Tag Archives: Artificial Intelligence

Dyson’s $499 Camera Toothbrush Is So Advanced, They Had To Redesign The Toothpaste For It

Posted on by

Dyson’s new $499 CameraJet hides a 1mm camera under the brush head, reading 28 images a second, and within 100 milliseconds of spotting a gap between your teeth it fires a cone of mouthrinse straight into it. That single lens is why Dyson also had to formulate its own non-foaming toothpaste, because bubbles blind a camera. It is also why the 12.5ml tank runs a diaphragm pump and pressure chamber, so the jet still fires when you tilt the handle sideways to reach your back molars. One optical decision rewrote the rest of the product. That cascade is the actual story here, and it is a lovely bit of design logic.

James Dyson unveiled it in Paris this morning, capping six years of work, 661 engineers and 38 patents. The vision system is called Gap Optical Targeting, trained on 470,000 dental images and running 16 million lines of code, all so a toothbrush can find the spaces nine in ten of us never floss. It ships today at $499, or £419.99 in the UK, in Ceramic Ultra Blue and Ceramic Pink. You also get a dock that refills the tank in three seconds and RFID brush heads that count their own cleans. Every one of those parts exists because somebody decided the brush should be able to see.

Designer: Dyson

The jet took years of refinement on its own, because the shape of the water matters as much as the camera aiming it. Traditional water flossers use what Dyson calls a needle jet, a narrow, piercing stream that punches through a plaque deposit without actually lifting it away. Dyson’s engineers went hunting for a shape that would sweep plaque off the tooth instead of tunneling through it, and landed on a broad conical spray fired at lower pressure. To test that idea without running a clinical trial every time they tweaked the nozzle, Dyson spent five years building its own synthetic plaque with the National University of Singapore’s Faculty of Dentistry. That is the unglamorous groundwork that never makes the keynote slide but decides whether the whole idea holds up.

None of that fluid engineering matters if the tank runs dry the moment you tilt your head back, which is the next problem the camera created. Conventional reservoirs use a straw-in-a-cup setup that draws air and loses pressure on an angle, fine for a kitchen gadget, useless inside a mouth that never sits still. Dyson’s fix is the anti-gravity tank, built around a pump that adjusts as you move so the jet fires whether the brush is upright, angled, or flat against a back molar. The brush head carries its own logic too, firmer bristles inside for stains, softer ones outside for the gumline, contoured to actually hug a tooth rather than sit flat against it. Even the sonic oscillation varies its angle mid-stroke so bristles do not stall in the tighter corners.

The camera does double duty as a diagnostic tool, streaming live through the MyDyson app so the brush briefly becomes a tiny endoscope. Dyson says nothing is recorded or stored, on the device or in the cloud, and a light on the handle shows when the lens is active. Cameras inside toothbrushes are not new. A Chicago dentist named Craig Kohler built one back in 2016 called the Prophix, a video-streaming brush that let people watch their own mouths on an iPhone in real time. It never went mainstream, and its own website still reads coming soon, a decade on. What Dyson brings that Prophix never had is the industrial muscle to turn that footage into an automated response, steering a jet instead of just showing a picture.

Dyson CameraJet in situ imagery, JPEG, PSD

Dyson has spent decades arguing that the dirt you cannot see is the dirt worth engineering around, first in carpets, then in bathroom air, now in the gap between your incisors. The CameraJet is that philosophy at its smallest scale yet, a lens standing in for the clear bin that made a bagless vacuum’s dust visible for the first time. What makes it worth writing about as design, rather than just as gadgetry, is how completely one sensor reorganized everything around it, the chemistry of the paste, the physics of the pump, the shape of the bristle. That is a rare thing to see fully worked through in an object that costs less than a laptop. Whether $499 buys cleaner teeth is a question for your dentist in six months. Whether it buys a masterclass in constraint driven design is not really a question at all.

The post Dyson’s $499 Camera Toothbrush Is So Advanced, They Had To Redesign The Toothpaste For It first appeared on Yanko Design.

This Startup Says Your House Should Be the Robot, Not ‘Have’ a Robot

Posted on by

Most of the AI money right now is betting on robots. Hundreds of embodied-AI startups and their backers are convinced that a bipedal machine will eventually become the household’s operating system, rolling from room to room and handling everything. Endong Zhang, the founder and CEO of LiveLarge Home Inc., thinks that’s the wrong frame, and he’s pretty candid about it.

A robot, he points out, has limited battery life, limited field of view, and can only be in one place at any given moment, which makes it a fairly weak candidate for running a multi-room household with several people in it. His counterproposal is that the building itself is the smarter infrastructure bet, and he’s building a company around exactly that.

Click Here to Know More

The company Zhang founded in 2023 spent its first few years looking more like a serious California construction operation than a tech startup, delivering more than 20 accessory dwelling units (ADU) and residential projects across the Bay Area and Los Angeles.

Endong Zhang – Founder and CEO, LiveLarge Home

In July 2026, LiveLarge revealed the Flagship L-Space, a standalone AI-native room environment and the clearest physical statement of Zhang’s thesis so far. Light, sound, air, privacy glass, and a voice assistant called Lila are coordinated as one system rather than a pile of separate apps, organized around three modes: Focus, Reset, and Passion.

Space AI: the invisible housekeeper

The conceptual core of Space AI is a deceptively simple question: can AI give every household some version of what a housekeeper provides? Zhang keeps returning to the housekeeper image because it captures something the smart home industry never quite cracked. A housekeeper reads the room, observes what’s needed, and acts without being told every small step. A traditional smart home, by contrast, receives a command or follows a fixed rule, and it doesn’t understand the situation, decide what should be done, or make a plan. “Without a human housekeeper,” Zhang says, “the home itself has to listen, observe, understand, decide and act.” That’s the jump LiveLarge is trying to make. LLMs, multimodal models, and a maturing robotics ecosystem are what make the timing feel genuinely different now rather than aspirationally so.

The room integrates a 5.1 surround sound system, adaptive lighting, fresh-air circulation, a 4K short-throw projector, and privacy glass into a single coordinated environment, all orchestrated by an AI assistant called Lila. The three modes, Focus, Reset, and Passion, were shaped directly by early users rather than invented in a whiteboard session. “People do not really want smarter devices,” Zhang told us. “They want to focus better, relax, recover, or spend time on something they care about.”

LiveLarge makes some deliberate engineering choices when it comes to privacy, with pretty straightforward reasoning: a room that watches, listens, and learns is exactly the kind of product that erodes trust the moment something leaks. The first rule is to do as much computation locally as possible, and L-Space uses an NVIDIA Jetson Orin as its local AI computer, with a large part of the sensing, data processing, storage, and inference happening inside the product. Sensitive information, including conversations, biometric signals, and usage data, can stay local.

The system uses voiceprints to recognize regular users, each user has a separate history and profile, and User A cannot access User B’s conversations or usage records. The full permission system is still in development, but the architecture already accounts for different access levels across family members, guests, and service providers. Zhang’s framing is tidy: “Personalization only makes sense within the permission given by the user.”

The Flagship L-Space is what Zhang calls a “concept car,” and the analogy holds up across more dimensions than the marketing might suggest. The Standard and Flagship models share the same core Space AI platform, but the key addition in the Flagship is vision: locally processed computer vision that lets Lila recognize individual users, read gestures, and actually observe what’s happening in the room rather than waiting to be addressed.

Vision is, in Zhang’s view, one of the most important sensors for a future AI-native home, and once AI can actually see a space, many services that are difficult today become possible. He’s testing facial identification and gesture control as more natural interaction layers on top of voice, and all of it runs on-device because visual data inside a home should never casually leave the building. “The engineering question is simple,” he says. “If AI can really understand a room, what useful services can it provide to the people inside it?”

The biggest lesson in getting from concept car to delivered product is that a lab demo and a product are different engineering objects. In the lab, every cable, firmware version, network condition, and room setup is controlled. A delivered L-Space has to survive shipping, installation, startup, software updates, and daily use. The customer also needs to set it up quickly, understand how to use it, and trust that the basic functions will keep working over time.

At current production volumes, many subsystems still use consumer-grade components, and the challenge of packing computing, storage, networking, IoT control, audio, projection, sensors, and actuators into a compact structure without compromising acoustics or airflow is, as Zhang puts it, closer to vehicle electronics than construction. “A feature is not finished when it works once,” he told us. “It is finished when it works repeatedly in the customer’s space.”

Technology is reshaping architecture and interior design

What the home looks like ten years from now, in Zhang’s telling, reads more like a tech product manifesto than an architecture brief. The house becomes a device in the same way a phone or laptop is a device: one integrated hardware and software system where every new appliance, robot, or sensor plugs into a shared runtime rather than living as yet another isolated smart product.

He also believes the intelligence shouldn’t stay tied to any specific building. “If you move, the home model, preferences and memory should move with you,” he says. “The new house becomes the new body, but the agent continues.” For designers, he sees a real reskilling moment coming, one where sensor placement, robot clearances, and interaction architecture start sitting alongside floor plans and material palettes as standard deliverables.

LiveLarge’s roadmap for the second half of 2026 includes bringing Space AI into ADU and broader residential products, which means the concept car phase has a fairly clear expiry date. You can see what the company is building at livelargetech.com.

Click Here to Know More

The post This Startup Says Your House Should Be the Robot, Not ‘Have’ a Robot first appeared on Yanko Design.

Apple TV 4K Is Coming September 9, With a Gemini-Powered Siri You Can Actually Talk To

Posted on by

Apple’s next Apple TV 4K is expected on September 9, and the headline feature is a Siri you can hold a conversation with. Siri AI runs on a custom 1.2 trillion parameter Google Gemini model inside Apple’s Private Cloud Compute, which Apple reportedly pays Google around a billion dollars a year to license. It should reach the public with tvOS 27 in mid September, most likely around Monday the 14th. Leaked code in the macOS 26.7 release candidate lists an unreleased controller as ATVRemote1,5. Four years after the last box shipped, the living room finally gets the good Siri, and the remote gets demoted.

The current model has been on sale since November 2022 with an A15 Bionic chip, making it the longest running Apple TV between updates. Apple raised its price anyway in June, from $129 to $199 for the Wi-Fi version and $149 to $249 with Ethernet, blaming the memory chip shortage. The new box is rumored to carry an A17 Pro chip or newer, 8GB of RAM, and Apple’s N1 wireless chip for Wi-Fi 7, Bluetooth 6, and Thread. Apple has confirmed exactly none of it. My interest is in the stick, because a conversational assistant changes what a remote is even for.

Designer: Apple

The Siri Remote has always been the awkward part of this product. The 2015 glass touchpad version was famously easy to hold upside down. The 2021 aluminum redesign fixed orientation with a circular clickpad and a jog ring, but it stayed thin, cold, and easy to lose in a couch. Nobody expects ATVRemote1,5 to look radically different, although one can hope. The rumor mill’s best guess is a second generation Ultra Wideband chip for Precision Finding, which is a nice way of saying Apple is finally building a remote finder into the remote instead of asking you to turn your house upside down, or worse, strap a clunky AirTag to it.

Siri AI’s 1.2 trillion parameter Gemini model dwarfs the roughly 150 billion parameter models Apple’s own Foundation Models run on, and Google built it specifically to handle summarizing and multi step planning, the tasks Siri has failed at for over a decade. Apple’s answer to the privacy question is architectural rather than rhetorical: queries route through Private Cloud Compute, Apple’s own server silicon, so the assistant is Google’s brain wearing Apple’s skin. Whether Apple TV gets the full version or a scaled down version tied to on device processing is still unconfirmed.

None of this works without new silicon, which is presumably why Apple TV and HomePod mini were left out of the WWDC keynote entirely despite tvOS 27 shipping Apple Intelligence frameworks in its beta code. Current Apple TV and HomePod hardware simply cannot run it. That is also the best explanation for why Apple hiked the price of four year old hardware by 54 percent in June rather than discounting it toward obsolescence. The company was buying time, not margin.

What I find genuinely interesting is the inversion. For a decade, the remote was the product’s weak point and the box quietly did fine. Now the box is getting the transplant everyone wanted, and the remote becomes almost optional, a fallback for when you would rather not talk to your television. If Apple pulls off a Siri that understands “put on the thing with the guy from that show” without three follow up questions, the aluminum wand in your couch cushions finally gets to retire.

The post Apple TV 4K Is Coming September 9, With a Gemini-Powered Siri You Can Actually Talk To first appeared on Yanko Design.

Plaud One AI Earbuds remember and summarize your conversations so you can focus on the work

Posted on by

Most wireless earbuds are designed to disappear into the background while you listen to music or take calls. Plaud One is more ambitious compared to your normal buds, as it wants to remember what happens around you and then help turn those conversations into actual work.

The AI earbuds can record face-to-face conversations, phone calls, or online meetings without needing to stay connected to a smartphone. That is possible because the charging case includes built-in 4G LTE connectivity through an eSIM. Plaud says it has arrangements covering more than 80 countries and regions, allowing the earbuds to upload recordings even when your phone or Wi-Fi connection isn’t nearby.

Designer: Plaud

Recording isn’t limited to the earbuds, either. The case has its own microphones, allowing it to work as a standalone recorder when placed on a meeting table. Users can start recording from the earbuds, the case, or the app, while a recording indicator on the case provides a visible cue that audio is being captured. The earbuds can also natively capture audio from video-calling services such as Google Meet and Zoom. Once a conversation has been captured, Plaud’s software can transcribe it, summarize the discussion, and identify useful information. The more interesting step is what happens next. Plaud Agent is designed to use that conversational context to prepare follow-up work, such as an email, report, document or presentation.

It can connect with services including Gmail, Google Calendar, Notion and Slack, potentially allowing a meeting to produce something useful before you even return to your desk. That makes Plaud One different from an AI recorder that simply leaves you with a transcript to sort through later. The company is positioning the wearable as an AI system that can remember information across conversations and use that context to help automate workflows. You can also interact with the assistant using a voice command or the dedicated button on the case.

There is still a conventional pair of wireless earbuds underneath all this. Plaud One uses Bluetooth 5.4 with multipoint connectivity, supports SBC and LDAC codecs, and includes active noise cancellation rated for up to 40dB of reduction. The earbuds are IP54-rated for dust and splash resistance. Battery life is rated at up to six hours for music playback with ANC off, while the case adds up to 36 hours of additional listening time. Storage is built into the system as well, with 16MB in each earbud and another 512MB in the case. Plaud says the case can hold up to 30 hours of recordings. Recording endurance from the earbuds varies depending on the use, with up to six hours for in-person recording and around three hours for calls and podcasts.

The $250 Explorer Edition comes with $200 worth of Plaud Credits for its agentic AI features, but that allowance is not unlimited. Heavier transcription and AI use can require paid plans or additional credits, making the ongoing cost an important part of the product’s proposition.

The post Plaud One AI Earbuds remember and summarize your conversations so you can focus on the work first appeared on Yanko Design.

This AI-Powered Nature Guide Is Designed to Reduce Children’s Screen Time

Posted on by

Most technology designed for children asks them to look down. Quest asks them to look around. At first, the idea of using artificial intelligence to reduce children’s screen time may sound contradictory. Yet that contradiction is exactly what makes Quest compelling. Rather than competing with the physical world for a child’s attention, this AI-powered nature companion redirects it towards leaves, insects, birds, rocks, textures, sounds, and the small details that often go unnoticed.

Designed for children aged six to eleven, Quest transforms outdoor discovery into an ongoing learning adventure. It is not a conventional educational device, nor is it simply a camera that identifies objects. Its purpose is more ambitious: to turn technology into a quiet bridge between curiosity and the natural world.

Designer: Viraj Gapchoop

During the day, children carry Quest while exploring parks, gardens, trails, or even their own backyard. When they encounter something unfamiliar, they can use the device to identify it. Quest’s AI-powered vision system recognizes plants, insects, birds, rocks, and other natural objects, confirming each discovery with gentle haptic feedback before responding through conversational audio.

The absence of a dominant screen is an important design decision. Instead of pulling the child into a visual interface, Quest keeps their attention on the object in front of them. Audio and touch become supporting cues, allowing the experience to remain rooted in the physical environment.

More importantly, Quest is not designed to answer every question immediately. It might ask a child to look more closely at the shape of a leaf, listen to the rhythm of a bird’s call, or compare the surface of two stones. Through these prompts, the device encourages observation before explanation.

This subtle shift matters. Rather than positioning AI as an all-knowing authority, Quest uses it as a conversational guide. The child is not simply given information; they are invited to notice, interpret, and investigate. Learning becomes participatory, and the moment of identification becomes an “aha!” experience earned through attention.

The concept is built around three principles: Capture, Converse, and Celebrate. Children capture what interests them, converse with Quest to better understand it, and later celebrate those discoveries through storytelling.

That final stage begins at night. When placed into its charging dock, Quest transforms into a bedside storytelling projector. The plants, insects, feathers, rocks, and animals discovered during the day become the ingredients for a personalized bedtime story, with the child placed at its center.

A beetle found beneath a stone might become the guardian of an underground kingdom. A feather collected on a walk could inspire a journey across the clouds. A flower discovered in the garden might hold the secret to restoring an enchanted forest.

In this way, Quest does more than reinforce factual learning. It gives emotional meaning to discovery. Information gathered during the day is reimagined through narrative, helping children connect memory, imagination, and knowledge. Nature is no longer something observed once and forgotten; it becomes part of a personal story.

The nighttime experience also transforms Quest from an individual learning device into a shared family ritual. Parents and children can revisit the day’s discoveries together, talk about what they found, and experience a story shaped by those moments. Bedtime becomes an opportunity for reflection, connection, and imagination rather than another isolated digital interaction.

This is where Quest’s human-centered approach becomes most visible. The product does not attempt to replace outdoor play, parental involvement, or a child’s imagination. It is designed to strengthen all three.

Still, designing AI for children requires restraint. Quest’s success would depend not only on what it can recognize or generate, but on knowing when to speak, when to ask a question, and when to remain quiet. If every encounter is interrupted by information, curiosity can quickly become instruction. The device must support independent exploration without over-directing it. That balance is central to the concept. Quest does not aim to make nature more digital. It uses digital intelligence to make children more attentive to nature.

Young children are often described as asking hundreds of questions each day, a reflection of their instinctive desire to understand the world. Quest is designed around protecting that instinct rather than replacing it with instant answers. It treats curiosity not as a problem to resolve, but as something to extend.

At a time when many digital products are designed to hold attention for as long as possible, Quest proposes a different role for technology. Its interface is deliberately quiet. Its value lies not in keeping children engaged with the object itself, but in helping them engage more deeply with everything beyond it. The most interesting thing about Quest, then, is not that it uses artificial intelligence. It is that the intelligence knows when to step aside.

The post This AI-Powered Nature Guide Is Designed to Reduce Children’s Screen Time first appeared on Yanko Design.

Canyon’s New Bike Knows a Crash Is Coming Before You Do

Posted on by

Road cycling has always been a sport that rewards the rider who pays the most attention. You watch the road, you watch the traffic, you watch the rider in front of you. You’re constantly processing information, making micro-decisions at every turn. Which is exactly why Canyon’s Predict concept bike, which made its debut at Eurobike 2026, feels less like a cycling product and more like a quiet but serious rethinking of what a bike is allowed to be.

The premise is deceptively simple: rather than reacting to danger, the bike anticipates it. Canyon built the Predict with a 360-degree sensor array, combining radar, cameras, and a multi-dimensional motion sensor, all working together under an Edge AI system that assigns risk scores to what it detects and communicates those scores to the rider in real time. Four radar sensors total, three at the front and one at the rear, plus a data screen built directly into the handlebar. The whole setup is compact, integrated, and far sleeker than you’d expect from something packing this much hardware.

Designer: Canyon

What makes the Predict genuinely interesting isn’t the specs list, though. It’s the shift in design philosophy behind it. Cyclists have had cameras and GPS computers bolted onto their bikes for years, but those tools are still fundamentally passive. You look at the screen when you want to. You check the rearview camera on your own time. The Predict doesn’t wait for you to ask. It tracks group-ride dynamics, advises on cornering speeds, and predicts hazardous road surfaces, all before the rider even registers a problem. The bike talks to you through haptic signals in the handlebars, which means your eyes don’t have to leave the road for the alert to land.

That detail matters more than people might realize. Most cycling safety gear operates on the assumption that the rider has a free moment to process information. The Predict is designed around the fact that sometimes you simply don’t. It’s a subtle but meaningful distinction, and it tells you a lot about where Canyon’s thinking really is right now.

The bike also pairs with the Canyon Stingr Smart Helmet, which features a drop-down visor with its own data visualization screen. Wearing that helmet while riding the Predict, you’d essentially have a full 360-degree awareness system wrapped around you, the kind of setup that until recently existed only in motorsport or military contexts. Seeing it applied to a road bike is either thrilling or slightly unsettling, depending on your disposition. I lean toward thrilling.

There’s a natural question worth naming directly: at what point does all this technology change the experience of riding rather than just enhancing it? Cycling has a deeply personal quality to it. The appeal, for a lot of people, is the relationship between body, machine, and road, the very human act of reading the environment yourself. The Predict does a portion of that reading for you, and I don’t think that’s a bad thing, but it’s worth sitting with.

My honest take is that Canyon isn’t trying to automate the rider out of the experience. They’re trying to buy back the rider’s attention for the things that matter most: the road, the ride, the moment. The safety layer runs in the background so you don’t have to. That’s a thoughtful design goal, and it shows in how cleanly the technology is packaged. Nothing about the Predict looks bolted on or retrofitted. It looks like what a bike is supposed to look like, with the tech quietly embedded inside.

The Predict is still a prototype. It’s not heading to market tomorrow, and real-world software testing still stands between this convention floor concept and a commercially viable product. But the hardware is largely functional today, the vision is clear, and the intent is serious. Canyon isn’t just imagining a safer bike. They’re doing the actual work to build one, and that alone is worth paying attention to.

The post Canyon’s New Bike Knows a Crash Is Coming Before You Do first appeared on Yanko Design.

Milan Smart Glasses pack a 12MP camera, live translation in a stylish ‘non-gadget’ design

Posted on by

Le Corbusier wore round glasses. John Lennon wore round glasses. Several decades of architects, artists, and intellectuals who wanted to look like they read a lot wore round glasses. There is a long, slightly pretentious tradition attached to the panto and round frame silhouette, one that signals “considered, creative, European in a very specific mid-century way.” L’Atitude 52°N, the Berlin-based eyewear brand behind the Departure Collection, named their panto frame Milan for a reason, and they loaded it with a 12MP Sony camera, an AI assistant, and real-time translation into five languages before shipping it out the door.

Milan launches August 25th in the US and Europe, and the hardware it carries makes a more interesting case than the marketing copy would suggest. A 12MP Sony IMX681 sensor with a 107-degree ultra-wide field of view captures 1080p HD video and high-resolution stills, with a 6nm AI chip handling noise reduction, color processing, and exposure in real time. The Goya AI assistant manages music, calls, navigation, and battery checks by voice, while an AI Tour Guide feature uses the camera to visually identify landmarks and deliver 30-second expert commentary on demand. Live two-way translation covers five languages, all through the frame’s open-ear speakers and five-mic array. It comes in Olive and Obsidian from $399, with the photochromic lens option jumping that to $449.

Designer: L’Atitude 52°N

Click Here to Buy Now: $449

Milan’s matte finish and rounded panto silhouette exist in deliberate contrast to the BERLIN frame’s angular, navigator-coded geometry: BERLIN for a classic utility vibe, MILAN for culture. The EMS TR90 frame and self-adjusting temples bring the total to about 50 grams with lenses, genuinely light for a frame carrying cameras, microphones, speakers, and a 6nm AI processor inside it. Two colorways launch, Olive and Obsidian, each paired with gradient tinted or photochromic lens options that auto-adapt to changing light. Stainless steel spring hinges carry an ultra-wear-resistant nano technology coating tested over 20,000 cycles, three times harder than steel by L’Atitude 52°N’s own testing. IP65 sealing and a UV-hardened finish handle the elements, so the build is more rigorous than the silhouette lets on.

The Sony IMX681 sensor sits at the bridge with a 107-degree field of view that approximates natural peripheral vision reasonably well, and the 6nm AI processor handles real-time noise reduction, exposure calibration, and color enhancement so stills are usable without post-processing. Portrait and landscape modes are both supported, which matters for anyone managing content across platforms with different aspect ratio preferences. A tap on the frame or a “Hey Goya” voice command triggers capture, keeping hands free and the moment intact. 32GB of onboard storage holds everything locally until you sync via the companion app. The 1080p video ceiling is the one number that will catch attention from anyone shopping around, given that rival frames at lower price points are already delivering higher resolutions.

The AI Tour Guide uses the camera’s visual recognition to identify landmarks and artworks, delivering 30 seconds of expert commentary after a tap or voice trigger. Live Translation handles two-way voice conversation in English, German, French, Italian, and Spanish, with each session running up to five minutes before auto-closing. General Goya tasks work hands-free via the “Hey Goya” wake phrase, covering navigation, music, battery status, call management, and quick Q&A. All of it runs directly through the glasses without needing to pull out your phone or break the flow of whatever you are doing. A group communication Intercom Strap accessory is also in development, which will add offline mesh-networked voice chat with up to eight people simultaneously.

Open-ear stereo speakers and a five-microphone array handle audio in both directions, with AI noise reduction keeping voice commands readable even in crowded environments. Privacy is handled with more care than most of the category bothers with: photos and videos live locally on the device and are deleted after syncing to your phone, and Goya AI interactions leave nothing stored in the cloud. A physical LED on the frame lights up whenever recording is active, and covering it stops the camera immediately, audio included. Milan is built for GDPR compliance, which reflects the brand’s German roots and resonates with European buyers especially. Wi-Fi 6 and Bluetooth 5.4 handle connectivity, and the 200mAh battery delivers up to six hours of mixed use, with an IP67-rated charging case offering eight to ten full top-ups and a 20-minute fast-charge to 70%.

At $399 base and $449 with photochromic lenses, Milan is priced at the premium end of what is now a genuinely crowded field, with strong hardware available at lower price points from several well-funded players. The value case lives in design and privacy architecture rather than raw specs. The panto silhouette is more considered and more wearable than most competing frames, and the local-only media storage model is among the tightest privacy implementations shipping in this category today. Six hours of mixed-use battery is workable for a day out, but worth factoring in for anyone planning longer trips. The AI feature set is a meaningful differentiator, particularly for solo travelers navigating unfamiliar cities who would otherwise juggle translation apps, maps, and camera apps simultaneously.

Milan is available from August 25th on the L’Atitude 52°N website, with pre-orders placed before that date qualifying for a $50 discount. Gen 1 Departure Collection buyers, covering both BERLIN and MILAN, get full access to Goya AI features out of the box, including Voice Activation, AI Tour Guide, Live Translation, and AI Image Recognition, with no subscription required. L’Atitude 52°N has committed to giving users advance notice of any changes to that access model. The smart glasses space is genuinely crowded in 2026, with major players from tech and fashion all competing for the same audience. Milan’s answer is a panto frame, a matte finish, and a camera that asks you to simply look at the thing you want to remember. Le Corbusier would probably have approved, and he was notoriously difficult to impress.

Click Here to Buy Now: $449

The post Milan Smart Glasses pack a 12MP camera, live translation in a stylish ‘non-gadget’ design first appeared on Yanko Design.

At $69.99, the Boya Notra Neo Is the AI Note-Taker That Students Can Finally Afford

Posted on by

The Plaud NotePin costs $169. Boya’s own Notra flagship launched in February 2026 at $149.99. For a stretch, if you wanted a dedicated AI recorder with real transcription and AI-powered organization, you were looking at spending north of $150, which quietly priced the category out of reach for the students and young professionals who arguably needed it most. Every new tech category does this: it launches expensive, finds its early adopters, and then figures out how to bring the price down. Boya just compressed that timeline significantly with the Notra Neo, a $79.99 AI recorder that retains the core AI note-taking experience of the flagship.

Launching September 4th at IFA 2026 in Berlin, the Notra Neo targets students, young professionals, multilingual users, and anyone who has been curious about AI note-taking but wasn’t ready to drop $150 on the experiment. The spec differences from the Notra are real but they’re calibrated: storage drops from 64GB to 8GB, the ambient pickup range goes from 10 meters to 5 meters, and battery life trims from 24 hours to 20. The core AI feature set remains available, with transcription across 140-plus languages, speaker identification, summaries, to-do lists, mind maps, and AI Search across your entire recording history. Automatic switching between Ambient, Phone Call, and Bluetooth recording modes means the device reads the situation and adjusts without you touching a setting. Pre-orders open August 24 at $69.99.

Designer: BOYA

Click Here to Buy Now: $69.99 $79.53 (12% off). Hurry, deal ends in 48-hour!

Boya’s automatic scenario detection handles context management silently: Ambient kicks in for face-to-face conversations, Phone Call for phone calls (obviously), and Bluetooth when you’re on a virtual meeting through your earbuds. In Ambient mode, the device picks up 360-degree audio within a 5-meter radius, covering most classroom or conference table setups without precise positioning. For first-timers already managing a phone, a laptop, and half a dozen tabs, that simplicity matters more than any individual spec. None of this requires navigating a settings menu mid-session or remembering which mode you last used. A similar friction-removal logic powered the REETLE E-Ink iPhone case YD covered last year, where a single dedicated record button built into the frame turned out to be the one design decision that made the entire product click.

Keeping the full AI feature set identical to the flagship Notra is genuinely the most interesting product decision Boya made with the Neo. Transcription runs across 140-plus languages and dialects with speaker labeling, making it a credible tool for multilingual classrooms, international teams, and professionals who regularly conduct conversations across more than one language. The AI layer generates summaries, to-do lists, mind maps, and insights from those recordings, with 60-plus professional templates calibrated for different scenarios. AI Search lets you query your entire recording history without scrubbing through audio, which shifts this from a recorder into a searchable knowledge base. Everything syncs and exports through the Boya OS App and Web, with format options for audio, transcripts, summaries, and mind maps.

The Notra Neo boasts a cleaner front face than its flagship counterpart, swapping the “BOYA Notra AI” label for a simple “Boya” wordmark on a smoother surface. The form factor follows the compact square silhouette now standard in this category, but the softer branding reads younger and more consumer. Nothing about the physical profile announces entry-level in the way cheaper electronics sometimes can’t avoid, which matters for users already self-conscious about carrying a dedicated recorder. At $80, the Notra Neo competes less with the Plaud than it does with the phone propped against a water bottle, and that’s a smarter fight to pick.

Pre-orders for the Notra Neo open August 24, 2026, at a discounted $69.99. At its budget-friendly price, this is the most accessible entry point into dedicated AI recording the category has seen, and for anyone who has been putting off the experiment, the excuse just ran out.

Click Here to Buy Now: $69.99 $79.53 (12% off). Hurry, deal ends in 48-hour!

The post At $69.99, the Boya Notra Neo Is the AI Note-Taker That Students Can Finally Afford first appeared on Yanko Design.

RayNeo Just Split Smart Glasses Into Two, and Neither One Compromises

Posted on by

The smart glasses category has long struggled with identity. Most devices try to do too much and end up doing nothing all that well. RayNeo’s answer to that is two entirely separate glasses for opposite ends of the spectrum: the iO as a discreet all-day AI companion and the GT Series as a portable personal theater. It’s a straightforward bet that specialization beats compromise.

The iO is the one you’d wear to a meeting and forget you have on. At 33 grams, with an aerospace-grade titanium frame and magnesium-aluminum alloy construction, it distributes weight evenly enough to disappear on your face. The display is invisible to anyone looking at you, a 97% transparent waveguide that delivers text at 1,300 nits of brightness while looking like ordinary eyewear to everyone else in the room.

Designer: RayNeo

What the iO actually does with that display is where things get interesting. Sit through a meeting, and it’s quietly logging the conversation, capturing context, and flagging action items that you can approve with a nod. Step up to give a presentation and its Auto-Teleprompter scrolls your script in sync with your speaking pace. For frequent travelers, it translates across 55 languages and 109 accents in near real time.

The iO doesn’t have speakers or a camera, which is a deliberate call. A four-mic array paired with a bone conduction sensor picks up your voice clearly even in noisy environments, and the 240mAh battery stretches to two days of regular use. Bluetooth 5.4 with MFi certification, IP54 water resistance, and a tactile Smart Crown rotary dial on the right temple complete the hardware.

The GT Series is a completely different proposition. Where the iO blends in, the GT and GT Max make no apologies for what they are: personal cinemas you wear on your face. The standard GT projects a 231-inch virtual screen at a 46° field of view, and the GT Max scales that up to a maximum of 307 inches at 59°, adjustable across three selectable sizes.

Paired with the RayNeo Pocket TV Pro, a Google TV-powered streaming device with a 6,500mAh battery, the GT Series is also the world’s first AR glasses system to support native Dolby Vision playback. That companion device powers both itself and the glasses for about 5.5 hours of streaming, which means actual Dolby Vision on a screen that fills more of your visual field than most living room televisions.

Audio in the GT Series comes from a quad-speaker system co-tuned by Bang & Olufsen, using a six-magnetic circuit design that delivers a 5dB boost in low-frequency sensitivity. Head-tracked spatial audio keeps the soundstage anchored as you move, and a dedicated Steady Mode in the 3DoF chip filters out turbulence and vehicle vibration, making it a natural companion for long flights or train journeys.

Both products go on sale September 4, 2026, via RayNeo’s website and Amazon. The GT starts at $299 and the GT Max at $399, with pricing for the iO still to be confirmed. Custom prescription lens inserts are supported across both GT models, and the GT Max comes in three IPD sizes to accommodate different face shapes. It’s a two-pronged launch that doesn’t hedge its bets on either end.

The post RayNeo Just Split Smart Glasses Into Two, and Neither One Compromises first appeared on Yanko Design.

Tecno Camon 50 Ultra 5G Review: A $420 AI Phone With Built-in Gemini, ChatGPT, DeepSeek, and Grok

Posted on by

PROS:


  • EllaClaw is one of the most impressive phone AI experiences available right now

  • Multi-model AI support, including Gemini, ChatGPT, Grok, DeepSeek, Meta AI, and Ella

  • Strong 2-day battery life from the 6500mAh cell

  • Competitive value in the mid-range segment

CONS:


  • Single speaker setup sounds tinny and lacks bass

  • Curved waterfall display is prone to accidental palm touches

  • No wireless charging or Qi2 (MagSafe)

RATINGS:

AESTHETICS
ERGONOMICS
PERFORMANCE
SUSTAINABILITY / REPAIRABILITY
VALUE FOR MONEY

EDITOR'S QUOTE:

The Camon 50 Ultra 5G feels like Tecno looked at the current AI phone race, skipped the walled garden nonsense, and built the first assistant experience that actually shows you why you’d want to use it in the first place.

Apple has Apple Intelligence. Google Pixel has Gemini. Samsung has Galaxy AI. Tecno has everything else. Packed into its EllaClaw agent is the ability to use Gemini, ChatGPT, Grok, and DeepSeek, because a walled garden never benefited anyone. Tecno’s Camon 50 Ultra 5G is a great phone, but what makes it worth doing a double take for is EllaClaw, an agentic AI built into the phone that offers a phone AI experience I have yet to see on any smartphone so far. EllaClaw is currently an invite-only closed Beta feature limited to just this phone, but seeing it in action gave me everything I needed to understand where other phones are going wrong with their on-board AI, and where they can learn from Tecno.

The wild part is how approachable it all feels. EllaClaw does not sit around waiting for you to guess the perfect command like some moody genie. It actively shows you what it can do, which makes the whole experience feel far less intimidating than Gemini or Apple Intelligence currently do. Within minutes, I was jumping between message summaries, creative tools, settings help, and different AI models without feeling like I needed a tutorial or a prayer.

Aesthetics

The moment I unboxed the Camon 50 Ultra 5G, I was hit with a serious case of déjà vu. The green variant has a gorgeous satin finish that gives it a silky, metallic glow under the light, and it immediately sent me down a memory lane trip to my old OnePlus 8 Pro which I faithfully used from 2020 to 2024. The color, the texture, even the way the camera module is arranged, it all felt so familiar and, frankly, so premium. Tecno absolutely nailed the flagship-esque feel here. The back is smooth to the touch, resists fingerprints reasonably well, and has a subtle elegance that a lot of phones in this bracket just do not bother with.

The front is where the design language gets a little more complicated, at least for me. Tecno went with a waterfall display, which means the screen curves dramatically around the sides. It looks stunning when the phone is off or when you are watching a video, creating an immersive, almost bezel-free experience. But I have always had a love-hate relationship with these displays. They are beautiful, no doubt, but they can be a magnet for accidental palm touches. While the software does its best to ignore them, I still found myself occasionally tapping something I did not mean to. It is a classic case of striking aesthetics meeting the messy reality of how we hold our phones. I do have to say that Tecno’s Liquid Glass homage does look much better on a waterfall display than on a bezel-ridden iPhone.

Ergonomics

That waterfall display has a big effect on how the phone feels in your hand. The curved edges melt away into the frame, making the Camon 50 Ultra 5G incredibly comfortable to hold. There are no sharp, flat corners digging into your palm, which is great for long sessions of scrolling or gaming. The trade-off, of course, is that the phone can feel a little slippery if you are not paying attention. It has that smooth, polished-stone feel that is lovely to touch but always feels one careless gesture away from a gravity experiment. It is a phone that practically begs to be put in a case, even if it feels like a shame to cover up the design.

The button layout is personally feels a little little bit tricky. On the left, you have a dedicated AI key that summons EllaClaw, a shortcut I ended up using constantly. You can remap it, but having a one-press gateway to the AI is a smart touch. The right side, however, is where my muscle memory started to complain. There is a power button and a fairly short, single-piece volume rocker. Because the rocker is slightly compact, I often found myself fumbling, unsure if my thumb was on volume up, down, or about to lock the screen. It is a small pain, but a tiny indent in the middle of the volume button would have made all the difference. The buttons themselves are slightly sharp and textured, a necessary choice for a phone with curved edges, but it does not quite solve the problem of telling them apart by feel alone.

Performance

EllaClaw is where the Camon 50 Ultra 5G earns its stripes. The dashboard (which you access by swiping towards the left – where you’d see Google Discover on most Android phones) keeps surfacing prompts, shortcuts, and suggestions, making the AI feel alive and accessible rather than buried under three layers of settings. Most phone AI treats users like they should already know what they want. EllaClaw meets you halfway, and that changes everything about how you interact with the phone.

The deeper layer, EllaClaw Pro, is genuinely impressive. It reaches into the phone’s apps and system functions in a way that feels ahead of where the conversation currently is. I asked it to summarize my text messages without opening the messages app, and it did exactly that, handling the whole thing on its own. That kind of agentic behavior is the next logical step for phone assistants, and right now, Tecno seems to be ahead of the pack. Nothing reassured me that the AI wasn’t reading my messages to enhance its database though, so the privacy conscious might have a thing or two to worry about.

Tecno’s decision to build EllaClaw as a multi-model hub is one of the smartest software calls I have seen this year. My unit offered access to DeepSeek, Gemini, ChatGPT, Grok, and Meta AI alongside Ella’s own model, all in one dashboard. Claude is conspicuously missing, but everything else is here and free to use (at least during the Beta period). Daily limits are applied on AI use which reset every 24 hours, and you can buy extra credits in case you overshoot. In all my hours of testing, I never needed to purchase extra credits, which means image generation, video generation, and sketch-to-photo tools were all available at no additional cost.

EllaClaw is still being refined, though. The dashboard lagged occasionally, and the AI sometimes took longer to load than felt comfortable. Restarting the app always sorted it out, but it is a reminder that this experience has not fully grown into its ambition yet. Users in certain regions may also need a VPN to get the full experience running, which is worth knowing upfront.

The camera system is built around a 50MP Sony LYT-700C main sensor with OIS, paired with a 50MP telephoto lens offering 3x optical zoom (70mm equivalent) and an 8MP ultrawide. That telephoto is a proper standout for a phone at this price point, and the software leans into it with confidence. You get 60x hybrid zoom, which is mostly digital exaggeration past the 3x mark, but the results between 3x and 10x are still usable in decent light. The main camera performs well across a range of scenarios, especially in good light, and night mode is aggressive without getting too soupy. The front camera is a 50MP sensor with autofocus, which makes a noticeable difference in selfie sharpness compared to the fixed-focus cameras on cheaper phones.

Tecno stuffs the camera app with modes including dual-cam, underwater mode, document scanning, super night mode, and super macro. FlashSnap is the standout, though. It watches your scene and captures the most dramatic moment automatically, so if someone is mid-jump or mid-throw, the phone clicks at exactly the right time without you having to think about it. It made me look like a considerably better photographer than I actually am. Video recording tops out at 4K at 30fps, which is fine but not particularly exciting.

Battery life is one of the phone’s most quietly satisfying qualities. The 6500mAh cell pushed comfortably through two full days of use without needing a charge, which is more than my current flagship iPhone manages. At some point I just stopped thinking about charging altogether, which is the best thing a battery can do. The phone supports 45W wired charging, which is fast enough to feel civilized without being record-breaking.

The display is a 6.78-inch AMOLED panel with a 1.5K resolution (1208 x 2644 pixels, 429 ppi) and a 144Hz refresh rate with LTPO support for adaptive refresh. It is vibrant, sharp, and performs well in low light. The 144Hz refresh makes scrolling and gaming feel smooth, and the high PWM dimming (2160Hz) helps reduce flicker for those sensitive to it. The curved waterfall edges are a design choice I have already aired my reservations about, and this section will not relitigate that argument. It looks great. It catches accidental touches. Moving on.

Speakers are a weak spot, and a noticeable one. There is a single grille at the bottom right, so stereo is off the table, and the audio leans tinny with very little bass to speak of. For a phone this polished everywhere else, it feels like a department that got the budget cut.

Gaming held up well for everything I tried. Tecno’s dedicated gaming software manages frame rates and battery draw sensibly, and the Dimensity 7400 handled casual to moderate sessions without complaint. The 144Hz display helps make fast-paced games feel responsive, and the phone stayed well within comfortable temperature ranges throughout. Nothing I played caused any real issues, thermal throttling never became a concern, and the large battery meant I could game for extended sessions without nervously watching the percentage drop.

Sustainability

Tecno is promising 3 years of Android OS upgrades and 5 years of security patches for the Camon 50 Ultra 5G, which puts it roughly in line with brands like Motorola and slightly ahead of OnePlus, though still behind Nothing’s 6-year security support.That is a meaningful commitment for a phone at this price, especially if you are someone who plans to use the device well beyond the usual two-year upgrade cycle.

Tecno also claims the battery is built for durability, marketing it as a “5-Year Durability Battery” with TÜV certification and a 72-month lag-free guarantee. Whether that holds true in practice is something only time will tell, but the fact that the company is willing to stake that claim suggests they are at least thinking about longevity beyond the usual marketing fluff. The 6500mAh capacity itself helps, because the less frequently you need to charge, the slower the battery degrades. If you are getting two full days of use out of a charge right now, that headroom means the phone should still feel usable years down the line when the cell inevitably loses some of its capacity.

Value

The Tecno Camon 50 Ultra 5G is priced at $420 for the 8GB + 256GB variant (that’s based on its ₹40,000 Indian price, since the phone hasn’t officially launched in the US). For a phone with a 50MP telephoto, a 144Hz AMOLED display, and a 6500mAh battery, that positioning feels competitive in the under-40K segment, especially when you factor in the AI ambitions.

What feels less competitive is the setup experience. When I powered on the phone and began the initial configuration, India was not listed in the countries menu. Hilariously enough, Palestine was available as an option. The world’s most populated country, nowhere on the list. I eventually had to select a different region just to proceed, which immediately felt awkward for a phone launching in India. Later, I also needed a VPN to access EllaClaw, which could have been a software issue tied to my initial region selection, or it could point to broader availability constraints. Obviously this won’t be the case for everyone since mine was a review unit with the EllaClaw beta program probably tied to a certain region. Either way, it is a frustrating first impression for a feature that is supposed to be the phone’s main selling point.

As for EllaClaw Pro and its credits system, Tecno does offer a credit-based structure for premium AI features like image generation and video creation. In my testing, I never hit a paywall or needed to purchase add-on credits, which suggests the included allocation is generous enough for typical use, at least initially. How sustainable that remains over time, or whether Tecno tightens the limits later, is harder to predict. For now, though, the experience feels surprisingly open, which is a rare thing in the AI phone space.

Verdict

The Tecno Camon 50 Ultra 5G delivers a lot of value for $420. The camera is genuinely impressive, with super night mode shots and 60x zoom that feel bonkers for this price range, and the battery life makes it easy to forget about charging anxiety for days at a time. But I am not recommending this phone for its camera, or its battery, or even its design, as much as I enjoyed all of those things. I am recommending it for EllaClaw, because it is the closest I have come to using a phone AI that feels like it was built for actual humans.

EllaClaw is everything Gemini and Apple Intelligence hope to achieve, and a bit more. I did not think I would see a phone that lets me choose the LLM for its AI assistant, swapping between models depending on what I need. I definitely did not expect an AI assistant dashboard to genuinely give me so much insight and value into how I can use the AI, rather than just sitting through a keynote, nodding along to feature lists, and then forgetting half of them the moment I leave the room. Tecno has cracked something here that bigger brands are still struggling with, which is making AI feel approachable, integrated, and genuinely useful instead of intimidating or buried.

The phone has a few tiny flaws that feel manageable if you’re more driven by the AI chops. The speakers are tinny, the curved display catches palm touches, EllaClaw occasionally lags, and the setup process was a bit of a hassle given I needed a VPN to get EllaClaw to work (but that’s just my China-bound review unit). But if you are looking for a phone that shows you what AI on mobile should feel like when it grows up, the Camon 50 Ultra 5G is worth your attention. It is not perfect, but it is ambitious in all the right ways. Google, Apple, and Samsung could surely take notes.

The post Tecno Camon 50 Ultra 5G Review: A $420 AI Phone With Built-in Gemini, ChatGPT, DeepSeek, and Grok first appeared on Yanko Design.