). Meta execs, including Rob Sherman, say the three-hour deadline is operationally challenging and sometimes unrealistic. They say it does not provide enough time to properly investigate and review flagged content, which could result in the removal of legitimate speech.  

Meta’s 2025-2026 US Political Strategy 

In the United States, beginning in 2025, Meta adopted a different quiet strategy intended to reduce political volatility and appease critics of previous moderation policies. Meta is phasing out traditional third-party fact-checkers and replacing them with a community-node system similar to X’s. The company is loosening its rules, allowing more types of speech. Now it focuses on enforcement, primarily serious violations and illegal content, rather than general misinformation. Threads and other Meta services have said they will not boost or promote political content.  

With these loser rules, Meta says it will still enforce policies against voter interference and serious election-related violence.  

The Quiet Struggle of Real-Time Moderation 

The challenge of moderating content in real time involves acting quickly, as in India’s three-hour rule, while still allowing enough time to avoid mistakes and ensure a fair process. Companies must act so quickly that they rely more on automated AI moderation, which can lead to mistakes, most notably when managing complex local or satirical political comments to avoid legal trouble and maintain their safe harbor protections. Like in India, companies often remove content first and review it later. This approach can sometimes silence legitimate political speech.  

Many have seen this move toward less intervention as Meta’s response to a more free speech-focused, right-leaning political climate in the United States.  

In summary, while India’s three-hour rule represents a concrete regulatory shift, the US scene is defined by Meta’s quieter, voluntary pivot toward user-migrated content and reduced intervention.  

Meta, the owner of Facebook, Instagram, and WhatsApp, announced a major strategy shift today: It is overhauling its content moderation policies to reduce rule complexity and address concerns about over moderation, especially around political and health topics.  

In a blog post titled “More speech, fewer mistakes matter,” new Global Affairs Officer Joel Kaplan explained three main changes to, in his words, undo the mission creep.  

• Meta will end its third-party fact-checking program and switch to a Community Notes model in the coming months. This change will allow users to annotate and contextualize posts, similar to how it works on x.com, giving the community more influence over flagging and clarifying information.  

• The Company will remove restrictions on topics that are part of mainstream discussions, allowing more content on widely discussed issues. Instead, it will focus on enforcing rules against illegal and serious violations, such as:  

Terrorism, child sexual exploitation, drugs, fraud, scams. As a result, users may notice fewer removals of popular or contentious posts, but enforcement will remain strict on the most serious offenses.  

• Users will be encouraged to personalize the political content they see by adjusting their settings. People will have more control over the political opinions and viewpoints appearing in their feeds. This approach enables users to tailor their experience but may also increase the consistency of their information bubbles.  

These changes are important, in part, because a new US presidential administration will take office later this month. Donald Trump and his supporters have said they want free speech to include a much broader range of opinions.  

Facebook has faced criticism from Trump and his supporters over the past few years, especially after the company banned Trump from its platforms as part of its content moderation efforts.  

The United States has strengthened its moderation after facing criticism for spreading misinformation, adding third-party fact-checking in 2016 during the U.S. presidential election.  

This led to the creation of an oversight committee and tools to help users control and report content.  

However, not everyone agrees with these policies. Some critics say the rules are too weak; others think they cause too many mistakes, and some believe the controls are too politically biased.  

Experts, like everyone else, have their own biases and perspectives. This showed up in the choices some made about what to fact-check and how to do it. Kaplan noted that Meta over-enforced its rules, limiting legitimate political debate and censoring too much trivial content. Meta now estimates that 1 to 2 of every 10 censored items were mistakes that didn’t violate policies.  

Some might say these challenges are meant to win over the new administration, but some of these plans have been in progress for a while.  

In the past year, Meta struggled to uphold its own rules. Nick Clegg admitted going too far with moderation, and the oversight board fell short of expectations.  

Now, with political leadership turn-overs, Meta appears to be moving toward a more hands-off approach.  

Meta’s platforms are built to be places where people can express themselves freely. That can be messy on sites where billions of people can have a voice. All the good, bad, and ugly is on display, but that’s free expression, Kaplan wrote.  

The oversight board said it welcomes the news that Meta will revise its approach to fact-checking with the goal of finding an expandable solution to increase trust, free speech, and user voice on its platforms. The board added that it would work with Meta to shape its approach to free speech, signaling a deliberate shift to align with evolving external pressures and internal priorities.  

CEO Mark Zuckerberg has signaled a stronger interest in working with, not battling, the Trump administration. Yesterday, the company appointed three new board members, including UFC head Dana White, a supporter of the incoming president, and last week, Meta replaced its non-time Public Affairs head Nick Clegg, promoting Kaplan to the role. Kaplan had already been part of the policy staff and was known as Meta’s most prominent Republican.  

Meta is also making another change to prevent being stuck in its own echo chamber. Kaplan said, “We will be moving the trust and safety teams that write our current policies and review content out of California to Texas and other U.S. locations.”

Source: Meta drops fact-checking, loosens its content moderation rules 

News Summary 

• AMD expands its Ryzen AI lineup with the new Ryzen AI 400 and Ryzen AI Pro 400 series desktop processors engineered for next-gen AI PC applications and supporting Copilot+ PC features.  

• OEM Partners are launching new AI PCs, including enterprise notebooks and mobile workstations, powered by Ryzen AI Pro 400 Series Mobile Processors.  

• Ryzen AI 400 series mobile processors offer up to 30% faster multi-threaded performance vs. other processors. This helps professionals finish demanding tasks more quickly and still enjoy all-day battery life.  

• AMD provides its widest range of enterprise PC solutions supported by the AMD PRO platform. This platform boasts security, manageability, and resilience for large-scale AI/PC rollouts.  

At Mobile World Congress 2026 in Barcelona, AMD introduced new additions to its Ryzen AI lineup, the Ryzen AI 400 series and Ryzen AI Pro 400 series desktop processors. These processors offer strong on-device acceleration and next-level performance, letting users run AI applications and large language models locally. They are also well-suited for challenging tasks like design and engineering. AMD is also expanding the Ryzen AI 400 series mobile lineup to include workstations.  

These processors enable OEMs to offer next-gen AI PCs from desktops to mobile workstations built for today’s workloads.  

Ryzen AI 400 series processors are the first desktop AI chips supporting Microsoft Copilot Plus PC features. A dedicated NPU delivers up to 50 TOPS of AI computing power, empowering users and, with on-device AI systems, improving privacy and control.  

Desktop PC is evolving from a tool you can use to an intelligent assistant that works alongside you, said Jack Huynh, senior vice president and general manager of the computing and graphics group at AMD. With the Ryzen AI 400 Series processors, the world’s first design to power new Co-Pilot Plus experiences on the desktop, we’re bringing powerful AI acceleration that enables our partners to build systems that empower both enterprises and consumers to do more and create more.  

The World’s First Copilot Plus Desktop Processor For Next-Gen AI Applications 

AMD Ryzen AI 400 Series Desktop processors are built to ensure scalable performance and smart features for professional tasks. They combine high-performance Gen5 CPU cores, AMD RDNA 3.5 graphics, and a dedicated AMD XDNA to NPU. This mix gives office professionals, developers, and power users the speed, efficiency, and local AI acceleration they need for:  

• Multi-tasking  

• Collaborating on software development  

• Data analysis  

• AI-assisted work  

These processors provide unwavering performance and on-device intelligence for today’s desktops.  

Ryzen AI 400 SE AM5 desktop systems with Ryzen AI 400 series processors are expected to be available from OEMs like HP and Lenovo starting in the second quarter of 2026.  

Extending Ryzen Ai Leadership Across Notebooks And Mobile Workstations. 

AMD’s Commercial PC lineup includes more Notebooks powered by Ryzen AI Pro 400 series mobile processors, such as the Ryzen AI 9 HX Pro 470, offering up to 30% faster multi-threaded performance than the Intel Core X7-3581. These devices provide AI acceleration, Co-Pilot Plus PC features, and all-day battery life.  

AI Pro 400 series mobile processors deliver next-gen local AI experiences on commercial notebooks and workstations. With up to 60 TOPs of AI computing power, they enhance business productivity and responsiveness.  

Zen AI Pro 400 series mobile processors will also be used in next-generation workstations, delivering the same performance to professional designs and supporting independent software vendors (ISVs). These mobile workstations are built to speed up professional applications that use all available computing resources – CPU, NPU, and GPU – for demanding engineering, creative, and technical tasks.  

Ryzen AI Pro 400 Series Mobile Workstation Availability 

While workstations with Ryzen AI Pro 400 series processors will be available from OEMs such as Dell Technologies, HP, and Lenovo in the second quarter of 2026, they will empower professionals with advanced AI capabilities in mobile form factors.  

Strengthening Enterprise Security and Manageability with AMD Pro. 

AMD Pro provides enterprise-level security, manageability, and reliability, along with hardware and software built to make IT operations easier and protect investments over time. AMD is improving the Pro platform by making both its hardware and software stronger to help IT teams manage large fleets of AI-enabled PCs. New remote management features give AI administrators better visibility, recovery options, and control so they can fix issues and keep business running without needing to visit each desk. Enabling seamless integration into existing enterprise environments and helping organizations protect their fleets with established security ecosystems.

Source: AMD Gives Consumers and Businesses More AI PC Options with Expanded Ryzen™ AI 400 Series Portfolio 

At MWC2026 in Barcelona, Lenovo is launching a new generation of adaptive AI devices and innovative concepts. The products are aimed at business professionals, creators, students, and gamers. Highlights include:  

• A modular PC  

• A glasses-free 3D Laptop  

• A foldable gaming handheld  

• The debut of Lenovo Qira  

These innovations signal a shift: personal computing is evolving toward adaptive systems that respond to individual needs. This theme of adaptability connects Lenovo’s new hardware and software launches.  

Lenovo is committed to making its technology increasingly personalized, proactive, and secure while creating an ecosystem that works smoothly across all devices.  

The AI era will not be defined by a single device or application, but by intelligent systems that work seamlessly across everything we use, said Luca Rossi, president of Lenovo’s Intelligent Devices Group. We are showing how Lenovo and Motorola are making this real. By merging adaptive hardware with unified AI that works across PCs, smartphones, tablets, wearables, and more. With new devices and the launch of Lenovo Qira, we are building a broad AI portfolio that creates increasingly connected and intuitive experiences for people.  

Lenovo Qira, Personal Ambient Intelligence. 

Lenovo Qira is personal ambient intelligence built into Lenovo and Motorola devices, not just an app added separately. It works across supported PCs, tablets, smartphones, and wearables to help users transition smoothly between tasks and devices, providing guidance based on what they want to do.  

In the coming weeks, Lenovo Qira will roll out across more than 20 devices in Lenovo’s PC portfolio, spanning the Yoga, Ideapad, Legion, and ThinkPad families via a combination of wireless updates and preloads.  

This includes the launch of the IdeaTab Progen 2, Lenovo’s first tablet to feature the experience. First Wave Support on PCs includes six languages across nine regions: English (US, UK, India), Spanish (US, Latin America, Spain), French (France), German (Germany), and Portuguese (Brazil).  

Lenovo Qira will continue to grow, adding support for more devices, features, and partnerships. In 2026, it will support more languages and be available on Motorola smartphones, helping build a unified AI ecosystem for both Lenovo and Motorola.  

Breakthrough Concepts: Exploring Adaptive Form Factors. 

ThinkBook Modular AI PC Concept 

Lenovo is presenting a modular AI PC concept designed for business professionals, emphasizing portability with the idea of “carry small, use big.” This device can be customized with different modules to fit various business needs.  

Concept features a 14-inch ultra-thin base system designed to support interchangeable display configurations, detachable input components, and modular I/O elements. A secondary display can be attached in multiple orientations or can replace the keyboard. By enabling workspace expansion to approximately 19 inches while maintaining portability, the concept illustrates how modular architecture could support evolving workflows and extended device lifecycles in AI-ready environments.  

Yoga Book Pro 3D Concept 

Lenovo is also showcasing a glasses-free 3D laptop concept that allows creators to view and manipulate content with real-time, in-depth detail directly on the display. This technology aims to make design and visual work more immersive for creative professionals.  

By combining dual displays, AI-powered 2D-to-3D conversion, and gesture controls, this concept uses two displays, AI-powered 2D-to-3D conversion, and easy creative tools to make immersive content creation smoother. It also shows Lenovo’s interest in spatial computing and its exploration of adaptable hardware.  

The foldable gaming handheld device can transform from a compact gaming console to a larger display. For desktop use, it supports multiple operating modes, such as:  

• Handheld play  

• Split-screen multitasking  

• Expanded gaming  

• Standard desktop computing  

Demonstrating its flexible design for both gaming and productivity.  

Commercial Portfolio: AI-Ready Platforms for Modern Work 

Shenoy continues evolving its commercial PC portfolio to support organizations adopting AI-enabled workflows at scale.  

The new ThinkPad T series has better serviceability, usability, and AI features. Some models have top iFixit repairability scores. These changes show Lenovo’s emphasis on long-term value, less downtime, and sustainable device management.  

The ThinkPad X13 Detachable extends flexible mobile-first productivity with integrated pen support and field-replaceable components in a lightweight design suited for frontline and hybrid professionals.  

Lenovo is launching the ThinkTab X11, a rugged Android tablet designed for industrial and front-line work.  

For small and medium-sized businesses, the ThinkBook 14 2-in-1 Gen 6 delivers multimodal versatility plus AI-enabled collaboration features, while the ThinkVision M16 portable monitor provides lightweight screen expansion to support flexible work styles.  

Its Commercial Portfolio Lenovo integrates security, manageability, and lifestyle services to help organizations adopt AI-enabled devices with certainty and control.  

The Yoga 9i 2-in-1 Aura Edition (14″ 11) combines a premium convertible design with immersive OLED visuals and intelligent features that streamline creative workflows. The Yoga Pro 7A (15”, 11) and IdeaPad Slim 5i Ultra (14”, 11) expand Lenovo’s AI-enabled laptop portfolio, harmonizing performance and portability for productivity and content creation.  

In tablets, the IdeaTab Pro Gen2 introduces intelligent learning. The IdeaTab Pro Gen2 brings smart learning and productivity features to a flexible tablet and is the first Lenovo Tablet with Qira (15” 11) gaming laptop. The Legion Tab (8.8”, 5) is a top-tier gaming tablet featuring a 3K display and optimized cooling.  

Advanced Smarter AI for all. 

At MWC2026, Lenovo is focusing on making AI accessible to everyone with new hardware. And Lenovo Qira, the company intends to create a future in which technology adapts to people, not the other way around.  

For additional details about all products and initiatives revealed at MWC2026, visit the Lenovo MWC2026 website.

Source: Lenovo Unveils Adaptive AI PCs, Modular Concepts, and Lenovo Qira Rollout at MWC 2026 

Amazon has changed the name of Project Kuiper to Amazon Leo. Amazon Leo has started beta testing with selective business customers in rural Western states, focusing on enterprise users. Plans are to expand to additional business customers in areas such as Illinois. The service is also being developed for residential users, with download speeds targeted at 100-400 Mbps and latency at 30-50 ms. Early tests use production hardware, and there are over 150 satellites in orbit as of late 2025.  

Major Updates and Benchmarks as of March 2026 

• Beta testing is underway with selected business customers focusing on enterprise connectivity in rural western states, with plans to expand.  

• Early reports: Place latency at 30-50 ms, competitive with LEO satellite broadband.  

• Load speeds for residential users are projected to range from 100 to 400 Mbps, with upload speeds from 20 to 100 Mbps. Business customer speeds and requirements may vary by enterprise service tier.  

• A major launch in December 2025 increased the number of Amazon Leo satellites in orbit to over 150. By the end of 2026, the network is expected to have more than 3,200 satellites.  

• Optical inter-satellite links enable fast, direct transfers, improving speed and latency.  

• The terminal or dish is expected to cost less than $400 to produce.  

Amazon Leo aims to compete directly with SpaceX’s Starlink. The service is initially targeting enterprise and business customers, particularly in rural areas, with plans to expand to serve rural residential homes.  

October 2023: Project Kuiper’s Protoflight mission has achieved key milestones, including successful satellite maneuvers and extensive system tests. These results are guiding our constellation management as we expand connectivity services. The final step is de-orbiting Kuiper site one and Kuiper site two, using onboard propulsion and atmospheric drag to lower them until they burn up at about 270 miles above Earth. This process underscores our commitment to space safety and sustainability.  

Optical Mesh Network In Space Will Support Global Connectivity 

Project Kuiper recently revealed the last critical system onboard its satellites with the announcement of an optical intersatellite link (OISL) capability that uses infrared lasers to send data directly between satellites. We’ve completed multiple successful demonstrations of OISL technology during our protoflight mission, maintaining 100 Gbps links over nearly 621 miles (1,000 km) during the test windows. The test results ensure that OISLs will be operational on our first production satellites in 2024. These high-speed laser cross-links will form a secure, resilient mesh network in space, allowing us to increase throughput and reduce latency across our satellite constellation. This provides greater flexibility for connecting Kuiper users across land, sea, air, and space.  

Project Kuiper Confirms 100% Success For Proto-Flight Mission 

Within 30 days of launching two prototype satellites, Project Kuiper reached a 100% success rate for its Protoflight mission. All major systems and sub-systems on both satellites worked as expected or better after launch. Our tests let the team confirm the design of our satellite network and demonstrate 4K video streaming and two-way video calls over it. You can learn more about our mission status and watch our network performance demonstrations.  

A Custom Satellite Propulsion System Allows For Safe Maneuvering In Space 

A series of recent tests demonstrated that Project Kuiper’s Custom Electric Propulsion System, designed and made in-house by Amazon, can safely maneuver ice operands in space. The results mean other satellites can use propulsion to achieve and maintain their assigned orbits, maneuver to avoid world space debris and other spacecraft, and actively de-orbit at the end of their missions. The achievement advances our goal of increasing space safety and sustainability.  

Project Kuiper Satellites Are Healthy 

Project Kuiper engineers have confirmed that our KuiperSat-1 and KuiperSat-2 satellites are fully activated, generating power independently and communicating with our Mission Operations Center.  

Project Kuiper makes first contact with prototype satellites. 

On Oct. 6 at 2:06 p.m. EDT, an Atlas V rocket from United Launch Alliance launched from Space Launch Complex 41 at Cape Canaveral Space Force Station in Florida. It carried two prototype satellites from Amazon’s Project Kuiper, our low Earth orbit LEO satellite broadband project, into space and deployed them at 311 miles above Earth. The launch marked the start of our Protoflight mission. Admiral Washington confirmed the first launch with Kuiper Set 2. This was when the satellite and one of our telemetry, tracking, and control (TT&C) antennas established a telemetry link for the first time. We first contacted Kuiper Set 1 at 2:54 PM EDT.  

First Contact is one of the key steps in our total flight mission. It lets us start downloading data about satellite health and set up regular communications with the satellites.  

Launch today started a new phase of our Prototip Flight mission, and there is a long way to go, but it’s an exciting milestone all the same, said Rajeev Bhatia, Vice President of Technology for Project Kuiper. I am tremendously grateful to the Project Kuiper team and to our partners at United Launch Alliance for their dedication in getting us to this point, and for helping us deploy our first spacecraft ever into orbit.  

On November 13, 2025, Project Kuiper was renamed Amazon Leo.  

On Monday, Amazon announced that its newly rebranded satellite internet service aimed at competing with SpaceX’s Starlink is now available for business trials.  

A select group of businesses will test Amazon Leo’s hardware and software in an enterprise preview before a broader rollout. Amazon said this will help gather feedback and refine solutions for specific industries before full launch.  

Earlier this month, Amazon renamed its satellite internet service from Project Kuiper to Amazon Leo and launched a new website. The name references low Earth orbit, where its satellites will operate within 1,200 miles of Earth.  

Six years ago, Amazon announced plans to build 3,236 low-Earth orbit satellites to deliver fast, low-latency Internet to consumers, businesses, and governments using square-shaped terminals.  

Since April, Amazon has launched over 150 satellites with partners including the United Launch Alliance and SpaceX.  

Amazon’s goal is to compete with Starlink, SpaceX’s market-leading service with nearly 9,000 satellites in orbit.  

Amazon has signed agreements with JetBlue, L3Harris, and Australia’s NBN. It is sending Pro Terminals and Ultra Antennas to businesses in the Enterprise Preview.  

On Monday, Amazon revealed its ultra-model offering up to 1 Gbps download and 400 Mbps upload speeds, powered by a custom silicon chip. Amazon claims it is the fastest commercial phased array antenna available.  

Amazon plans to expand the program to more customers as Leo’s coverage and capacity grow.  

Amazon has not announced consumer pricing or availability.

Source: Amazon starts letting businesses test rebranded Leo satellite service as it plays catch-up to Musk’s Starlink 

The latest updates from Project Kuiper’s satellite test mission

This week, Microsoft announced Sudo for Windows. You can try it out in the Windows 11 Insider program, Preview Build 26052.  

Sudo stands for “super user do”. It’s a command-line tool usually found on Linux systems. On its GitHub page, Microsoft said it did not work on the Linux Sudo project. Instead, Sudo for Windows is a Windows-specific implementation of the Sudo concept. In some cases, it behaves differently from Linux’s Sudo. Some Linux Sudo scripts may not work directly with Sudo for Windows without modification.  

Microsoft built Sudo for Windows mainly to make things easier. It lets users run elevated commands directly from an un-elevated console session, according to the announcement. The announcement also explains how to enable Sudo in Windows 11 preview to build 26052. This is done through the For Developers settings. Users can choose how Sudo for Windows opens: in a new window, in a pop-up, or inline.  

Sudo for Windows Security 

The inline option gives the most familiar interface; however, it may have fewer security controls than the other two options. All three options still elevate the Sudo process to admin privileges using user account control. Microsoft included this warning in its Sudo for Windows documentation.  

Sudo for Windows can be used as a potential escalation-of-privilege vector when enabled in certain configurations. You should be aware of the security considerations when enabling Sudo on your machine.  

Right now, Sudo for Windows is just available on Windows 11 preview to build 26052 and higher; however, the document indicates that it may become available on Windows 10. Microsoft doesn’t have plans to add it to Windows Server. Microsoft is also working to make Sudo for Windows an open-source project on GitHub.  

The Windows 11 Version 24H2 Preview 

This week, Microsoft released Windows 11 preview build 26052 to Windows Insider Program testers in the Canopy and Dev channels. The build is also known as Windows 11 v24h2, which is expected to be the commercial release coming this fall.  

Microsoft now releases major Windows client updates once a year, not twice; however, it still adds new Windows features each month on update Tuesdays. These happen on the second Tuesday of every month and usually include security patches.  

The Windows 11 Build 26052 preview includes some interesting features, but not all of them may make it to the final release. Microsoft is testing automatically opening Co-Pilot when Windows starts on widescreen devices. Co-Pilot in Windows will also launch if you drag an image file onto the Co-Pilot icon in the taskbar.  

The registry editor in this build has an option to limit a search to the currently selected key and its descendants. App developers can access a Power Grid Forecast API to optimize local power grid costs for app users during background tasks. Microsoft is also permitting developers to secure Windows keys with virtualization-based security via a cryptographic API.  

The control panel now includes a new color management screen that lets users set up color profiles for connected displays. Microsoft will also enable users to join Wi-Fi networks by scanning QR codes.  

This build adds more detailed controls for low-energy audio, allowing users to pair Windows audio streaming with supported hearing aids. There is also a new voice clarity feature that shows how apps use different audio processing modes, and this preview build deprecates some features. For instance, the Windows Mixed Reality app has been deprecated, and a Windows Mixed Reality headset will not work starting with this build, although HelloLens, Microsoft’s mixed reality headset, isn’t affected. Microsoft also axed the on-demand use of Microsoft Defender Application Guard as a feature with this build release. Microsoft Edge for Business described Azure Secure Enterprise Browser as its replacement for this feature.

Source: Sudo for Windows Available in Windows 11 Preview 

In late 2025 and early 2026, Google is developing new power grid resilience tools through Tapestry. This major project at Alphabet X is described as a Google Maps for electrons. The goal is to build digital twins of electricity distribution networks, offering operators a clearer way to manage local energy resilience.  

Building on this, Tapestry incorporates AI to create Unified Grid models. These models represent the grid’s physical structure and trace how energy travels across it, similar to how Google Maps displays roadways.  

By illustrating the flow of electricity using familiar mapping techniques, Tapestry provides operators with intuitive visibility into the system.  

In August 2025, Tapestry took its first significant step beyond high-voltage transmission planning, shifting focus to lower-voltage distribution networks. The team partnered with Vector in New Zealand to strengthen local network durability.  

By February 2026, this collaboration had enabled Tapestry’s digital twin systems to make virtual copies of distribution networks. These digital twins enhanced load management, supported improved planning, and delivered up to 20% faster restoration times.  

Forecasting Abilities: The AI tools can rapidly simulate complex scenarios for example, predicting low wind generation during heat waves up to 30 times faster than previous techniques. These tools analyze integrated data from weather systems, energy demand, and distributed resource outputs to optimize real-time grid management and proactively identify potential disruptions.  

Overall, this technology anchors Google’s broader initiative to use AI and mapping tools to make the power grid more reliable, supporting progress toward 24/7 carbon-free energy.  

History, a GoogleX project focused on the electric power grid (the network that delivers electricity from producers to consumers), has mainly partnered with others to bring artificial intelligence (AI tools to transition problems) (issues related to moving electricity over long distances). In Chile, the National Grid Operator uses Tapestry’s tools for yearly transmission planning. North America’s largest grid operator, PJM, a regional transmission organization, is also using Tapestry AI to help manage its large interconnection backlogs (the queue of projects waiting to connect to the grid).  

But as the project’s transmission efforts unfold, Tapestry has also been quietly developing tools for the distribution grid. Today, Latitude Media has learned that Tapestry is unveiling a key milestone in that work: a partnership with a New Zealand distribution service.  

Vector, the largest of New Zealand’s 29 distribution utilities, is now using Tapestry’s grid management and planning tools for daily operations. This marks the first wide use of the technology on a distribution network. The grid-aware Tapestry AI inspection tool has already cut Vector’s average inspection time from 45 minutes to about 5 minutes per asset. This faster, more accurate process gave Vector the insight it needed to use Tapestry’s grid planning tool. That tool helps stimulate future scenarios to plan for resilience and reliability.  

Tapestry’s transmission tools actually grew out of its distribution-focused work, according to Page Crahan, Tapestry’s general manager. When Tapestry started at X, the earliest goal was to address distribution-level challenges first.  

Distribution Grades are less understood, less measured, and less mapped with high confidence than the transmission network, Crahan told Latitude Media. One of the things that is really challenging for network distribution operators is getting a high-confidence representation of their current network from which they can make decisions.  

While these early years saw progress in distribution tools, development lagged; meanwhile, the global energy landscape shifted rapidly.  

More importantly, the conversation about load growth changed. In 2018, people focused on load growth from crypto or electric vehicles, but within a few years, load growth from artificial intelligence and industrial electrification became bigger and more urgent. At the same time, advances in AI and machine learning were also changing. Tapestry’s work greatly improved the team’s capabilities, Crahan said.  

Tapestry looked at that trend, probably a little bit early, and we knew that there was an all-hands-on-deck moment for transmission planning coming immediately. She explained, “When I think about managing resources on our team and where we should focus, it wasn’t about distribution being solved, so we should stop and put our pencils down. It was more about doubling down on things that seemed really urgent at the time.”  

Working Around the Data Problem 

In 2019, when Crahan and her team first met with Vector, the utility was touring innovation hubs around North America. The utility was looking for tools to prepare for future electrification needs in Auckland.  

Shortly after those initial meetings in 2019, the COVID-19 pandemic began, which changed how the teams could work together.  

Things went a little more slowly at the beginning, Crahan said. The upshot of that, she added, was that the Tapestry team really understood the problem before we started building things, because it was the best we could do remotely.  

For example, Tapestry and Vector initially set out to build a distribution planning tool. As they worked together throughout the slower months of the pandemic, they realized that Vector first needed a better understanding of the network’s immediate status before planning for future expansion. That need led to the creation of the grid-aware tool. The automation of inspections and defect detection is great for preventive maintenance, Crahan said, but, more importantly, it provides critical information to drive the planning tool.  

RIDA web enables partners to pull together images of their assets (from utility poles to transformers) from multiple sources, including satellite and street-view imagery, as well as images taken by field teams (in the case of Vector, the utility’s helicopter and drone images). The tool combines those various forms of visual inputs into a single view to simplify inspections.  

Crahan said the most important part of training the model was expert annotation. Experienced Vector field crews labeled features in images to teach the system important details. This human-in-the-loop method, she added, was essential. Ask someone who’s spent their entire career evaluating and maintaining a network that can look in less than 15 seconds and see things that are extremely difficult to train a machine learning model to do.  

The Distribution Impact 

With that foundational data in place, the team then turned to Tapestry’s grid planning tool. It was clear, however, that they could just replicate existing transmission-focused tools. Distribution required its own models, interfaces, and workflows tailored to its unique operations, Crahan explained.  . anna kopf nudes

Transmission and distribution planning do share key steps, she added, like:  

• Preparing future scenarios  

• Running power flow and economic simulations  

• Analyzing system constraints  

But the networks are fundamentally different.  

Transmission planning tools, such as Tapestry in Chile and PJMs, simulate scenarios for high voltage, long-distance power flows. It looks at large-scale expansions, such as load growth from industry or population centers. In contrast, a distribution planning tool must account for many local issues and operational constraints. These are often handled by different tools, which makes building scenarios and models more complicated, Crahan said.  

Now, seven years after Tapestry began deploying its tools on a distribution grid, it highlights the project’s broader strategy with Vector. Crahan said Tapestry is beginning to connect transmission and distribution tools and simulations to create an efficient, singular solution.  

Tapestry considers this deployment proof that AI can help the industry meet energy demand, not just increase it. According to Crahan, a field worker in Auckland may now perform tasks faster and more easily thanks to this machine learning work.

Source: Exclusive: Google’s grid moonshot is now tackling distribution 

In 2013, Caleb Hicks, who later founded School AI, taught nearly 300 students daily. He could connect with the top and bottom performers, but many students in the middle were overlooked. This concerned him. As class sizes grew and budgets shrank, teachers like Hicks found it harder to support every student. 

When ChatGPT launched in 2022, some teachers were concerned about cheating and safety. Still, Ben Hicks, drawing on his instructional design experience at Apple, saw how AI could help create more personalized learning and give teachers better resources. 

Building on that insight, he started School AI in 2023. The platform gives teachers quick updates on student progress and offers students bespoke support. In just two years, School AI has reached over 1 million classrooms in more than 80 countries and formed over 500 education partnerships. It uses OpenAI models to bring advanced technology to teachers and students. 

“We’ve worked hard to ensure AI doesn’t just do the work for you. If AI simply gives students the answer, we’ve missed the mark. Teaching is about coaching and keeping students engaged,” said Nate Sanders, Chief Experience Officer at School AI. 

Lesson 1: Building Trust by Keeping Teachers Involved 

To build trust, School AI’s system mirrors a real classroom. Teachers use DOT, an interactive assistant, to create captivating learning spaces for students. 

A teacher can ask Dot to create a reading activity for students at three different levels. Dot quickly builds a lesson. Teachers can also add interactive apps, letting students create, play, and learn in ways that match the lesson goals. 

Students use Sidekick and AI Tutor, powered by GPT-4.0 and GPT-4.1, to complete these lessons. Sidekick adapts to each student’s learning style, giving guidance, pacing, and encouragement along the way. 

As students work, teachers keep informed. Every School AI interaction is visible, giving teachers early insight into student needs before small gaps grow into larger ones. Built-in safeguards keep School AI safe, transparent, and consistent with classroom goals. 

One student who had just arrived in the US and spoke only Dari used Sidekick for instant translation. Within weeks, he was joining group activities, making friends, and feeling included. Early, confident engagement like this helps students build a base for long-term success. 

Lesson Number 2: Matching Models to Actual Tasks 

For teachers, the main question is not what AI can do, but how it can truly improve learning instead of just providing answers. 

“If AI gives the student the answer, we’ve failed,” says Hicks. “The point of teaching is to coach and keep students engaged in their work.” 

School AI made educator oversight a core feature, rather than relying on simple prompts and responses. Every student input goes through an agent graph with specialized nodes that use models, tools, or safeguards before providing organized support for real learning. 

OpenAI supports every part of this process: 

• GPT-4o drives DOTS Conversational Interface and real-time logic behind lesson construction and response generation.  

• GPT-4.1 handles more complex tasks, such as scaffolding and multi-step math problems. 

• Image generation creates custom visuals, such as photosynthesis diagrams or historical maps, to support lessons. 

• Text-to-Speech (TTS) provides spoken feedback in more than 60 languages. 

The system uses smart routing. Complex tasks go to GPT-4.1 or GPT-4o, while simpler ones are handled by smaller or nano models. This approach keeps costs stable while securing accuracy. 

Up-to-date information and nuance are especially important in schools. Algorithm outputs are visible in logs and shared with teachers in real time. Administrators get a consolidated report. This feedback loop supports School AI’s main idea: AI should coach students, not just give them answers. 

Lesson 3: Using One Tech Stack to Grow Quickly 

At School AI’s latest product showcase, over 10,000 educators attended. Just before the event, the team realized they were still limited by consumer-level restrictions. 

“We reached out to our OpenAI contact to see what we could do,” said Sanders, Chief Experience Officer at School AI. Within 10 minutes, they upgraded our storage and increased our limits as we started using GPT-4.1, assuring a smooth event as new models became more affordable. School AI reduced its costs from nearly $1 per student to just a fraction of that. This change allows the team to invest in the future and grow wisely, which is especially important in education, where budgets are tight. 

“We chose OpenAI because their models present unmatched accuracy, nuance, and flexibility. Hicks says we decided to scale with them because the support, we’ve received is second to none.” 

Gazing Forward: Molding the Future of Education 

For teachers, AI can be a valuable partner. It gives them more time for the important human side of teaching. Some teachers say School AI saves them over 10 hours a week. More importantly, they use that time to step in earlier, offer faster support, and spend more meaningful one-on-one time with students. said she used to depend on test scores to spot students who were struggling, but with school AI, she noticed a student who had stopped asking questions and joining discussions. That small sign led to a check-in and early help that might have been missed before.  

Student behavior is changing, too. Engagement is rising in lessons that use AI, and Sidekick is helping students become more confident and independent. The same Dari-speaking student who once needed real-time translation is now joining group work, joking with classmates, and feeling more confident. 

As more schools start using School AI, leaders are using live data to see what works and where extra support is needed, including new features for learning at home. School AI now connects students, teachers, and families through one trusted system. 

This mission has always been about helping every student feel noticed. Since HICS with OpenAI, we have been able to deliver on that promise consistently at the system-level schools’ needs.

Source:    SchoolAI’s lessons in building an AI platform that empowers teachers