Table of Contents

Meta has officially introduced its latest large language models Llama 4 Scout and Llama 4 Maverick which it describes as its most advanced multimodal AI systems to date. These models are capable of understanding and integrating different types of inputs such as text, images, video, and audio. According to Meta, both models excel in handling this kind of complex, cross-format content, offering improved performance across various tasks.

Importantly, both Scout and Maverick are being released as open-source, making them accessible to developers and researchers around the world. This move signals Meta’s continued support for collaborative AI development, especially at a time when big tech firms are competing to lead the AI race.

Preview of Llama 4 Behemoth and Meta’s AI Ambitions

Alongside the official release, Meta also gave a preview of Llama 4 Behemoth, a much more powerful model designed to train and guide future generations of AI systems. Though earlier reports suggested delays due to performance concerns in math and reasoning tasks, Meta appears to be pushing forward with confidence. Behemoth is being positioned as a kind of ‘teacher’ model for improving the next iterations of AI.

In line with this strategy, Meta has committed to investing as much as $65 billion in AI infrastructure this year, reflecting the intense pressure among major tech companies to show real returns on their AI investments. Also check out Dailly blog page here

Meta launched its most advanced open-source AI models yet, Llama 4 Scout and Maverick, and previewed Llama 4 Behemoth. These multimodal systems are part of Meta’s major push into AI, supported by a $65 billion investment in infrastructure.

Growing National Security Concerns

The ban comes amid rising concerns from U.S. officials and lawmakers over DeepSeek’s potential threat to sensitive government information. The low-cost AI model, developed in China, triggered a major selloff in global markets in January, as investors worried about its impact on the United States’ AI dominance.

In February, Congressmen Josh Gottheimer and Darin LaHood, members of the House Permanent Select Committee on Intelligence, introduced legislation to ban DeepSeek from government devices. Earlier this month, they also urged U.S. governors to ban the AI model from state government-issued devices, warning of potential data exposure to the Chinese Communist Party (CCP).

Statewide Bans and Legislative Push

Several U.S. states, including Virginia, Texas, and New York, have already banned DeepSeek on government devices. Additionally, a coalition of 21 state attorneys general has called on Congress to take further legislative action.

In a March 3 letter, lawmakers expressed concerns about DeepSeek’s ability to collect sensitive data, stating:
“By using DeepSeek, users are unknowingly sharing highly sensitive, proprietary information with the CCP such as contracts, documents, and financial records. In the wrong hands, this data is an enormous asset to the CCP, a known foreign adversary.”

Future Implications

While the Commerce Department’s ban is now in effect, it remains unclear how extensively the restriction applies across the entire U.S. government. With pressure mounting from lawmakers and state officials, the push to prohibit DeepSeek at the federal level may soon gain momentum, further escalating U.S.-China tensions in the AI sector.

Table of Contents

Google has accused Meta and other social media companies of pushing legislative proposals that shift the burden of child safety onto app stores rather than addressing platform-specific risks. In a blog post published on March 12, Google argued that such laws introduce privacy risks for minors while failing to tackle the root causes of online harm.

Utah’s Age-Verification Law Sparks Debate

The controversy follows Utah’s recent legislation, which makes it the first US state to require app stores like Google Play and Apple App Store to verify users’ ages and obtain parental consent for minors before allowing app downloads. While Meta, Snap, and X praised Utah’s new law, Google has criticized it as “concerning”, claiming it shifts responsibility away from social media platforms and places unnecessary burdens on app store operators.

Google’s Criticism and Alternative Proposal

Google’s Public Policy Director, Kareem Ghanem, argued that age verification should not be the sole responsibility of app stores. Instead, Google proposed an alternative legislative framework where only certain app developers those offering potentially risky content would be required to implement age verification measures. This approach, Google claims, would prevent excessive data collection from children and their parents while still ensuring appropriate safeguards.

Privacy Concerns Raised by Apple

Apple has also expressed concerns about excessive data collection, highlighting that many US children do not have government-issued IDs. This means parents may have to submit additional sensitive documents just to allow their children to use basic apps. Apple argues that this poses a greater privacy risk rather than enhancing child safety. Check Latest Apple news here

Meta’s Response to Google’s Claims

Meta spokesperson Jamie Radice welcomed Google’s willingness to share age information with developers but questioned how Google would determine which apps should receive such data. She reaffirmed Meta’s stance that parents should have the final say over children’s app access. According to Meta, app stores should be required to obtain parental consent before allowing minors to download apps, ensuring that families, not companies, make these decisions.

The Challenge of Age Verification

The issue of age verification remains complex, with no universal solution in place. Utah’s law does not specify a clear process for ensuring accuracy, only requiring app store operators to use “commercially available methods that are reasonably designed to ensure accuracy.” As more US states consider similar laws, the debate over whether social media companies or app stores should be responsible for child safety continues to in…

Comment down what you think ? Also dont forget to visit our PC Build Store

While Lenovo has not revealed detailed technical specifications, the AI Stick is capable of delivering up to 32 TOPS (Tera Operations Per Second) , a metric commonly used to measure AI performance. Lenovo also mentioned that the AI Stick will perform better when connected via its dedicated power adapter, though it remains unclear how much performance will be impacted when plugged into a standard USB-C Thunderbolt port.

A unique feature of the AI Stick is its ability to plug directly into a wall socket, providing maximum performance for demanding AI tasks. This makes it an ideal solution for users with older processors or slower NPUs, ensuring that they can still benefit from modern AI features without upgrading their entire system.

Alongside the AI Stick, Lenovo also introduced the AI Display, a curved monitor with a built-in discrete NPU. Similar to the AI Stick, the AI Display allows users with older laptops and PCs to access AI-powered enhancements without needing a high-end, AI-ready computer.

While still in development, Lenovo’s AI Stick and AI Display highlight the company’s commitment to making AI accessible, allowing users to upgrade their devices with AI capabilities without investing in entirely new hardware.

Powered by Intel’s fastest Arrow Lake-HX CPU, the Core Ultra 9 275HX, the Blade 18 is paired with up to an Nvidia GeForce RTX 5090 (24GB VRAM) What else do you expect?. Other GPU options include the RTX 5080 (16GB) and RTX 5070 Ti (12GB). To support its high-performance components, the laptop delivers a total power allocation of 280W, requiring significant cooling upgrades. As a result, the Blade 18 is 47% thicker and 50% heavier than the Blade 16, measuring 27.94 mm thick and weighing 3.1 kg (7.06 lbs).

The 18-inch display is a standout feature, offering a dual-mode refresh rate, allowing gamers to switch between 4K at 120Hz for high-resolution gaming and 1080p at 440Hz for ultra-fast competitive gameplay. The panel boasts a 3ms response time and 100% DCI-P3 color gamut coverage, ensuring high accuracy and vibrant visuals.

Connectivity options on the Blade 18 are extensive, including three USB 3.2 Gen 2 Type-A ports, a Thunderbolt 5 port, a Thunderbolt 4 port, a 2.5Gb Ethernet port, an HDMI 2.1 port, and an SD card reader. The laptop also features a 5MP webcam with Windows Hello support for secure facial recognition login.

Powering the system is a 99Wh battery, supported by a massive 400W power brick. The top-tier configuration, featuring the Core Ultra 9 275HX, RTX 5090, 32GB DDR5-5600 RAM, and a 2TB NVMe SSD, is priced at $4,499 (₹3.73 lakh). Pre-orders are currently live on Razer’s official website, making this a top choice for those seeking desktop-level performance in a laptop form factor.

YouTube channel BuildZoid and a viral Reddit post by an alleged Intel engineer argue that Nvidia’s power delivery system is flawed. BuildZoid points to the removal of shunt resistors, which are crucial for monitoring power flow and preventing overloads. Older GPUs in the 30-series and earlier generations used multiple shunt resistors to balance power draw, but with the RTX 50-series, Nvidia has reportedly reduced their number, making it impossible for the card to detect cable failures. If wires break, the GPU cannot regulate power properly, potentially causing dangerous overheating or even fire hazards.

Adding to the concerns, the 5090 Founder’s Edition is said to use only one shunt resistor, increasing the risk further. Some third-party GPU manufacturers, like ASUS ROG Matrix 4090 and Astral 5090, have added extra safety mechanisms, but these do not fully eliminate the risks posed by improper power regulation.

A Reddit user claiming to be an Intel electrical engineer further explained that 12VHPWR connections operate at their absolute amperage limit, unlike 8-pin and 6-pin connectors, which have built-in safety margins. The micro-fit 12VHPWR connectors reportedly carry 600W at dangerously close tolerances, leaving no room for failure, which is why melting incidents are still occurring despite Nvidia’s connector redesign.

Nvidia has not issued an official statement on why these failures are happening. Initially, the company blamed user error for the melting cables seen in the RTX 4090, but it later introduced the 12V-2×6 revision to improve pin length and connection stability. While some suspect that inferior metals used in third-party cables might be part of the issue, the lack of proper voltage regulation on Nvidia’s own cards remains a major concern.

For now, potential RTX 5080 and 5090 buyers may want to proceed with caution, considering alternative power connectors where available or waiting to see if further hardware issues emerge in the coming weeks. Given the limited stock and high demand, many interested buyers are already on hold, making it an ideal time to monitor how widespread these failures become before making a purchase decision.

Designed for gamers, AI professionals, and content creators, the 9100 PRO offers sequential read speeds of up to 14,800 MB/s and write speeds of up to 13,400 MB/s , almost twice as fast as the 990 PRO. Additionally, random read/write speeds reach up to 2,200K and 2,600K IOPS, ensuring smooth multitasking and ultra-fast data processing.

To maintain high-speed performance, Samsung has incorporated an advanced heat management system. The 1TB to 4TB models come with an 8.8mm thick heatsink, while the 8TB variant features an 11.25mm thick heatsink, enhancing power efficiency by 49% compared to previous models. This ensures reduced energy consumption without compromising performance.

The 9100 PRO also marks Samsung’s first-ever 8TB NVMe SSD, catering to professionals who require massive storage. It is compatible with laptops, desktops, and gaming consoles, making system upgrades easier. Additionally, Samsung’s Magician software offers tools for optimization, data migration, and advanced security.

Samsung 9100 PRO SSD Price in India & Availability

Samsung will launch the 9100 PRO SSD in four storage options:

• 1TB – ₹14,999
• 2TB – ₹25,499
• 4TB – ₹49,999

These variants will be available from March 18, 2025. The 8TB model is expected to be released in the second half of 2025. With its breakthrough speed, efficient cooling, and expanded storage, the 9100 PRO SSD sets a new benchmark for high-performance storage solutions in India.

Panos Panay, SVP of Devices & Services at Amazon, emphasized the significance of this upgrade, stating, “With 600 million Alexa devices now out in the world, the latest advancements in generative AI have unlocked new possibilities enabling us to reimagine the experience in our pursuit of making customers’ lives better and easier every day.”

Pricing & Availability

Alexa+ will not be free for regular Amazon users, but Prime members will get access at no additional cost. Non-Prime users will have to pay $19.99 (₹1,660) per month. This pricing strategy appears to be Amazon’s way of incentivizing more users to subscribe to Amazon Prime, which costs $14.99 (₹1,250) per month or $139 (₹11,500) annually in the US.

The AI-powered assistant will begin rolling out in the US in the coming weeks as part of an early access period, with a wider release planned over the next few months. Initially, Amazon will prioritize Echo Show 8, 10, 15, and 21 device owners, but a global rollout plan has yet to be announced.

Expanded Access Across Devices & Platforms

Alexa+ will be accessible on all Echo devices, along with a new web-based platform (Alexa.com) and a mobile app available on the Apple App Store and Google Play Store. This expansion ensures that users can access Alexa+ from virtually anywhere, even without an Echo device.

New AI Capabilities & Smarter Interactions

Powered by advanced AI, Alexa+ can understand natural, informal speech, even when users speak in half-formed thoughts, slang, or complex ideas. It can maintain context between conversations, making interactions more fluid and natural. Users can also ask Alexa+ to remember specific details, allowing for a more personalized experience.

Beyond conversations, Alexa+ can control smart home devices, play music, order groceries, send reminders for ticket sales, and even use Ring doorbells for security alerts. Its enhanced agentic powers enable it to autonomously complete tasks on behalf of users. For example, Alexa+ can search the web, find an air conditioner repair service via Thumbtack, authenticate, schedule an appointment, and notify the user once it’s done all without manual input.

With Alexa+, Amazon is aiming to revolutionize the voice assistant experience, offering a smarter, more proactive AI that can anticipate user needs and perform complex tasks seamlessly. As AI-powered assistants become more sophisticated and deeply integrated into daily life, Alexa+ positions itself as a powerful competitor in the evolving AI landscape.

The Tecno Spark Slim will feature a 6.78-inch 3D Curved AMOLED display with a 144Hz refresh rate and 4,500 nits peak brightness, ensuring a vibrant and smooth visual experience. The phone’s design also emphasizes sustainability, as it is built using recycled aluminum and features a stainless steel unibody for durability.

In terms of camera capabilities, the Spark Slim is equipped with a 50MP main camera, a 50MP ultra-wide lens, and a 13MP front camera for selfies. Additionally, the phone incorporates an interactive light band, creating a dynamic and humanized lighting effect.

Despite its ultra-thin design, Tecno has managed to pack a 5,200mAh battery, which is supported by 45W fast charging. Powering the device is an octa-core processor, though Tecno has yet to reveal the specific chipset model.

With its combination of sleek design, high-performance display, advanced camera system, and powerful battery, the Tecno Spark Slim is set to make an impact at MWC 2025, further pushing the boundaries of smartphone innovation.

News of Skype’s shutdown has sparked nostalgia and emotional reactions across social media, particularly on X (formerly Twitter), where users reminisced about its impact. Discord, a platform widely used by gamers, acknowledged Skype’s influence by posting, “Thank you for walking so we could run. RIP to the OG gaming chat app.” Another user reflected on Skype’s decline, writing, “Skype had a 17-year head start and cracked under pressure in 2020. Its fumble during the pandemic will be studied for centuries.” Many shared personal experiences, with one user highlighting Skype’s role in keeping international students connected, saying, “Skype was a savior, making it possible to call home for free when calling cards used to cost $20-25 for just 15 minutes.” Others humorously noted that even NBA star LeBron James has outlasted Skype in relevance.

Once a dominant force in online communication, Skype reached its peak in 2010 with 660 million users. However, competition from platforms like Zoom, Google Meet, FaceTime, WhatsApp, and Facebook Messenger gradually diminished its relevance. By 2015, its active users had dropped to 300 million, and by March 2020, at the height of the pandemic, the number had declined further to just 100 million, as video conferencing alternatives gained popularity. Microsoft, which acquired Skype for $8.5 billion in 2011, now sees little reason to continue the service, especially with Microsoft Teams offering a more integrated experience for both professional and personal use.

Microsoft’s Jeff Teper, president of 365 collaborative apps and platforms, assured users that they will not be forced into Teams if they prefer to move their data elsewhere. However, the company is positioning Teams as a natural successor, emphasizing its seamless transition process for Skype users. With Skype’s discontinuation, an era of internet-based communication history comes to an end, as Microsoft shifts its focus toward a more modern and business-friendly messaging ecosystem.