Apple reportedly removed several high-memory Mac Studio and Mac mini options as AI demand and memory shortages strain desktop Mac supply.

The post Mac Studio, Mac mini Buyers Are Losing Options Amid AI Demand appeared first on TechRepublic.

Alphabet is closing in on Nvidia’s market value as Google Cloud growth, AI investments, and custom chips fuel Wall Street optimism.

The post Alphabet Poised to Overtake Nvidia as the World’s Most Valuable Public Company appeared first on TechRepublic.

New Texas filings suggest Elon Musk’s proposed Terafab chip plant could cost up to $119 billion, raising stakes for AI and semiconductor supply chains.

The post Elon Musk’s Texas Chip Plant Could Cost $119B, Filings Show appeared first on TechRepublic.

Android 17 rumors point to Motion Assist, App Bubbles, native app locking, Gemini updates, and Android XR news ahead of Google I/O 2026.

The post Android 17: Everything We Know About Google’s Biggest Year Yet appeared first on TechRepublic.

Headphones and earbuds have, over the last couple of years, become staples of this fast-paced world for good reason. The little audio gear essentials can do a multitude of tasks with just the push of a tactile button or pinch gesture. They can trigger smart assistant support for a smarter you, and if this concept were to be imagined, they can give you a pair of smart eyes, too.

The idea of a pair of earbuds with integrated cameras is not new, as Emil Lukas imagined, and now another concept reinforces the merit in having a pair of lenses on earbuds. Dubbed Lightwear, the earbuds look something straight out of a sci-fi flick, but underneath, they are a pair of smart assistant earphones that enhance your environmental perception in real time.

Designer: Suosi Design

Touted as the world’s first AI-powered earphones, Lightwear comes with a set of HD camera lenses to expand the sense perception as a vision module to interpret the surroundings and deliver the desired result. One can detail them via voice command about any information required in real time, and the buds respond with a detailed set of instructions or navigation guidance. Having gesture control support, the buds can control the connected home devices remotely using just gesture commands. All the data fed into the smart data system is end-to-end encrypted and stored locally. For enhanced privacy and protection, the sensitive data is automatically cleared on a scheduled cycle.

Compared to Emil’s version, these earbuds have a very downplayed camera presence, which I prefer. They look and feel just like any normal earbud, but have a function that makes them stand out from most pairs of earbuds that have the predictable features. Unlike other AI-powered earbuds, these stand true to their name as they come with the added visual apparatus to put forth better results. The use of AI functions is not limited to the earbuds, since the charging case does the same. This removes the use-case scenario to just when the earbuds are being worn. Loaded with highly sensitive microphones, the AI features can be triggered anytime the user wants. Privacy is also taken care of, as the user can opt to activate the fingerprint unlock module to prevent any unauthorized use.

These have an over-the-ear design, reminiscent of the way IEMs sit flush on the ears. The battery resides in those lobes, and although the designers don’t specifically talk about the usage time, these should last longer than TWS earbuds. Nor is there a specific word about the sound quality, ANC levels, or the app features. But then it’s just a concept centred on the form factor and usability.

The post AI-powered earbuds with built-in camera expand your capabilities in the real world first appeared on Yanko Design.

A new phishing kit named Bluekit offers more than 40 templates targeting popular services and includes basic AI features for generating campaign drafts. [...]

We keep building AI to do more. More answers, more speed, more certainty. Designer Mehrnaz Amouei looked at that trajectory and asked a fundamentally different question: what if we built AI to be more present instead? The result is POCO, a soft robotic companion that might be one of the most quietly radical design concepts to emerge in recent years. It doesn’t talk over you, doesn’t flood you with information, and it doesn’t pretend to know things it doesn’t know. POCO sits with you. Literally.

At its core, POCO is a soft, tactile object that pairs with a smartphone, which serves as its computational brain and face. A soft textile body wraps around the device, transforming rigid, glass-and-metal technology into something that moves, breathes, and gestures in response to your presence. Together, they create something that sits somewhere between object, creature, and companion, and that deliberate ambiguity is very much intentional. You’re not quite sure what to call it, and that’s entirely the point.

Designer: Mehrnaz Amouei

Amouei developed POCO through research at the University of Illinois at Chicago, grounding the project in studies on loneliness and trust. Her findings indicated that people don’t actually want AI that projects certainty or control. They want availability and responsiveness. They want something that shows up without taking over. From those findings came the concept of “constructive interdependence,” a design philosophy where POCO’s limitations aren’t bugs to be patched but features embedded directly into the interaction model itself. The robot communicates what it can and cannot do through its behavior and physical states, which is a level of honesty you don’t often get from technology that typically overpromises and underdelivers.

I think that matters more than it might initially seem. The dominant conversation around AI right now is almost entirely about expansion: more capability, more integration, more autonomy. POCO pushes back on that without being preachy about it. It reframes the question of what good AI design actually looks like, and the answer it offers isn’t “smarter,” it’s “more trustworthy.” That is a genuinely different value system, and it feels overdue.

The sustainability dimension is also worth paying attention to. Rather than introducing new hardware and generating more electronic waste, POCO repurposes a device most people already own. That decision isn’t just a nice bonus; it’s built into the concept from the start, aligning with the UN’s Sustainable Development Goals around mental well-being and responsible consumption. In product design terms, that means the project was developed with a broader cultural and environmental context in mind, not just a user persona sitting in a lab.

Physically, POCO responds to touch, movement, and environmental cues. It adapts to a user’s preferences while maintaining a consistent identity, which is a surprisingly nuanced balance to strike in any product, let alone one sitting at the intersection of soft robotics and emotional design. Because interaction happens through touch rather than voice commands or screen taps, there’s an intentional slowing down embedded in the experience. You can’t rush a tactile exchange the same way you can type faster or speak louder. That shift from speed to presence feels like a meaningful counter-proposal to how most tech is currently designed. We’ve grown so accustomed to interfaces that demand our attention that a device asking only for our company reads almost as radical.

POCO has already earned an Honorable Mention from the International Design Awards and drawn coverage from major design publications. Whether it ever moves into consumer production remains an open question. But as a design statement, it’s doing exactly what the best concept work should: prompting us to reconsider what we actually want from the technology we live with, and whether expanding capability was ever really the right goal. Maybe the most interesting AI isn’t the one that knows the most. Maybe it’s the one that knows when to just stay close.

The post Forget Smarter AI, This Robot Thinks Presence Is the Point first appeared on Yanko Design.

The ChatGPT maker indicated that its new partnership with Amazon will help enhance its enterprise market share while highlighting key constraints in its Microsoft tie-up.

OpenAI’s Codex gets a memory upgrade, improving developer workflows by boosting effectiveness and efficiency.

Security researchers discovered a remote code execution (RCE) vulnerability in Apache ActiveMQ Classic that has gone undetected for 13 years and could be exploited to execute arbitrary commands. [...]

OpenAI CEO Sam Altman recently admitted that ChatGPT doesn't have the capability to keep track of time, but the chatbot insists it does.

Harness multiple top-tier models like GPT‑4o, Claude, Gemini, and more in one unified platform, now $55.

The post Stop Juggling AI Tools — This Lifetime Deal Puts GPT‑4o and More in One Place appeared first on TechRepublic.

Every time you type a prompt into ChatGPT, something happens somewhere far away. Servers spin up. Electricity moves. Carbon gets generated. The whole transaction is so clean and invisible on your end that it might as well not be happening. That’s by design, and it’s worth thinking about. Although with the way we use technology these days, we seldom think about the consequences on our environment.

London-based creative studio Oio wants to change that, starting with a small 3D-printed box and a bright yellow pinwheel. Their project, the Hot Air Factory, is a domestic AI device that processes your questions and requests locally, without connecting to the cloud, and every time it thinks, it physically exhales. Hot air pushes out of the top of the device and spins that cheerful little pinwheel. The harder it thinks, the faster it spins. You’re watching computation happen in real time, which turns out to be a surprisingly powerful thing.

Designer: Oio

The concept is simple: make the invisible visible. We know AI uses energy. We’ve read the headlines. But knowing abstractly that data centers are energy-hungry is different from watching a pinwheel turn every time you ask your AI assistant to summarize something. One is a statistic. The other is a moment of honest accountability.

What makes the Hot Air Factory smart, beyond its obvious design appeal, is how it translates cost into human-readable terms. It doesn’t give you kilowatt-hours because most people have no idea what that means. Instead, it tells you something like “that prompt cost the equivalent of brewing a cup of tea” or “watching Netflix for five minutes.” Suddenly the math becomes personal. Suddenly you start wondering whether you really needed a 500-word AI response to a question you could have Googled.

Oio co-founder Matteo Loglio describes it as “a small, domestic AI that reveals the hidden energy cost behind every prompt.” The factory also lets you dial up or down the level of intelligence it uses. Want a quick answer? Use a lighter model, spend less energy. Need something more complex? Crank it up, and watch that pinwheel work for it. You can even schedule your heavier prompts for the night shift, when energy is cleaner and the grid is quieter. These are design decisions that carry real ethical weight, and they’re baked in with zero condescension.

The playfulness and the seriousness aren’t in conflict here. They’re exactly the point. The Hot Air Factory is built in a Frutiger Aero visual language, all soft curves and clean optimism, the kind of aesthetic that makes you want to put it on a shelf next to your plants. But underneath that approachable exterior is a genuinely complicated machine running open-source large language models on a local GPU. It looks like something a friendly robot would carry. It functions like a small act of protest.

AI companies have very little incentive to make their energy costs legible to users. Invisibility is convenient. It keeps things frictionless. It keeps you prompting without thinking about the bill. A report from the US Department of Energy projected that by 2028, data centers could account for 12% of total electricity consumed in the US. That’s not a small number, and it keeps growing every time we treat AI like it runs on good intentions and cloud magic.

The Hot Air Factory isn’t saying AI is bad. It isn’t demanding you stop using it. What it’s doing is quieter and more persuasive than that. It’s asking you to look. To see. To feel, just a little, what your digital habits cost in the physical world. That’s the argument made not through a lecture or a campaign, but through a yellow pinwheel spinning in your living room.

Design can do that. Sometimes a small, well-made object says more than a policy paper ever could. The Hot Air Factory is currently looking for collaborators to help bring it to a wider audience, still working its way from experiment to something anyone can own. If the goal is conscious computing, the first step might just be this: a tiny box, a spinning fan, and the quiet discomfort of watching a machine breathe.

The post A Tiny Pinwheel Is Doing What AI Giants Won’t first appeared on Yanko Design.

Lenovo and HP have both devised on-device AI assistants designed to make your digital life easier. I dug into the features to discover similarities and differences.

Microsoft is changing how OneDrive handles deleted files. Soon, cloud deletions won't hit your local Recycle Bin, forcing you to use the web for recovery.

An AI expert mapped more than 80 Copilot products, sparking debate over whether it’s branding brilliance or confusion.

Copilot's terms of use indicated that it's for entertainment purposes only. However, Microsoft has indicated that the phrasing is a legacy language from when Copilot originally launched as a search companion service in Bing.

Hackers are exploiting a maximum-severity vulnerability, tracked as CVE-2025-59528, in the open-source platform Flowise for building custom LLM apps and agentic systems to execute arbitrary code. [...]

Microsoft MVP Lance McCarthy highlights how easy it actually is to add AI to an app using Microsoft's Windows AI APIs during development. I want to see more of this type of AI and less of the unecessary bloat that's giving it a bad name.

Your AI PC's Neural Processing Unit (NPU) is a useful piece of hardware, but are you making the most out of it? I dug up 7 apps that make good use of the NPU; which ones are you using?