Category: 4. Technology

Meta Ray-Bans have often been very difficult to find in stock at both retailers and online stores for the past two years — and they rarely go on sale. But one of the best ways to not only get a 20% discount but also find some of the styles that have been discontinued or are rarely ever in stock is to use the little-known Meta Refurbished AI Glasses online store. 

I recently ordered a pair of Meta Ray-Bans in the Skylar style with the Shiny Caramel finish with Polar Brown lenses. If you’ve followed Meta Ray-Bans, then you’ve probably seen these before because they are in a lot of the Meta promotional materials. The frames themselves are semi-transparent, so you can see a lot of the technology inside. And the lenses have that amber tint that’s popular in lots of sunglasses because it gives the world a bit of warm orange glow. 

I was able to buy these for $263, even though they normally retail for $329. Other configurations, like the black with blue transition lenses, were as much as $76 off. As of the writing of this article, Meta’s refurbished store was selling the Wayfarer, Headliner, and Skyler styles. The Wayfarer is the most common style, while the Headliner and Skyler can be more difficult to find.

When my transparent brown Meta Ray-Bans arrived, they were in the original box and it was unopened, with the perforated pull tag still in perfect condition — no different than the box of the original pair of black Meta Ray-Bans I bought when the product first launched two years ago.

When I opened the box, everything was in perfect order, just like a new pair. All the accessories, booklets, polishing cloth, and even the paper protectors over the arms of the glasses were all just as they are in the packaging of a brand new pair of Meta Ray-Bans that you’d buy at retail or online. 

I pulled the little protective tab inside the case to activate charging and plugged the case into USB. After about 30 minutes, I opened up the Meta AI app, tapped the glasses button on the far right, hit the plus sign in the upper right corner, tapped on the Ray-Ban | Meta button, and the software immediately recognized that there was a pair of glasses in a case nearby. I accepted the prompt and tapped through the menus. 

I pulled the glasses out of the case and examined them, and they were in perfect condition. I put them on and immediately started using them like the other pairs of Meta Ray-Bans I’ve been testing and using for the past couple years. 

What does refurbished actually mean for a tech product?

So if these are just like new and sold at a discount, then what does it mean when you buy a refurbished tech product? These are typically products that have been lightly used and a customer returned or traded in for another product, or that originally had small defects. They go back to the manufacturer for inspection, cleaning, repair (if needed), and another round of quality assurance. They are then repacked, often in the same original packaging as new products. However, both the US and the European Union have regulations that require refurbished products to be marked and sold as refurbished, and can’t be passed off as new products. That’s why they are typically discounted, even though in many cases they are as good as new. 

In the case of refurbished Meta Ray-Bans, when you buy through the Meta Refurbished AI Glasses online store, then Meta offers a 30-day return policy and a 1-year warranty. Based on my experience, I have no problem recommending that you buy a refurbished pair of these smart glasses, if one of the styles you like is available.

Buying advice

Keep in mind that Meta is likely to unveil its next-generation Meta Ray-Bans Smart Glasses on September 17 at its Meta Connect event. These could have new features like longer battery life and better cameras — two of the items on my wish list. It’s unclear if they will be more expensive, but it appears possible if they are integrating better technology and there’s also the uncertain US tariff situation that could impact the cost of parts. 

The recently released Oakley Meta Smart Glasses have some nice tech upgrades and retail for $400-$500. That could be an indication of where the prices of the new Meta Ray-Bans are going to land. If that’s the case, the deal on these refurbished Meta Ray-Bans could look even better.

Sports Interactive

FM26

Gwen Archer

Charles Thorp is the Fitness and Commerce Editor at Men’s Health, where he shares the best product recommendations in gym equipment, recovery tools, supplements, and more. Following an early life in athletics, Charles became a NASM-certified trainer and began writing programs alongside the most respected coaches in the world.  

Since entering the world of fitness content, Charles has had the opportunity to learn from and train alongside high performance individuals from the NFL, UFC, NBA, Formula 1, CrossFit, US Olympics, and Navy SEALs. When he’s not writing about training programs or gear, he can be seen at the gym or in the wild, putting them to the test.

To make such a network possible, so-called quantum network nodes — that can store quantum information and share it via light particles – are needed. In their latest work, the Innsbruck team led by Ben Lanyon at the Department of Experimental Physics of the University of Innsbruck demonstrated such a node using a string of ten calcium ions in a prototype quantum computer. By carefully adjusting electric fields, the ions were moved one by one into an optical cavity. There, a finely tuned laser pulse triggered the emission of a single photon whose polarization was entangled with the ion’s state.

The process created a stream of photons; each tied to a different ion-qubit in the register. In future the photons could travel to distant nodes and be used to establish entanglement between separate quantum devices. The researchers achieved an average ion-photon entanglement fidelity of 92 percent, a level of precision that underscores the robustness of their method.

“One of the key strengths of this technique is its scalability,” says Ben Lanyon. “While earlier experiments managed to link only two or three ion-qubits to individual photons, the Innsbruck setup can be extended to much larger registers, potentially containing hundreds of ions and more.” This paves the way for connecting entire quantum processors across laboratories or even continents.

“Our method is a step towards building larger and more complex quantum networks,” says Marco Canteri, the first author of the study. “It brings us closer to practical applications such as quantum-secure communication, distributed quantum computing and large-scale distributed quantum sensing.”

Beyond networking, the technology could also advance optical atomic clocks, which keep time so precisely that they would lose less than a second over the age of the universe. Such clocks could be linked via quantum networks to form a worldwide timekeeping system of unmatched accuracy.

The work, now published in Physical Review Letters, was financially supported by the Austrian Science Fund FWF and the European Union, among others, and demonstrates not only a technical milestone but also a key building block for the next generation of quantum technologies.

Google has issued a security alert to its 2.5 billion Gmail users, urging them to update their passwords and enhance account protection following an increase in “successful intrusions” by password hackers.

The tech giant is encouraging users to be proactive and remain vigilant for suspicious account activity, and to install additional safety measures—such as two-factor authentication—if they haven’t already done so.

Hackers often obtain Gmail credentials by tricking users through phishing emails that lead to fake login pages or prompt them to disclose sensitive information like two-factor authentication codes.

One group, known as ShinyHunters—allegedly named after the Pokémon franchise—has been active since 2020 and linked to several high-profile breaches involving AT&T, Microsoft, Santander and Ticketmaster.

Despite growing awareness around online security, Google reports that while many users already have strong and unique passwords, only about one-third update them regularly—leaving a significant window of vulnerability.

In an advisory, Google also addressed a separate breach involving its Salesforce database, where attackers used social engineering tactics. In these cases, cybercriminals posed as IT support—a method that proved highly “effective in tricking employees,” according to the company.

While the compromised data in this incident was largely publicly available, Google warned that the tactics used could evolve into more serious and targeted attacks in the future.

“We believe threat actors using the ‘ShinyHunters’ brand may be preparing to escalate their extortion tactics by launching a data leak site (DLS),” Google stated in a blog post published in June.

Google also warned, “These new tactics are likely intended to increase pressure on victims, including those associated with the recent UNC6040 Salesforce-related data breaches.”

All users potentially affected by the incident were notified by email on Aug. 8.

As digital threats continue to evolve, Google is strongly encouraging its global user base to take proactive steps to secure their accounts.

Google advises users to not only use complex passwords and enabling two-factor authentication, but to also remain alert for phishing attempts, unsolicited login prompts, and suspicious activity that may indicate a breach attempt.

Tactile sensors are widely used in robotics, prosthetics, wearable devices, and healthcare monitoring. These devices detect and convert external stimuli such as pressure and force into electrical signals, facilitating effective environmental detection. Scientists have made extensive efforts to improve the performance of tactile sensors in terms of sensing range and sensitivity. In this context, mechanical metamaterials are highly promising. Specifically, auxetic mechanical metamaterials (AMMs)-possessing a negative Poisson’s ratio-enable inward contraction and localized strain concentration upon compression. These counterintuitive behaviors render them lucrative options for designing sensors and actuators with excellent properties.

However, existing AMM technology suffers from fabrication and integration challenges.

Addressing this knowledge gap, a team of researchers from the Seoul National University of Science and Technology, led by Mr. Mingyu Kang, the first author of the study and a Master’s course student in the Department of Mechanical Design and Robot Engineering, and including Dr. Soonjae Pyo, an Associate Professor in the Department of Mechanical System Design Engineering, have proposed a novel 3D AMM-based tactile sensing platform based on a cubic lattice with spherical voids and fabricated using digital light processing-based 3D printing. Their breakthrough findings were made available online and published in the journal Advanced Functional Materials on July 6, 2025.

The researchers explored the tactile sensing platform, utilizing 3D-printed auxetic metamaterials in both capacitive and piezoresistive sensing modes. While the sensor responds to pressure via electrode spacing and dielectric distribution modulation in the first mode, the latter mode leverages a conformally coated network of carbon nanotubes that alters resistance under load.

“The unique negative Poisson’s ratio behavior utilized by our technology induces inward contraction under compression, concentrating strain in the sensing region and enhancing sensitivity. Beyond this fundamental mechanism, our auxetic design further strengthens sensor performance in three critical aspects: sensitivity enhancement through localized strain concentration, exceptional performance stability when embedded within confined structures, and crosstalk minimization between adjacent sensing units. Unlike conventional porous structures, this design minimizes lateral expansion, improving wearability and reducing interference when integrated into devices such as smart insoles or robotic grippers. Furthermore, the use of digital light processing-based 3D printing enables precise structural programming of sensor performance, allowing geometry-based customization without changing the base material,” remarks Mr. Kang.

The team showcased two proof-of-concept scenarios highlighting the novelty of their work: a tactile array for spatial pressure mapping and object classification, as well as a wearable insole system with gait pattern monitoring and pronation type detection capabilities.

According to Dr. Pyo: “The proposed sensor platform can be integrated into smart insoles for gait monitoring and pronation analysis, robotic hands for precise object manipulation, and wearable health monitoring systems that require comfortable sensing without disrupting daily life. Importantly, the auxetic structure preserves its sensitivity and stability even when confined within rigid housings, such as insole layers, where conventional porous lattices typically lose performance. Its scalability and compatibility with various transduction modes also make it suitable for pressure mapping surfaces, rehabilitation devices, and human-robot interaction interfaces that require high sensitivity and mechanical robustness.”

In the next decade, auxetic-structured 3D-printed tactile sensors could form the backbone of next-generation wearable electronics, enabling continuous, high-fidelity monitoring of human movement, posture, and health metrics. Their structural adaptability and material independence could drive the creation of custom-fit, application-specific sensors for personalized medicine, advanced prosthetics, and immersive haptic feedback systems.

As additive manufacturing becomes more accessible, mass-customized tactile interfaces with programmable performance may become standard in consumer products, healthcare, and robotics.

Valve has started to comply with the UK’s Online Safety Act, by rolling out a requirement for all Brits to verify their age with a credit card to access “mature content” pages and games on Steam. UK users won’t even be able to access the community hubs of mature content games unless a valid credit card is stored on a Steam account.

While platforms like Reddit, Bluesky, and Discord have opted for age verification checks using selfies, Valve is restricting its age checks to just credit cards, according to a support article. “Among all age assurance mechanisms reviewed by Valve, this process preserves the maximum degree of user privacy,” says Valve. “Having the credit card stored as a payment method acts as an additional deterrent against circumventing age verification by sharing a single Steam user account among multiple persons.”

In the UK you need to be 18 years of age to obtain a credit card, so this passes the age checks onto banks instead of Valve having to perform them. If you don’t have a credit card then you won’t be able to access mature content games or pages on Steam, as there’s no other way of verifying your age.

Microsoft has also started rolling out Xbox age verification in the UK. While the age verification checks are optional right now, they will become a requirement to access a variety of Xbox services in early 2026, when additional parts of the UK’s Online Safety Act come into force.

Forget Windows 95, it’s 30 years since Doom was released on the Super Nintendo Entertainment System. And thanks to the Raspberry Pi RP2350 microcontroller, the game is back in cartridge form.

Doom on the SNES was a programming tour de force that arguably kicked off the “But can it run Doom?” meme. The first-person shooter, with its atmospheric soundtrack and 3D graphics, had taken the world by storm after its release in 1993. It was ported to many platforms, but squeezing it onto an SNES cartridge was a challenge.

The port was accomplished thanks to a Super FX chip, a graphics acceleration coprocessor developed by Argonaut and Nintendo, in the game cartridge, but the experience was not as fluid on SNES as on high-end PCs of the time and some content was missing.

Thirty years on, Limited Run Games (LRG) is having another crack at the old warhorse, this time with updated cartridge hardware, while still ensuring it will work on the vast majority of SNES consoles.

According to Randal Linden, who also worked on the original port in the 1990s, the update replaces the Super FX chip with an inexpensive RP2350. As well as simulating the Super FX chip, the RP2350 also performs “graphics format conversion operations” and provides a high-speed interface to multiple memory devices.

It’s important to note that RP2350 is not running the game itself, although Doom has been ported to the hardware. Rather, it is simulating and emulating the Super FX chip. The setup has allowed the developers to further optimize the code and reach a heady 20 frames per second.

Speaking to John Linneman of The Digital Foundry, Linden explained how it works:

Controls can be via a standard controller or mouse, and there’s also rumble force feedback. Missing levels have been added, as well as Episode 4.

As for why performance is capped at 20 frames per second, Linden explained: “A single screen’s graphics data requires about 30 kilobytes of transfer. With the reduced display size – thanks to letterboxing at the top and bottom, and pillarboxing on the sides – we’re able to transfer roughly 11 kilobytes per vertical blank.

“Multiply that by three vertical blanks (which corresponds to 20 frames per second), and you get around 33 kilobytes of data, which is exactly what’s needed to transfer the next frame. That’s how we’re hitting the theoretical throughput limit.”

The world of retro gaming and creating new games to run on old systems is thriving. There are few better ways to celebrate 30 years since Doom made its debut on the SNES than with a whole new cartridge containing a thoroughly-modern RP2350 taking on the duties of the innovative Super FX silicon. However, there doesn’t appear to be a delivery date for the cartridge and LRG did not respond to The Register’s queries. ®

Eddie AI has introduced a plethora of updates that should make the platform faster, more compatible, and easier to use. From broader NLE support to smarter media handling, the latest version is designed to fit more naturally into diverse editing workflows. Let’s see what’s new (again!) with Eddie AI.

Over the past year, we’ve covered how AI-powered tools are reshaping post-production, from automated rough cuts to text-based editing. We’ve written quite often about Eddie AI here at CineD, and noted its ability to accelerate the often time-consuming process of sorting, cutting, and syncing footage. In the last couple of months, they’ve added AI text-based editing to their rough cut mode, plus support for 30 languages. And it doesn’t seem to stop – we’ve got 50 new updates to take a look at! As a reminder, here’s Johnnie and Nino discussing what Eddie AI can do:

What is Eddie AI

So if you’re not yet familiar with Eddie AI, it’s a cloud-based platform that combines machine learning with text-driven video editing. You upload your footage, generate searchable transcripts, and quickly assemble rough cuts by working directly with text.

What’s in the latest release

Expanded NLE Integration and File Compatibility

• Eddie now edits in 30+ languages. Import your interview in any one of dozens of new languages. (Except Klingon.)
• Blackmagic Cloud and NAS Support: Eddie AI now supports Network Attached Storage (NAS) and Blackmagic Cloud. Boom.
• NLE Proxy: If you make proxies in your NLE, Eddie now pulls those instead of the source media, making it swifter to import.
• Updated Eddie extension for Resolve: Eddie is more stable for DaVinci Resolve imports and performs better with large projects.
• FCP Simple Timeline: If you cull interviews on the timeline first in FCP, you can now drop in FCP timelines to Eddie’s Rough Cut mode. Support already exists for Adobe Premiere Pro and DaVinci Resolve.
• Precise Timecode and Reference: Relinking is smoother now that start timecode inconsistencies across vendors have been addressed.
• Support for Apple App Store’s DaVinci Resolve: Compatibility issues with the App Store version of Resolve have been fixed.
• Optimized Clip Importing: Large batches of clips for B-roll logging import more efficiently, avoiding timeout issues.
• High Framerate Support: 90, 100, 119.88, and 120 FPS are now supported.
• Fixed Premiere Pro Motion Effect: The latest Premiere Pro bug affecting Motion Effects on edit sequences is resolved.
• Preventing Metadata Duplication: Eddie now detects and avoids re-importing FCP clips with existing metadata.

Enhanced User Experience & Account Management

• Pay-as-you-go, No Sub required: Buy credits without a subscription and use them as needed.
• Simplified Subscription Management: In-app unsubscribe and reactivation, plus direct billing portal access.
• Broadened Language Support: Define a default language for transcription.

Multicam and Project Workflow Refinements

• Improved Multicam Error Handling: Clearer error and warning messages with more flexible import options.
• Robust B-roll and A-roll Processing: Smoother performance, even under heavy loads.
• Automated Multicam Activation: Automatic switch to multicam mode when importing multicam clips.
• Enhanced Multicam Podcast Sync: Better angle detection for podcast cuts.

Refined Export and Media Handling

• Advanced MP4 Export and Captioning: Transparent outline styling and a new font for multilingual support.
• YouTube Video Import: Direct import of YouTube videos for editing, including vertical clip creation for social platforms.
• Better FCP Captions: Cleaner captions with fewer orphaned words.
• Robust Export Stability: Dozens of backend updates for smoother exporting.
• Smart Rough Cut Export: Chapter names now appear as ranged timeline markers.

Key Bug Fixes and Performance Enhancements

• Improved B-roll Analysis: Better support for pro codecs.
• Text Editing Reliability: More stable text-driven edits.
• Multicam Project Stability: Fixes for crashes related to R3D files and multicam audio thumbnail bugs.
• Mobile Upload Improvements: Better performance for Chrome and Android uploads.
• Streamlined UI: Cleaner transcoding, uploading, and sharing interfaces.

Who is Eddie AI for?

One editor reflected that after feeding a 100 GB+ interview into Eddie, it isn’t trying to be your editor – it’s trying to be your assistant editor. And that sentiment echoes through all its updates: Eddie isn’t meant to replace your intuition, your narrative instincts, or your human touch. It’s built to enhance them, to hand over the groundwork so you can focus on telling the story that only you can tell.

Price and availability

All the new features are now live for Eddie AI users. To try it out and to get more information, please visit the Eddie AI website.

Are you currently using Eddie AI in your projects? Have you noticed a significant difference in your workflow? Let us know in the comments!

The road to SAP S/4HANA is paved with data, and for many organizations still navigating this digital transformation, the journey is about a fundamental rethinking of their most critical asset. The challenge isn’t just moving to the cloud, but what to bring to the new system and what to leave behind.

Speaking ahead of Data Migration International’s (DMI’s) Digital Lounge @Lakeside 2025 event in Switzerland, Thomas Iseler, Vice President of Strategic Partners at DMI, discussed the real-world data challenges on the path to a modern ERP.

“For the longest time data has been an afterthought,” Iseler explained. “Then the shortcut was always to take all data.” However, he added that this approach can create a ripple effect of problems, from sluggish system performance to bloated project timelines and costs. Moreover, in today’s AI-driven landscape, that old mindset is a recipe for failure.

According to Iseler, a successful SAP S/4HANA data strategy rests on three key pillars:

1. Embrace a Clean Data Core. The first rule of a modern migration is to be selective. “You should only migrate data that is relevant and that is needed,” Iseler emphasized. Bringing decades of historical clutter into a new, agile environment undermines the purpose of the transformation. Iseler added that focusing only on necessary data is key for system and process performance as well as the speed and complexity of the migration itself.
2. Prioritize Data Quality. While good data has always been necessary for reporting and transactions, AI raises the stakes exponentially. “With the growing importance of AI, the relevance and importance of data quality has grown significantly,” Iseler noted. Without clean, reliable, and harmonized data, any investment in AI-driven business outcomes is built on a shaky foundation.
3. Solve the Legacy Data Puzzle. Historical information is often essential for compliance, audits, and analytics, and traditional archiving is not a smart and efficient option. Leaving it scattered across aging, expensive-to-maintain legacy systems is a huge risk. The goal, Iseler said, is to figure out “how to make that data readily accessible, and how to make storing it transparent and cost-effective.”

Ultimately, the move to SAP S/4HANA is an opportunity to break from the past. It’s a chance for organizations to transform data from a liability into their most powerful asset. As Iseler concluded, with the rise of AI, data has moved from the periphery to the core. “It’s a complete reversal from being an afterthought. Now it has to be the center of the strategy.” For organizations embarking on their transformation, making data the central character in their story is the key to a successful outcome.

Watch the full interview here.

What This Means for ERP Today Insiders

Seek a unified partner rather than a point solution. Modern digital transformation is a complex web of interconnected data challenges, encompassing migration and quality, as well as archiving and compliance. Engaging multiple vendors for different pieces of the puzzle adds complexity, risk, and cost. For example, DMI is a strategic partner with a single, unified platform, JiVS, that manages the entire data lifecycle. It handles migration, data quality, harmonization, masking, and legacy data retirement from a single point of contact. This approach simplifies governance, reduces project timelines, and ensures consistency, proving critical for organizations looking to build a flexible and future-proof data architecture.

Your AI and cybersecurity posture depends on data strategy. An organization’s future capabilities in AI and its current security resilience are directly tied to how it manages data today, especially historical data. AI is useless without high-quality, relevant, and accessible data. By focusing on what Iseler describes as a “clean data core” and ensuring data quality is paramount during the transformation, DMI’s approach helps build the essential foundation for reliable AI and analytics. Additionally, legacy systems are a huge cybersecurity risk. They are often unpatched, difficult to monitor, and managed by a shrinking pool of experts. Decommissioning these systems and moving historical data to a modern, secure, and audited platform dramatically reduces the attack surface and helps ensure regulatory compliance for decades of data.

The time to move to the cloud is now, and it’s more feasible than you think. The primary roadblocks to migration, such as cost, complexity, and downtime, are directly addressed by a smart data strategy. By migrating only a small fraction of relevant data and intelligently managing the rest, the migration project becomes faster, cheaper, and less risky. Furthermore, a lean, clean system in the cloud is more agile, scalable, and ready to connect to modern innovations like SAP Business Technology Platform (BTP) and Business AI.