Home Blog Page 7

Aliro 1.0 Is Here And It’s Killing the Key Card

Liam Hayes
-
0
Aliro 1.0 Is Here And It’s Killing the Key Card

TLDR

  • The Connectivity Standards Alliance has officially launched Aliro 1.0. An open standard that turns your phone or smartwatch into a universal digital key.
  • Apple, Google and Samsung all back it with credentials stored natively in Apple Wallet, Google Wallet and Samsung Wallet — no app or internet required.
  • Certified smart locks and enterprise hardware are expected to hit shelves later in 2026, signaling the end of proprietary key card systems.

Your phone already pays for your coffee, boards your flight and stores your ID. So why are you still fumbling with a plastic key card to get into your office? Aliro 1.0 is the industry’s answer and it’s a big one.

On February 25, 2026, the Connectivity Standards Alliance (CSA). The same body behind the Matter smart home standard officially released the Aliro 1.0 specification. An open, interoperable protocol designed to make your smartphone or wearable work as a universal digital key across virtually any door: homes, offices, hotels, campuses, and government buildings.

What Exactly Is Aliro?

Aliro 1.0
image source – aliro official

Aliro is an industry-wide communication and credential standard for digital access control. In plain English: it’s the technology that allows your iPhone, Android phone, or Galaxy Watch to unlock compatible doors regardless of the brand of lock or the building you’re entering.

Before Aliro, digital key systems were a fragmented mess. Every lock manufacturer had its own app, its own cloud system and its own credentials. Aliro replaces all of that with one universal layer similar to what Matter did for smart home devices like lights, thermostats and plugs.

Aliro is solving the fragmentation that has held back digital key adoption, said Tobin Richardson, President and CEO of the CSA. By connecting the access control industry directly to leading mobile wallet ecosystems. It delivers a secure, frictionless experience that goes well beyond the front door.

How Does Aliro 1.0 Actually Work?

Aliro 1.0
image source- official aliro

Aliro supports three wireless technologies, each serving a different use case:

  • NFC — Tap your phone to the reader for instant access, just like a contactless payment
  • Bluetooth Low Energy (BLE) — Initiate access from a few feet away without touching anything
  • BLE + Ultra-Wideband (UWB) — Fully hands-free; the door unlocks as you walk toward it

Your digital key is stored directly inside your device’s native wallet Apple Wallet, Google Wallet, or Samsung Wallet. Meaning it works completely offline. No dedicated app. No cloud dependency. No internet required.

Security is handled through asymmetric cryptography, ensuring that communication between your device and any Aliro-certified reader is private, tamper-resistant and verified.

Who Is Behind Aliro?

This isn’t a startup project it’s a coalition effort. Seven founding companies pooled their core intellectual property to build the spec: Apple, ASSA ABLOY, Google, Infineon Technologies, Last Lock, Samsung and STMicroelectronics. Over 220 CSA member companies contributed to its development.

Hardware makers already lined up for Aliro 1.0 certification include Allegion, Aqara, HID, Kastle, Kwikset, Nuki Home Solutions, NXP Semiconductors, Nordic Semiconductor, and Qorvo covering everything from residential smart locks to enterprise-grade access readers. Products like the Aqara Smart Lock U300 and Nuki Smart Lock Pro have already launched with Aliro compatibility in mind.

What Does This Mean for You?

For homeowners, it means your smart lock can finally work natively with your phone’s wallet no proprietary hub required.

For enterprise IT and facilities managers, it means the end of expensive, vendor-locked badge systems. No more replacing every reader when you switch providers.

For travelers and hotel guests, it brings one-step mobile check-in and room access without downloading a hotel app.

The first wave of Aliro-certified hardware is expected to reach consumers and businesses later in 2026. With platform support from Apple, Google, and Samsung rolling out in the coming months.

What’s Coming Next?

Aliro 1.0 is built to evolve. The CSA has confirmed that future versions will tackle secure key sharing letting you digitally grant someone temporary or permanent access. Along with additional use cases and full backward compatibility with 1.0-certified hardware. Within the smart home world, Aliro integrates through Matter. which has already incorporated Aliro provisions in its latest specification.

The certification program is now open, and the industry momentum behind Aliro is unlike anything the access control space has seen before. If the Matter parallel holds, widespread adoption could come faster than most expect.

Physical keys had a good run. Aliro is what replaces them.

You might be interested in following article

The Anti-SMS Manifesto: Why I Switched to Hardware Keys

Sources

Samsung Adds Perplexity AI to Galaxy And It’s a Big Deal 2026

Maya Kapoor
-
0
Samsung Adds Perplexity AI to Galaxy And It’s a Big Deal 2026

TLDR: Samsung officially added Perplexity AI as a built-in agent on Galaxy S26 devices. You can activate it with Hey Plex or by holding the side button. It works across Samsung apps, powers Bixby’s web search and sits alongside Gemini and Bixby giving you real AI choice on one device.

So What Actually Happened?

At Galaxy Unpacked on February 25 in San Francisco, Samsung confirmed something that had been rumored for months. Perplexity AI is now a core part of the Galaxy experience. Not just a downloaded app, not a shortcut. An actual system-level agent built right into One UI 8.5.

If you’ve never used Perplexity before, think of it like Google Search but smarter, faster and it actually gives you real answers instead of a wall of links.

How Do You Use It?

It’s pretty straightforward. Just say Hey Plex or press and hold the side button on your Galaxy S26. That’s it. You’re in.

From there, Perplexity can help you out directly inside apps you already use every day:

  • Samsung Notes — ask questions while you’re writing
  • Gallery — search or get context about your photos
  • Calendar & Reminder — get smart suggestions in real time
  • Clock — yes, even there

No switching apps. No copy-pasting. It just works where you already are.

Wait — What About Bixby and Gemini?

Perplexity AI to Galaxy
image source- samsung.com

They’re still there, and that’s kind of the point. Samsung isn’t trying to replace anyone. Instead, Galaxy AI acts like a smart coordinator. It figures out which assistant is best for the job and sends your request there.

  • Bixby handles your phone settings and everyday device tasks
  • Gemini digs into complex reasoning and productivity
  • Perplexity jumps in when you need real-time answers from the web

Samsung said nearly 80% of users already use more than two AI tools daily. This setup finally makes that feel seamless instead of chaotic.

The Bixby Glow-Up Nobody Talked About

Here’s the part that didn’t get enough attention. Perplexity now powers web search inside Bixby too. So even if you never say Hey Plex a single time, Bixby’s search results are going to be way better than before.

For anyone who gave up on Bixby years ago. This might actually be the reason to give it another shot.

Why Samsung Did This

This isn’t just a cool feature drop. It’s a strategy. Apple is still betting big on Siri doing everything in-house. Samsung is going the opposite direction build an open ecosystem and let the best AI tools win.

With Galaxy AI features set to hit 800 million cumulative devices this year. The Perplexity integration is one of the biggest AI rollouts in consumer tech, full stop.

Should You Care about Perplexity AI to Galaxy

If you’re on a Galaxy S26 absolutely. If you’re on an older Galaxy flagship. Samsung hinted that updates are coming, so keep an eye out for One UI 8.5 rolling out to your device.

Either way, your Galaxy is about to feel like a completely different phone. And for once, the AI hype actually seems worth it.

You might be interested in following article

Perplexity Computer Is Here And It Just Changed What AI Can Do

FAQ

What is Perplexity AI on Samsung Galaxy S26?

Perplexity AI is a system-level AI agent built into the Galaxy S26. It delivers real-time web answers directly inside Samsung apps like Notes, Gallery, and Calendar no separate app needed

How do you activate Perplexity on Galaxy S26?

Say Hey Plex or press and hold the side button on your Galaxy S26. Perplexity will launch instantly inside whatever app you’re already using

Does Perplexity AI work on older Samsung Galaxy phones?

Currently, Perplexity AI integration launches with the Galaxy S26 series. Samsung has hinted that older Galaxy flagships may receive the feature through future One UI 8.5 updates.

Perplexity Computer Is Here And It Just Changed What AI Can Do

Sophia Lin
-
0
Perplexity Computer Is Here And It Just Changed What AI Can Do

TLDR:

  • Perplexity just launched Perplexity Computer. A cloud-based AI system that runs 19 models simultaneously, each handling a different part of your project.
  • It comes with a built-in file system, real-time browser, CLI tools and personal connectors, making it capable of managing entire workflows end-to-end.
  • Currently available to Max subscribers at $200/month with Pro access rolling out after load testing is complete.

I’ll be honest. When I first heard Perplexity Computer I thought it was just a rebranded chatbot with a fancy name.

It’s not.

What Perplexity launched this week is something genuinely different. An AI system that doesn’t just answer your questions. It runs things for you. We’re talking research, coding, writing, deploying all handled by 19 AI models working together behind the scenes, each doing what it does best.

Think of it like a full team working on your project. One person reasons through the problem. Another writes the code. Another handles the content. Perplexity Computer is that team minus the Slack messages and missed deadlines.

So What Is Perplexity Computer?

At its core, Perplexity Computer is a cloud-based AI system that orchestrates multiple AI models simultaneously to complete complex, multi-step tasks.

CEO Aravind Srinivas put it well when he quoted Steve Jobs: Musicians play their instruments. I play the orchestra. That’s exactly what this product does it plays the orchestra of 19 AI models so you don’t have to.

Here’s what it has access to:

  • A file system to store and manage your project files
  • CLI tools to run commands and automate tasks
  • A real-time browser to research and pull live information
  • Personal connectors to plug in your own tools and data

Put it all together and you’ve got an AI that can take a project from idea to finished product without you babysitting every step.

How the 19-Model System Actually Works

Perplexity Computer
image source- perplexity

Here’s the part that makes Perplexity Computer stand out from everything else out there.

Instead of relying on one AI model to do everything, Perplexity routes each part of your task to the model best suited for it. One model handles reasoning. Another handles code. Another handles writing. The system manages what they call sophisticated token management. Basically making sure the right model gets the right job without wasting resources.

FeatureWhat It Does
Multi-model orchestration19 AI models running in parallel
File system + CLIManages full project workflows end-to-end
Real-time browserResearches live data while working
Personal connectorsIntegrates your own tools and data
Usage-based pricingYou only pay for what you actually use

It’s less like using an AI tool and more like delegating to an AI team.

Pricing — Who Can Use It Right Now?

Right now, Perplexity Computer is available exclusively to Max subscribers, which runs at $200/month.

Pro subscribers are next in line Srinivas mentioned they’ll get access once the team is satisfied with load testing. The product is web-only at launch and requires you to sign in.

On pricing, Srinivas made an interesting point: usage-based pricing is the right business model for AI instead of ads. Meaning you pay based on how much you actually use it not a flat fee regardless of activity. That’s a shift worth paying attention to if you use AI tools heavily.

Who Is This Actually For?

Let’s be real — $200/month isn’t for everyone. But if you fall into any of these categories, it’s worth a serious look:

  • Developers who want to build and deploy without juggling 10 different tools
  • Content creators and bloggers managing research, drafts and publishing pipelines
  • Entrepreneurs and solopreneurs running lean without a full team
  • Freelancers handling complex, multi-step client projects

If your work involves switching between tools constantly. Perplexity Computer is built to collapse all of that into one place.

The Bigger Picture

Perplexity Computer didn’t launch in a vacuum. The company has been on a serious run lately upgrading Deep Research, launching a Model Council feature, and confirming that its Comet AI browser is hitting iPhone in March.

On top of that, Perplexity’s AI assistant is being baked directly into Samsung’s Galaxy S26 lineup, activated by saying “Hey Plex.” That’s a massive distribution play.

The message from Perplexity is clear: they’re not building a search engine anymore. They’re building an operating system powered by AI.

You might be interested in following article

Genspark vs Perplexity: Why Search Agents Are Better Than Answer Engines

FAQ

What is Perplexity Computer?

It’s a cloud-based AI system from Perplexity that uses 19 AI models working together to handle entire projects research, coding, writing, and deployment autonomously.

How many AI models does it use?

It orchestrates 19 models simultaneously, with each model assigned to the task it performs best.

How much does Perplexity Computer cost?

It’s currently available to Max plan subscribers at $200/month, with Pro access coming soon.

Is Perplexity Computer available on mobile?

Not yet — it’s web-only at launch and requires a sign-in to access.

Apple Is Building AI Wearables And It’s More Than Just a Pin

Maya Kapoor
-
0
Apple Is Building AI Wearables And It’s More Than Just a Pin

TLDR:

  • Apple is working on three AI wearables: an AI pin, smart glasses and camera AirPods
  • The AI pin is AirTag-sized with two cameras, three mics and always-on Siri
  • Smart glasses are targeting a 2027 launch with voice AI and live translation
  • Camera AirPods could arrive as early as 2026
  • Apple is racing against OpenAI and Meta in the AI hardware space
  • All devices will connect to iPhone and power a smarter, more visual Siri

Apple is quietly working on something big. According to recent reports, the company is developing not one but three AI-powered wearable devices. A smart pin, smart glasses and camera-equipped AirPods. If things go to plan, some of these could be in your hands as early as 2026 or 2027.

Here’s everything you need to know.

The AI Pin:

AI Wearables
image source- humane ai pin

Let’s start with the one that got everyone talking.

Apple is reportedly developing a thin, circular disc about the size of an AirTag made of aluminum and glass. That you can clip to your shirt, bag, or even wear as a pendant around your neck.

The specs are surprisingly detailed for something still in early development:

  • Two cameras — a standard lens and a wide-angle lens, always scanning your surroundings
  • Three microphones to pick up audio around you
  • A speaker for back-and-forth Siri conversations
  • A dedicated low-power chip, similar to what’s inside AirPods with your iPhone handling the heavy lifting
  • Magnetic inductive charging, similar to Apple Watch

The whole idea is for this pin to act as the eyes and ears of your iPhone. Feeding Siri real-world context so it can actually be helpful in everyday situations.

One important note though: development is still in early stages and Apple could cancel this product if it doesn’t meet their standards.

Why Apple Is Moving So Fast

Apple isn’t doing this just for fun. There’s real competitive pressure pushing things along.

OpenAI is reportedly launching its first AI wearable device in 2026. A collaboration with legendary designer Jony Ive. Meta already has its Ray-Ban smart glasses on the market and they’re selling well. Apple clearly doesn’t want to be left behind in the AI hardware race.

There’s also the cautionary tale of Humane’s AI Pin hanging over this whole category. Humane a startup founded by two former Apple employees launched their own AI pin back in 2024 and it flopped badly. It sold fewer than 10,000 units, faced harsh criticism for slow performance and poor battery life and the company was eventually sold to HP for just $116 million.

Apple is betting they can succeed where Humane couldn’t by tying the pin deeply into the iPhone ecosystem and using a far more powerful AI backend.

Apple Smart Glasses: The Real Prize

AI Wearables
image source- freepik.com

The AI pin is interesting, but Apple’s ai smart glasses might be the most exciting product of the three.

Apple has already handed out prototypes to its hardware engineering team and is targeting a 2027 launch, with production potentially kicking off as early as late 2026.

These glasses won’t have AR screens in the lenses at least not yet. But they’ll still pack some useful features:

  • Two cameras — one high-res for photos and video. One dedicated to giving Siri visual awareness of your environment
  • Voice-based Siri for calls, music, directions and questions
  • Live translation, context-aware reminders and the ability to scan physical text like event flyers and add them straight to your calendar
  • Premium materials with multiple frame sizes and colors. Apple’s clear edge over Meta’s Ray-Bans

Apple considered partnering with an existing eyewear brand but decided to design their own frames entirely in-house. Expect a premium feel and a premium price tag to match.

Camera AirPods Are Coming Too

Of the three products, AirPods with built-in cameras are the furthest along in development and could arrive as early as 2026.

These cameras aren’t for taking selfies. They’re low-resolution sensors designed purely to give Siri passive visual awareness of what’s around you. like your own personal AI assistant that can actually see the world.

The Bigger Picture

All three of these products share one core idea: give Siri real-world vision so it can be genuinely useful not just a voice assistant that answers trivia questions.

Apple is also reportedly building a brand new, more conversational version of Siri for iOS 27 that will serve as the brain powering all of these devices.

This is Apple’s answer to a bold question: what does AI look like when it lives on your body, not just in your pocket?

Whether every product makes it to store shelves or not, one thing is clear. Apple is deadly serious about AI wearables and the next couple of years are going to be very exciting to watch.

Stay tuned to TechGlimmer for the latest updates on Apple’s AI hardware lineup.

Sources

Emergent AI Review 2026: I Tried Building an App Without Writing a Single Line of Code

Kaus
-
0
Emergent AI Review 2026: I Tried Building an App Without Writing a Single Line of Code

TL;DR: Emergent AI is a no-code app builder powered by multiple specialized AI agents. It’s ideal for entrepreneurs, freelancers and creators who want to build and deploy real apps without writing code. It hit $100M ARR in just 8 months, backed by SoftBank and Google. Best for MVPs and lightweight apps not for complex enterprise builds. Free plan available. Worth trying.

I’ll be honest when I first heard about Emergent AI, I was skeptical. Another no-code tool promising to build apps in minutes? I’ve seen that story before. But after spending time actually using it, my perspective shifted. Here’s my real, unfiltered take.

What Is Emergent AI?

Emergent is an AI-powered app builder that turns your plain-English idea into a fully working, deployable application no coding required. It launched in mid-2025 and has already crossed $100M in annual recurring revenue in just 8 months, making it one of the fastest-growing AI startups ever.

It’s backed by heavy hitters like SoftBank, Khosla Ventures, Lightspeed, and Google’s AI Futures Fund. Which tells you this isn’t just hype.

How Emergent AI Actually Works (Real-Life Example)

Let’s say you want to build a simple expense tracker for your small business. You type something like:

“Build me an expense tracker app where I can add expenses by category, see a monthly summary, and export to CSV.”

Emergent’s AI agents immediately get to work:

  • A Planning Agent breaks your idea into structured requirements
  • A Frontend Agent designs the user interface
  • A Backend Agent sets up the database and logic
  • A Testing Agent checks for bugs
  • A Deployment Agent publishes it live

Within minutes, you have a working app with a real URL you can share. No Figma. No GitHub. No developer needed.

I tested this with a basic productivity tool and the output genuinely surprised me. It wasn’t perfect on the first try. I had to refine it with follow-up prompts but the core app was functional and clean. That’s more than I expected.

What is Vibe Coding? How fast AI is Changing the Way We Build Software

Who Is Emergent Built For?

This tool is genuinely useful for:

  • Entrepreneurs who want to validate an app idea without hiring a developer
  • Freelancers who want to build client tools or micro-SaaS products
  • Content creators and bloggers building digital products (lead gen tools, calculators, etc.)
  • Small business owners who need internal tools like dashboards, trackers, or booking forms
  • Students and beginners exploring app development without code

If you’re a developer, you might find it limiting for complex projects. But for MVPs and lightweight apps, it’s powerful.

Emergent vs lovable vs bolt.new vs cursor

Here’s a quick, honest look at how Emergent stacks up against its main rivals:

FeatureEmergentLovableBolt.newCursor
No-code friendly✅ Yes✅ Yes✅ Yes❌ Needs coding
Mobile app builder✅ Yes❌ Limited❌ No❌ No
Multi-agent system✅ Yes❌ No❌ No❌ No
Best forNon-devs, MVPsWeb appsQuick prototypesDevelopers
Free plan✅ Yes✅ Yes✅ Yes✅ Yes

Emergent’s biggest edge is its multi-agent architecture instead of one AI trying to do everything, specialized agents handle each part of the build. The result is noticeably more polished than competitors for non-technical users. Lovable is strong for web apps, but Emergent pulls ahead when you need a full-stack app with a backend and real deployment.

What I Liked

  • The output is surprisingly production-ready for simple apps
  • You can iterate by just typing what you want to change
  • Mobile app deployment is a genuine differentiator
  • The platform has a massive community of 6M+ builders to learn from

What Could Be Better

  • Complex, custom logic still hits limitations
  • Free plan has usage caps. You’ll hit them quickly on larger builds
  • Occasionally requires multiple prompt refinements to get the exact layout you want

Is It Worth It?

If you’re a non-developer with an app idea. yes Emergent is absolutely worth trying. The free plan is enough to test your concept and the paid plans are reasonable for what you get. It won’t replace a full development team for enterprise software. But for MVPs, internal tools and digital products? It’s one of the most capable tools available in 2026.

The fact that it went from zero to $100M ARR in 8 months isn’t just a fun stat. It means millions of real people are building real things with it. That’s the most honest endorsement possible.

Sources

  1. Emergent AI Official Website
  2. Emergent on Y Combinator
  3. Reuters: Emergent Raises $70M from SoftBank & Khosla Ventures
  4. TechCrunch: Emergent Hits $100M ARR in 8 Months
  5. Vibe Coding Tools Comparison 2026

Claude Code to Figma: How Code to Canvas Changes the Design-Dev Workflow Forever

Kaus
-
0
Claude Code to Figma: How Code to Canvas Changes the Design-Dev Workflow Forever

TLDR: Figma partnered with Anthropic to launch Code to Canvas. A feature that lets you take live, running UI built in Claude Code and drop it straight into Figma as fully editable design layers. No screenshots. No copy-paste chaos. Just a clean bridge between the code you write and the design your team collaborates on.

If you’ve ever built something in code, sent a screenshot to a designer and watched that context completely fall apart in translation this one’s for you.

Figma just quietly solved one of the most frustrating parts of the modern dev workflow.

On February 17, 2026, Figma and Anthropic announced a new integration that lets developers send live browser UI directly into Figma as editable design frames. No image exports. No Figma recreations from scratch. The actual rendered UI pulled from your localhost, staging or production environment lands in Figma as real, manipulable layers.

It’s called Code to Canvas and if you work with Claude Code. It’s already available.

The Real Problem This Solves

Here’s the honest reality most articles won’t say: the design-dev handoff has been broken for years.

Developers build something that works. Designers then try to recreate it in Figma from a screenshot or a Loom video. Half the context gets lost. Annotations are made on the wrong version. Engineers go back and change things that were already signed off.

This is the loop everyone has learned to live with. Code to Canvas breaks it.

Instead of exporting a screenshot, you type three words. “Send this to Figma” and your running UI appears in the Figma canvas as structured editable frames. Text stays as text. Buttons stay as buttons. Auto-layout is preserved. Your team can immediately start annotating, comparing and riffing on what’s actually been built not some approximation of it.

How It Works (Step-by-Step)

You don’t need to be a Figma power user to set this up. Here’s what the process looks like in practice:

  1. Install the Figma MCP server — run claude mcp add --transport http figma https://mcp.figma.com/mcp in your terminal
  2. Authenticate — open Claude Code, type /mcp, select the Figma server, and complete OAuth in your browser
  3. Build your UI in Claude Code and preview it in the browser
  4. Type “Send this to Figma” — Claude captures the live rendered state using the generate_figma_design tool
  5. Your UI appears in Figma as fully editable frames — ready for your team to annotate, copy, and iterate on
  6. Round-trip back to code by referencing the updated Figma design via the MCP server in your next Claude Code prompt

The whole setup takes under two minutes if you already have Claude Code running.

What Makes This Different From a Screenshot Tool

Claude Code to Figma
image source- figma.com

A lot of people will read the headline and think “Cool, it’s basically Snagit with extra steps.” It’s not.

The difference is in what the output is. When Code to Canvas captures your browser, it doesn’t flatten everything into a PNG. It reconstructs the UI as native Figma structure editable text, component buttons and auto-layout containers. Designers can duplicate frames, tweak spacing, try alternate color palettes and compare four different layouts side by side all without asking a developer to push a single line of code.

This matters most for multi-screen flows. You can capture an entire five-screen onboarding sequence in one session and Figma preserves the order and context across all of them. Your whole flow lives on one canvas, editable and shareable.

Key Capabilities at a Glance

  • Live UI capture — grab any running browser UI, including dynamic states that are hard to mock manually
  • Fully editable frames — output is native Figma, not screenshots or flattened images
  • MCP-powered — built on the open Model Context Protocol, making it extensible and interoperable
  • Multi-screen capture — capture several screens in one session while maintaining sequence and context
  • Works with Figma Make — whether you start in Claude Code or Figma Make. you can bring previews onto the canvas for iteration

The Honest Limitations

No integration is perfect, and this one has some real friction worth flagging.

The biggest issue is the round-trip. Going from Claude Code → browser → Figma → back to Claude Code involves at least three tools and five context switches. Each handoff risks losing information — particularly anything involving business logic, event handlers, or state management. Figma layers don’t carry your React state. When changes come back to code, the AI re-translates visual decisions from scratch.

A few other things to keep in mind:

  • The generate_figma_design tool is currently Claude Code-only — it doesn’t work with Cursor, VS Code, or Codex yet
  • The Desktop MCP server requires a paid Figma plan with Dev or Full seat
  • Windows/WSL users face additional friction — the Desktop MCP server doesn’t support WSL configurations
  • The get_code tool can struggle when Figma elements have annotations attached

These aren’t dealbreakers — but they’re real, and you’ll run into them.

Why Figma Is Making This Move Now

This launch isn’t just a product update. It’s a strategic response to a genuine existential threat.

With tools like Claude Code, v0 and Cursor letting developers ship polished UIs in hours, the traditional design first, then build pipeline is being skipped entirely by a growing chunk of teams. Figma CEO Dylan Field has been vocal about this — arguing that the canvas isn’t less relevant because AI makes building faster. It’s more relevant because it’s now the place where human judgment, exploration, and collaboration happen.

“Speed without exploration can turn into momentum in the wrong direction. That’s the moment when code needs to meet the canvas.” — Dylan Field, Figma CEO

By embedding Figma directly into the Claude Code workflow, Figma isn’t just adding a feature — it’s repositioning itself as the collaboration layer on top of AI-generated software.

Should You Use Claude Code to Figma

If you’re a developer using Claude Code who collaborates with designers or PMs. yes, absolutely try it. The setup is fast, the output is genuinely useful, and having an editable Figma frame instead of a screenshot will save you real back-and-forth time.

If you’re a designer working with a code-first team — this is worth bringing up in your next standup. Instead of waiting for a mockup that matches what’s actually been built. You can work directly from the live product.

If you’re a solo builder or indie hacker — the round-trip might add more overhead than it’s worth right now. But keeping an eye on how this matures is definitely worth it.

FAQ

Do I need a paid Figma plan to use this?
The Send this to Figma feature via Remote MCP works with free plans. However, the Desktop MCP Server. which supports selection-based workflows requires a Dev or Full seat on a paid Figma plan.

Can I capture an entire app flow, not just one screen?
Yes. Code to Canvas supports multi-screen capture in a single session. preserving the sequence and context so the full user journey stays coherent in Figma.

Does this work with tools other than Claude Code?
Right now, the generate_figma_design tool is Claude Code-only. Other Figma MCP tools work with Cursor, VS Code, and Codex but the Code to Canvas direction is exclusive to Claude Code for now.

What happens to my business logic when I send to Figma?
It doesn’t transfer. Figma captures the visual structure only not state management, event handlers, or component logic. The round-trip back to code means the AI will re-interpret visual changes into implementation.

Is this the same as Figma Make?
No. Figma Make lets you start from a prompt and build a UI inside Figma. Code to Canvas is the reverse. It starts with already-built code and brings it into Figma. Both paths lead to an editable Figma canvas, just from different starting points.

Sources

Is Apple Losing Popularity in 2026? Here’s the Real Answer

Kaali Gohil
-
0
Is Apple Losing Popularity in 2026? Here’s the Real Answer

TLDR — Quick Hits:

  • 📉 Apple faced real criticism: slow AI rollout, outdated Siri, limited customization
  • 🤝 Siri is getting a full Gemini-powered upgrade in 2026
  • 🎬 Apple Creator Studio launched January 28. 10 pro apps for just $12.99/month
  • 📱 iOS 26 brings the biggest iPhone redesign since 2013
  • 🥽 Vision Pro is finally getting serious content and controller support
  • 💰 Apple hit a record $143.8 billion in revenue in Q1 2026 — up 16% year-over-year

Let’s be honest. For the past couple of years, there’s been a real conversation happening online and it’s not entirely unfair. Apple fans started asking the question nobody wanted to say out loud: is Apple falling behind?

The AI boom caught Apple flat-footed. Competitors were shipping features that genuinely wowed people while Apple was still talking about what Siri might do someday. But here’s where the story gets interesting Apple isn’t just course-correcting. It’s quietly building something bigger. Let’s break it down.

Where Apple Was Genuinely Struggling

No sugarcoating here. Apple had some real problems heading into 2025 and early 2026.

Siri was embarrassing. While ChatGPT, Gemini and Claude were handling complex multi-step tasks and holding natural conversations. Siri was still fumbling basic requests. Users noticed. Tech reviewers didn’t hold back. The gap between Apple’s AI assistant and the competition became a running joke and rightfully so.

Customization was almost non-existent. Android users had been personalizing their phones for years. Widgets, custom launchers, icon packs, lock screen flexibility. iPhone users got… slightly different wallpapers. For a premium product the rigidity felt tone-deaf.

Vision Pro launched without enough to do. The hardware was genuinely impressive. The content library? Thin. The controller support? Missing. For a $3,500 device, it felt like Apple shipped the future without the present-day reason to buy it.

China revenue took a hit. Apple’s market share in China fell as domestic brands like Huawei surged. Combined with tariff pressures adding roughly $900 million per quarter in costs. Apple was dealing with real financial headwinds in one of its most important markets.

Why the Narrative Is Shifting Fast

Here’s what the doom-and-gloom crowd is missing. Apple doesn’t panic it pivots. And in 2026, the pivots are significant.

Siri Finally Gets a Brain Powered by Gemini

Is Apple Losing Popularity in 2026
image source- apple.com

This is the biggest AI story Apple has had in years. Apple is partnering with Google to power a fully overhauled Siri with Gemini’s large language models. Which is bringing genuinely conversational, multi-step AI assistance to every iPhone. The upgraded Siri is expected to arrive in spring 2026 and iOS 26.4 is already opening CarPlay to third-party AI chatbots including ChatGPT, Claude, and Gemini. It is giving users real options while the new Siri is finalized.

This isn’t Apple admitting defeat. It’s Apple making the smart business call: instead of spending hundreds of billions building a model from scratch. They’re integrating the best available technology and focusing their energy on privacy, on-device intelligence and user experience. That’s actually a solid strategy.

Apple Creator Studio A $13/Month Gift for Creators

On January 28, 2026, Apple launched Apple Creator Studio. A bundle of 10 professional creative apps for just $12.99 per month.

Here’s what you get in one subscription:

  • Final Cut Pro — professional-grade video editing with new AI features like Transcript Search, Visual Search and Beat Detection
  • Logic Pro — industry-standard music production
  • Pixelmator Pro — powerful image editing for Mac and iPad
  • Motion & Compressor — motion graphics and video export tools
  • MainStage — live performance software for musicians
  • AI-powered tools across Keynote, Pages, and Numbers. Including text-to-image generation powered by OpenAI models

For context: Adobe Creative Cloud costs $60+ per month. Apple just offered a legitimate alternative at $12.99. Students and educators pay just $3/month. For YouTube creators, podcasters and content marketers. This is a genuinely compelling package and it’s only going to grow.

iOS 26 — The Most Customizable iPhone Ever

Is Apple Losing Popularity in 2026
image source- apple.com

iOS 26 is Apple’s biggest visual overhaul since iOS 7 in 2013. The new Liquid Glass design language has been applied across the entire interface and customization options have expanded significantly.

Recent updates through iOS 26.2 and 26.3 added:

  • Expanded Lock Screen clock transparency controls
  • Camera swipe shortcuts directly from the Lock Screen
  • New screen flash notification options for accessibility and style
  • Slide-to-stop gestures with single-tap alternatives

This isn’t just cosmetic. Apple is finally giving users real control over how their phone looks and behaves. Closing the gap with Android in ways that long-time iPhone users have been requesting for years.

Vision Pro Is Actually Getting Good

Is Apple Losing Popularity in 2026
image source- apple.com

visionOS 26 is a serious update for Apple Vision Pro. Key additions include:

  • Spatial widgets that stay anchored in your physical space
  • Dramatically improved Personas — more natural hair, skin and expressions with full side-profile rendering
  • PlayStation VR2 Sense controller support — opening the door for a new category of spatial games
  • 180° and 360° video support from GoPro, Insta360 and Canon
  • Spatial scenes using generative AI to add depth to personal photos

This is the content and hardware ecosystem Vision Pro needed at launch. It’s arriving now and Apple is actively courting developers and enterprises with new spatial APIs.

Sum up on Is Apple Losing Popularity in 2026? with real numbers

Apple’s revenue in Q1 2026 hit a record $143.8 billion. A 16% year-over-year increase driven by strong iPhone 17 sales and a services business that crossed $30 billion in a single quarter for the first time. Net profit came in at $42.1 billion, up 19%. These aren’t the numbers of a company losing relevance.

The question was never really “Is Apple losing popularity?” The better question is “Was Apple moving fast enough?” In 2025, the honest answer was no. In 2026, they’re finally running.

Sources

Sarvam AI Review: India’s Own AI Platform Is More Serious Than You Think

Arjun Patel
-
0
Sarvam AI Review: India’s Own AI Platform Is More Serious Than You Think

TLDR: Sarvam AI is India’s government-backed full-stack sovereign AI platform built specifically for Indian languages, enterprises and government use. It’s not just another AI startup. It was officially selected by the Government of India to build the country’s first homegrown large language model. Its latest launches Sarvam Akshar and Sarvam Edge. show it’s rapidly moving from research into real-world deployment.

Most of the AI conversation in 2026 still revolves around OpenAI, Anthropic, and Google. But quietly and with serious intent. India has been building its own answer and it’s further along than most people outside the country realize.

Sarvam AI is a Bengaluru-based startup that started as a research lab and has grown into what the Indian government has officially designated as the builder of India’s sovereign large language model. That’s not a marketing claim. That’s a government contract and it changes how you should look at this company.

What Sarvam AI Actually Does

At its core, Sarvam is a full-stack AI platform designed to serve India’s specific needs. That means models fluent in Indian languages, built on Indian data deployed on Indian infrastructure and governed under Indian data laws.

Where most global AI tools treat Indian languages as an afterthought a bolt-on translation layer — Sarvam builds from the ground up for languages like Hindi, Tamil, Telugu, Bengali and more. Its models reportedly outperform leading frontier models on Indian language benchmarks. while remaining cost-effective enough for population-scale deployment.

The platform serves three audiences: enterprises, governments and developers. Each getting tailored access to Sarvam’s model stack and infrastructure.

The Sovereign AI Mission: Why This Matters

In April 2025, the Government of India officially selected Sarvam under the IndiaAI Mission to build the country’s first indigenous foundational AI model. The goal isn’t just a chatbot. It’s three model variants built from scratch:

  • Sarvam-Large — for advanced reasoning and generation
  • Sarvam-Small — for real-time interactive applications
  • Sarvam-Edge — for compact, on-device tasks without internet dependency

This makes Sarvam one of the very few AI companies in the world with a direct national mandate. For context, it’s the equivalent of France or Germany commissioning a domestic frontier model rather than licensing GPT-5. The stakes and the expectations are high.

What Launched in February 2026

Sarvam AI
image source- sarvam ai

Sarvam had a big week leading into India AI Summit 2026. Three significant releases dropped within days of each other.

Sarvam Akshar (launched February 15) is a document intelligence workbench built on Sarvam’s Vision model. It handles layout-aware extraction, grounded reasoning and automated proofreading across Indian language documents. Think health reports, insurance forms, prescriptions, academic records all processed with native language accuracy rather than OCR guesswork.

Sarvam Edge (launched February 14) is arguably the more exciting release for everyday impact. It runs AI models directly on smartphones and laptops — fully offline, no cloud required. Key specs worth knowing:

  • Speech recognition across 10 Indian languages in a 74 million parameter model
  • Text-to-speech in 10 languages within just 60MB of storage
  • Time-to-first-token under 300 milliseconds
  • Works with noisy backgrounds, telephony audio and multi-speaker environments
  • Zero per-query cost since everything runs locally

For rural India, small businesses and low-connectivity environments. This is genuinely transformative not just a tech demo.

Sarvam Kaze, unveiled at the India AI Summit is the company’s entry into AI wearables. Smart glasses that PM Modi was photographed wearing. It signals that Sarvam isn’t staying in the cloud-and-API lane.

Real Deployments, Not Just Demos

One thing that separates Sarvam from many AI startups is that it’s already running at scale in production. Tata Capital’s Chief Digital Officer credited Sarvam’s multilingual conversational AI with enabling personalized, product and segment-specific conversations across the customer lifecycle for consumer loan products.

On the government side Sarvam signed MoUs with the governments of Tamil Nadu and Odisha for sovereign AI partnerships. With Tamil Nadu committing ₹10,000 crore to build a full-stack Sovereign AI Park in Chennai.

Sarvam vs. Gemini vs Chatgpt

What You’re ComparingSarvam AIGlobal Platforms (OpenAI, Gemini)
Indian Language SupportNative, benchmark-leadingAdd-on or partial
Data SovereigntyFully within IndiaStored outside India
Government BackingOfficial IndiaAI Mission mandateNone
On-Device AISarvam Edge (offline capable)Limited or unavailable
Target MarketIndia-first, Indic use casesGlobal, English-first
Deployment OptionsCloud, VPC, On-PremisesMostly cloud-only

Who Should Pay Attention to Sarvam

If you’re a developer, enterprise or tech enthusiast in India or covering the Indian AI market Sarvam is not optional reading. It’s central to where India’s AI stack is headed over the next decade.

For global observers, it’s a case study in what sovereign AI actually looks like in practice: not just a policy buzzword, but a full platform with real government contracts, real enterprise customers and real on-device models shipping in 2026.

The company is moving fast. Three major product launches in one week, a national LLM in development, state-level government partnerships and now wearables. Sarvam is building like it has something to prove. Given what’s at stake for India’s AI independence, that urgency makes complete sense.

Sources

  1. Sarvam AI Official Website
  2. Government of India IndiaAI Mission — Sarvam Selection
  3. Sarvam Akshar & Sarvam Edge Launch — News9Live
  4. Sarvam Edge On-Device AI Details — India TV
  5. Tamil Nadu Sovereign AI Park MoU — TechCircle
  6. Sarvam Sovereign AI State Partnerships — Sarvam Blog
  7. Sarvam Kaze AI Glasses — Business Standard
  8. Sarvam AI Full Overview — upGrad Blog
  9. India’s Sovereign LLM Context — Sarvam Blog

Written for TechGlimmer | February 2026 | Category: AI

Google Lyria 3: The AI That Makes Music From a Text Prompt

Sophia Lin
-
0
Google Lyria 3: The AI That Makes Music From a Text Prompt

TL;DR

  • Google Lyria 3 new ai inside the Gemini app. It generates original 30-second songs from text, photos or videos
  • Built by Google DeepMind. It auto-creates lyrics, vocals, melody and even album art
  • Available free on desktop for users 18+, mobile coming soon
  • Every track carries a SynthID watermark to identify AI-generated content
  • Google won’t copy specific artists. It creates something inspired by a mood or genre instead
  • Biggest winners: content creators, podcasters, social media marketers
  • Apple also launched an AI music feature this week. But it only curates playlists. It doesn’t create music

Music used to take years to learn. Now it takes a sentence.

Google just launched Lyria 3 inside its Gemini app and it’s one of the more quietly significant AI releases of 2026. You type a prompt or upload a photo and within seconds. You get a 30-second original song, complete with lyrics, vocals and a full instrumental arrangement. No studio. No instruments. No music degree required.

I’ve been following AI tools closely for a while now and this one feels different. Here’s why it matters.

What Is Google Lyria 3?

Lyria 3 is Google DeepMind’s most advanced music generation model. Now built directly into the Gemini app. It lets anyone not just musicians create original tracks using nothing but a text description, a photo or even a short video clip.

You could type something like: “An upbeat Afrobeat track for a summer road trip.” Gemini processes that prompt and returns a fully produced 30-second song melody, beat, lyrics and all.

What makes Lyria 3 a step up from earlier versions is that you no longer have to write your own lyrics. The model generates them automatically based on your prompt. You also get direct control over style, vocals, tempo and mood. It even generates album artwork alongside the track.

The feature is currently available for free on desktop for users 18 and older, with mobile rollout coming soon.

How It Actually Works

The process is simple:

  • Text prompt — Describe a genre, mood, memory, or vibe
  • Image or video upload — Drop in a photo and Gemini scores a matching soundtrack
  • Style controls — Adjust tempo, vocal style and genre within the prompt
  • Output — A 30-second track with auto-generated lyrics and music

Every track generated by Lyria 3 is embedded with a SynthID watermark. An invisible digital signature developed by Google DeepMind. This means AI-generated music can be identified and traced back to its origin. Which is a big deal for copyright and authenticity reasons.

Google has also been careful about artist imitation. If your prompt includes a specific artist’s name. Gemini won’t copy their voice or style directly. Instead it creates something inspired by that mood or genre. The tool is designed for original expression not reproduction.

How Lyria 3 Could Change the Music Industry

Google Lyria 3
image source- freepik.com

This is where things get interesting and a little complicated.

For content creators, it’s a game-changer. If you run a YouTube channel, podcast or post Reels and TikToks. You know the headache of finding royalty-free background music. Lyria 3 solves that entirely. You describe the vibe you want and you get a custom track that fits your content. No licensing fees, no copyright strikes.

For indie artists and hobbyists, it lowers the barrier to entry. Someone with a creative idea but no production budget can now bring a musical concept to life in seconds. That’s genuinely new.

For the traditional music industry, it raises serious questions. Labels and professional composers have already been watching AI music tools like Suno and Udio with concern. Google bringing this directly into a mainstream app used by hundreds of millions of people accelerates that pressure significantly. The music industry shifted from litigation to licensing partnerships with AI companies recently but the pace of adoption may outrun those agreements.

Google’s integration of Lyria 3 also makes it the first major tech platform to bundle music generation into a general-purpose AI assistant. OpenAI and Anthropic are focused on text and reasoning. Google is building a creative production suite. That strategic difference matters.

Apple’s Approach: A Quick Contrast

Apple also made moves this week with “Playlist Playground” in iOS 26.4, which uses Apple Intelligence to generate 25-song playlists from text prompts. But Apple is curating existing songs, not creating new ones. It’s a discovery tool, not a creation tool. The distinction is significant. Google is playing in a completely different league here.

Who Benefits Most Right Now

  • Content creators making YouTube videos, Shorts, Reels or podcasts.
  • Social media marketers needing quick custom audio for branded content.
  • Bloggers and website owners wanting background music for video content.
  • Casual users who want a fun, personalized way to express a memory or feeling.
  • Small businesses that can’t afford custom music production.

FAQ

Is Google Lyria 3 free?
Yes, currently free inside the Gemini app for users 18 and older.

Can I use Lyria 3 music on YouTube without copyright issues?
Google designed Lyria 3 for original expression and all tracks are SynthID watermarked. However, YouTube’s exact monetization policies for AI-generated audio are still evolving always check current guidelines before publishing.

Will AI replace musicians?
Unlikely in the traditional sense. Lyria 3 is best for short, functional tracks not full albums or emotionally nuanced compositions. Think of it as a tool that helps more people make music, not one that replaces artists entirely.

Is Lyria 3 available on mobile?
Desktop first, with mobile rollout expected in the coming days.

Sources

Gemini 3.1 Pro Just Arrived And Developers Are Already Impressed

Sophia Lin
-
0
Gemini 3.1 Pro Just Arrived And Developers Are Already Impressed

TLDR: Google’s Gemini 3.1 Pro doubles the reasoning ability of its predecessor. Matches frontier model quality at half the cost and is already rolling out to developers and subscribers today. If you use AI for coding, research, or content creation this upgrade matters.

I’ve been following AI model releases closely for the past couple of years and most updates are honestly forgettable. A small benchmark bump here, a UI tweak there. Gemini 3.1 Pro is different. Google didn’t just polish the edges they took their Deep Think reasoning engine and baked it into the everyday model. That’s a meaningful shift.

Let me break down what’s new, what it actually does in real life and whether it’s worth your time.

What Is Gemini 3.1 Pro?

Gemini 3.1 Pro is Google’s latest flagship AI model, officially released on February 19, 2026. It’s built on the same foundation as Gemini 3 Pro but with one major difference. The reasoning upgrades from Gemini 3 Deep Think are now integrated into the standard model.

In plain terms: the version of Gemini that used to require a separate, more expensive research-focused mode is now just… Gemini. Same price. Smarter brain.

It’s available right now in preview through Google AI Studio, Vertex AI, Gemini CLI and the Gemini app for Pro and Ultra subscribers.

Gemini 3.1 Pro vs Gemini 3 Pro — What Changed?

Here’s a side-by-side breakdown of the key differences:

FeatureGemini 3 ProGemini 3.1 Pro
ARC-AGI-2 Score31.1%77.1%
Humanity’s Last ExamNot leading44.4% (top-tier)
Reasoning EngineStandardDeep Think–level
Agentic Workflow SupportBasicFully optimized
Pricing (per 1M tokens)$2 in / $12 out$2 in / $12 out
AvailabilityGenerally availablePreview (Feb 2026)

The ARC-AGI-2 jump is the headline stat going from 31.1% to 77.1% means the model is now dramatically better at solving logic problems it has never encountered before. That’s not a marginal improvement. That’s a different category of tool.

Real-Life Benefits — What Can You Actually Do With It?

Gemini 3.1 Pro

Benchmarks are useful context, but what most people care about is: does this make my day easier? Here’s where Gemini 3.1 Pro genuinely delivers.

Build Apps and Dashboards Without Deep Coding Skills
One of the most impressive demos involved the model pulling live NASA telemetry data and building a fully functional. Real-time dashboard tracking the International Space Station’s orbit — from a plain-language prompt. If you run a business and need custom internal tools or want to automate data reporting, this kind of capability used to require a developer. Now it doesn’t.

Animated Web Graphics From a Text Description
Type a description, get back a fully animated SVG graphic built entirely in code. That means crisp visuals at any screen resolution, no video hosting costs and lightning-fast load times. For bloggers and content creators, this is a genuine time-saver on visual production.

Smarter AI Agents That Actually Follow Through
Where most AI models struggle is on multi-step tasks. They start strong and lose the thread halfway through. Gemini 3.1 Pro is specifically optimized for agentic workflows, meaning it plans, uses tools, checks its own work and completes tasks end-to-end with far more reliability. Developers building automation pipelines or AI assistants will feel this improvement immediately.

Research and Analysis at Expert Level
Scoring 44.4% on Humanity’s Last Exam a benchmark designed to challenge PhD-level knowledge puts Gemini 3.1 Pro among the strongest research tools available right now. Whether you’re a student, medical professional or anyone working with dense technical material. You’ll get noticeably better answers on hard questions.

Creative Work With Genuine Context Awareness
This model doesn’t just follow instructions literally. It reads tone, intent and context. Feed it a creative brief or a writing style reference, and it builds outputs that actually match the vibe not just the keywords. That matters a lot for anyone doing brand writing, portfolio work or long-form content.

Who Should Actually Use This?

Gemini 3.1 Pro is a strong fit for:

  • Developers building AI agents, automation tools or multi-step applications
  • Content creators and bloggers who want smarter writing assistance and visual tools
  • Researchers and students dealing with advanced, technical subject matter
  • Business owners who want to automate workflows without hiring a dev team

At the exact same price as Gemini 3 Pro — $2 per million input tokens, $12 per million output tokens. There’s no financial reason to stick with the older version once it fully rolls out.

Is It Worth Switching?

If you’re already using Gemini 3 Pro through the API or a Google subscription, the switch is essentially free and automatic as the preview rolls out broadly. The improvements in reasoning and agentic performance are real, not just marketing numbers.

For those still on GPT or Claude this release puts Gemini back in serious contention. Independent testing from Artificial Analysis shows Gemini 3.1 Pro leading six out of ten categories in their Intelligence Index. While costing less than half of comparable frontier models to run. That’s a compelling combination.

Google says the full rollout will follow after the preview phase, with further improvements specifically targeting agentic use cases. Given what this preview already delivers, the full release should be worth watching closely.

Sources

1...678...24Page 7 of 24