Tag:

Google

Texas Longhorns quarterback Arch Manning (#16) runs up field for a first down during the CFP Semifinal Cotton Bowl Classic football game between the Ohio State Buckeyes and Texas Longhorns on January 10, 2025 at AT&T Stadium in Arlington, TX
Gaming Gear

Fox and Google Reach Agreement to Keep Fox Channels on YouTube TV

by admin


Fret not, football fans, Fox channels are staying on YouTube TV. Google and Fox reached an agreement to end their dispute that could have led to the removal of Fox channels from Google-owned YouTube TV. 

Google announced the deal last night in a YouTube blog post stating, “We’re happy to share that we’ve reached an agreement with Fox to keep their content on YouTube TV, preserve the value of our service for our subscribers and offer more flexibility in the future. This means that Fox channels, including the Fox Broadcast Network, Fox News, and Fox Sports, remain available for our subscribers along with 100+ channels and football fans will not miss any of the action this weekend.”

With the first big Saturday of college football this weekend — starting with No. 1 Texas facing off against No. 3 Ohio State at noon on Fox — and the first Sunday of NFL football next weekend, the pressure was on for Google and Fox to get a deal done. And after a short-term extension that allowed negotiations to extend past a Wednesday deadline, the two parties were able to renew their contract without any disruption in service.

Enjoy the football this weekend, folks.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google adds iPhone-like ‘Calling Cards’ to its Phone app
Gaming Gear

Google adds iPhone-like ‘Calling Cards’ to its Phone app

by admin


Google’s Phone app is adding “Calling Cards” that let you customize the appearance of contact screens for incoming calls. They’re similar to the Contact Poster feature that iPhone users have had since 2023, allowing Google Phone app users to replace the teeny contact photos that appear when someone is calling you with full-screen images and stylized names.

The update is part of Android’s Material 3 Expressive design language overhaul, which Google used to test a revamped Phone app interface in June. Calling Cards started appearing in beta versions of Google’s Contacts and Phone apps earlier this month, but now they’re getting a public release in version v188 of the Phone app. Google says Calling Cards will be available worldwide, and are being rolled out “in phases,” so they might take a while to appear for everyone.

When the feature becomes available, Phone app users will see a banner on the Home tab that reads “Introducing calling card: Customize how you see your contact when they call you.” Tapping on this takes users to the Calling Card page, but you can also navigate to it manually in Contacts. From there, Calling Cards can be created for each contact by selecting an image from the device’s camera, gallery, or Google Photos, and choosing a font and color option that will be used to display the contact’s name.

Unlike Contact Posters on iOS, you can’t design your own Calling Card that will appear for other contacts when you call them. Google’s Calling Cards only let you set customized screens that are specific to your device, so you’ll have to set these for every individual person if you want to use the feature. Those contacts can’t edit how their Calling card appears on your device, however, so you can have some fun with how you customize them.

This is rolling out alongside a new “Take a message” feature for the Phone app that automatically answers and transcribes voicemails when you miss a call. Users can record a custom greeting that callers hear when leaving a voicemail for Take a Message, or select from one of the available greeting presets. Transcripts and voicemail audio can be found in the Phone app Recents tab, and Google says that all messages are “stored privately on your device.” The feature is available on Pixel 4 phones or newer, and on Pixel Watch 2 models or newer when paired with Pixel 6 or more recent Google phone models.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Decrypt logo
GameFi Guides

Google Gemini and Elon Musk’s Grok Are Gaining on ChatGPT

by admin



In brief

  • ChatGPT still leads, but Google’s Gemini and Musk’s Grok are closing fast, per Andreessen Horowitz’s Top 100 AI Apps.
  • The AI app market is stabilizing—fewer new entrants on web, more originality on mobile.
  • China’s AI giants and “Brink List” newcomers show the next wave of global challengers.

For more than a year, OpenAI’s ChatGPT has been the undisputed heavyweight of consumer generative AI.

But according to the latest “Top 100 Gen AI Consumer Apps” report from venture capital giant Andreessen Horowitz, which analyzes two-and-a-half years of AI usage data, the challengers are finally starting to close the gap. Google’s Gemini and Elon Musk’s Grok are climbing the charts, signaling that rivals are coming for OpenAI’s crown.

That said, the Gen AI ecosystem is showing signs of stabilization. The web list saw 11 newcomers compared to 17 in March 2025, indicating less churn. The mobile app market, however, saw 14 new entries, partly due to app stores cracking down on “ChatGPT copycats,” making room for more original apps.

The report also seems to be somewhat at odds with a June SimilarWeb analysis that showed that OpenAI’s GPT was eating the web, with some 5.5 billion visits a month.

The big takeaways

That could be explained, of course, by how rapidly the AI landscape is shifting. Google has been making significant moves, with four products entering the web list for the first time. Gemini, its general LLM assistant, now ranks second on the web, capturing about 12% of ChatGPT’s web visits.

Other notable Google products include AI Studio (developer-oriented, top 10 web) and NotebookLM (#13 web), which has seen steady growth. On mobile, Gemini is also #2, with strong Android usage (nearly 90% of its MAUs).



While ChatGPT still leads among general LLM assistants, Google, xAI, and Meta are closing the gap.

X’s Grok jumped from no mobile app in late 2024 to 20 million MAUs, ranking #4 on web and #23 on mobile. This surge was fueled by the release of Grok 4—with improved reasoning, real-time search, and tool integration—and the introduction of AI companion avatars. Meta AI, however, has seen more subdued growth, ranking #46 on the web list and missing the mobile cutoff. DeepSeek and Claude have seen mobile usage flatten, while Perplexity continues to grow.

The AI world, of course, includes significantly more users than those who use the dominant platforms in the west. Chinese AI apps are gaining significant traction globally. Three China-serving companies—Quark (#9 web, #47 mobile), Doubao (#12 web, #4 mobile), and Kimi (#17 web)—are in the web top 20, largely due to China being the largest market and restrictions on non-Chinese LLMs.

Additionally, a substantial portion of the web list and 22 of the top 50 mobile apps—especially in photo/video editing, with Meitu contributing five entries—are developed in China and are now “exported” globally. Chinese video models, in particular, show an advantage, potentially due to more research focus and fewer IP regulations.

“Vibe coding” platforms are generating strong user engagement and revenue retention. Lovable and Replit debuted on the main list, while Bolt, previously a newcomer, is now on the “Brink List.” These platforms are also boosting traffic for other AI products and infrastructure providers like Supabase.

The report, now in its fifth edition, continues to anoint “All-Stars.” Fourteen companies have consistently appeared in all five editions of the web top 50, earning them “All-Star” status. These include general assistance (ChatGPT, Perplexity, Poe), companionship (Character AI), image generation (Midjourney, Leonardo), editing (Veed, Cutout), voice generation (Eleven Labs), productivity (Photoroom, Gamma, Quillbot), and model hosting (Civitai, HuggingFace). These All-Stars primarily hail from the U.S., UK, Australia, China, and France.

And finally, to track momentum at the edges, Andreessen Horowitz now publishes a “Brink List”—the five web and five mobile apps closest to breaking in. From the previous cycle, three “almosts” actually made it: Lovable (#22 web), PolyBuzz, and Pixverse. The message is clear: Today’s near-misses can be tomorrow’s breakouts.

The bottom line

ChatGPT still leads—but Gemini and Grok’s rise proves the fight is no longer one-sided. As Andreessen Horowitz’s top 100 shows, the consumer AI ecosystem is growing up, but it hasn’t stopped mutating. The giants may be closing the gap, but the next big breakout could still come from the “brink.”

Generally Intelligent Newsletter

A weekly AI journey narrated by Gen, a generative AI model.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google Pixel 10, Pixel 10 Pro, and Pixel 10 Pro XL Review
Product Reviews

Google Pixel 10, Pixel 10 Pro, and Pixel 10 Pro XL Review

by admin



Photograph: Julian Chokkattu

However, Camera Coach—which launches in a preview (sort of like a beta)—has a Get Inspired button that uses generative AI to deliver some photos it thinks you might like to try and mimic. These photos are often quite a bit different from the originally scanned image, and I found these less helpful. I think Camera Coach is a great way to teach someone about their phone’s camera capabilities, because most people barely scratch the surface, but I don’t think this generative add-on was really necessary.

Then there’s Pro Res Zoom, which is conflicting. On the Pixel 10 Pro and Pro XL, you can digitally zoom in anywhere from 30X to 100X, and the phone runs through more than 200 frames, blending images, and using generative AI to fill in the details. The results are spectacular. Take a look at the image of the Chrysler building in Manhattan, which I captured from Greenpoint, Brooklyn, across the water at 100X zoom. I’ve compared the image with real photos of the Chrysler building, and the results match up. It still leaves a weird taste in my mouth. The composition is mine, but a part of me feels like it isn’t my photo. (Note: Google says it’s not designed to work on people.)

Pro Res Zoom (100X) on Pixel 10 Pro XL.

Camera Coach on Pixel 10 Pro XL.

Lastly, there’s video capture. Google has made strides over the years in improving the video output of its phones, but it has largely started relying on Video Boost. Once enabled, this sends your footage to the cloud for processing, making the clips brighter, sharper, more colorful, and better stabilized. (It’s exclusive to the Pro models.)

The videos I’ve shot in the past week do genuinely look great once they’ve been put through the Video Boost ringer, but I still find the iPhone delivers better native footage, with better stabilization. You also have to account for the fact that some of these boosted videoclips arrived the next day for me (though you still have access to the original). It’s a smart solution, but I’d like to see Google improve the native video capture. Case in point: The Galaxy S25’s video footage was brighter, less grainy, and better stabilized than the Pixel 10’s.

The AI Assist

Photograph: Julian Chokkattu

Finally, on to the software. It’s probably not a coincidence, but both Google and Apple redesigned their operating systems this year, and I think Google’s Material 3 Expressive design language came out on top. It’s bubbly, colorful, fun, and playful. Apple’s Liquid Glass feels a bit more stale to me.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google Pixel 10 Pro review: AI, Qi2, and a spec bump too
Product Reviews

Google Pixel 10 Pro review: AI, Qi2, and a spec bump too

by admin


Last year, Google proved it could make a phone that looks and feels like a true flagship, despite the software feeling like an AI jumble. This year, the Pixel 10 Pro starts to put AI features together in a way that actually makes sense — and it manages to upgrade the hardware a bit, too.

Google has finally locked in a high-end finish and feature set for this phone, and it feels more polished than ever. There’s Android’s version of MagSafe, a flagship-worthy processor, and an excellent camera. All the stuff you’d want from a phone that starts at $999.

$999

The Good

  • Qi2 with magnets!
  • Some legitimately handy AI features
  • Great camera with upgraded portrait mode

The Bad

  • Battery life is just okay
  • Some AI features still feel like gimmicks
  • For real, what is a photo?

The 10 Pro also represents a baby step from AI’s party trick era to becoming genuinely useful on a mobile device. Magic Cue, which aims to surface information from your email and calendar contextually, lives up to its name for the most part — like the time it offered to put a coffee meetup with a friend on my calendar as we were hashing out the details over text. But alongside great features like Magic Cue, you’ll still find some that feel more like they’re there to satisfy an internal mandate to put AI into every available nook and cranny.

Between the hardware upgrades and a slightly more cohesive software experience, there’s something pretty special here. The Pixel 10 Pro is a phone with solid upgrades, though it doesn’t quite feel like a must-upgrade as long as your current device is working fine. It’s a glimpse of the future, with all the messiness of now mixed in there, too.

AI gets more useful on the Pixel 10.

From the outside, the Pixel 10 Pro is very hard to tell apart from its predecessor. That’s just fine; this is a good template to keep working from. Inside is another story, and that story begins with Tensor G5.

I get the sense that Google’s fifth custom chipset is the one that the company has been waiting on. Tensor G5 is the first made by TSMC, and it seems to be the key to unlocking a lot of on-device AI. Magic Cue, for example, runs in the background on device. A camera feature we’ll get to later runs a diffusion model on device. Same with the phone call translator that mimics the sound of your voice. It’s more than just an impressive list; running these entirely on the phone means your data is much more private than if it had taken a trip to the cloud.

Running these features entirely on the phone means your data is much more private than if it had taken a trip to the cloud

Just like last year, the Pro variant of the Pixel 10 comes in two sizes. Both the Pro and Pro XL come with 16GB of RAM, and variants with 256GB or more storage use faster UFS 4.0 memory. They all have the new chip, and while I can’t say I found previous Pixels to be laggy, this one seems snappy. It stutters a bit with dense, media-heavy web pages like character builds on Icy Veins. Android Authority has a good rundown on the nuts and bolts of Tensor G5, and confirms that there’s no ray tracing support. For what it’s worth, Diablo Immortal runs just fine on the 10 Pro. The phone also doesn’t seem to heat up quite as much or as quickly as previous Tensor-powered Pixels either, which have a reputation for running hot. I used the 10 Pro as a hotspot outside on a warm morning without a problem — something I’ve had less success with on previous Pixel phones.

The 10 Pro’s battery capacity is a little bigger this time around: 4870mAh versus 4700mAh, likewise 5200mAh compared to 5060mAh on the XL. Maybe it’s all that on-device AI, but battery life seemed a little worse than usual despite the slightly upgraded capacity. I tested it with the always-on display enabled and the highest screen resolution available. On a lighter day with the 10 Pro, I found the battery running down to around 50 percent by bedtime; a heavier day with a decent chunk of hotspot use brought it down into the 30s by night. That’s fine, and within the realm of a modern flagship phone, but not exactly stellar.

The Pixel 10 Pro doubles as a clock on Google’s Qi2 charging stand with this screen saver enabled.

I’m still testing the 10 Pro XL’s battery stamina. Given all of the resource-intensive, personalized AI features on board, a week just wasn’t enough to draw solid conclusions about battery life on two different phones. I’ll be updating this review soon (with the help of my colleague Dominic Preston who’s also testing the XL model).

There’s not a lot to say about the Pixel 10’s other marquee addition, Qi2 support with built-in magnets, other than it’s great and I’m glad it’s here. On the 10, 10 Pro, and 10 Pro Fold, you get 15W wireless charging, and on the Pro XL up to 25W with the Qi2.2 standard. I throw my phone on a wireless charger at the end of the day so slower charging never bothers me, but the magnets sure are nice to have. I never put a case on my phone either, for better and worse (mostly worse), so I’m thrilled I can reap the benefits of what is essentially MagSafe on an Android phone without having to use a case like you do with Samsung’s latest flagships. Hardware is hard, and it took Google a minute to get here, but this year’s Pixel phones stand out in a way that has eluded their predecessors.

That’s just a good-looking phone.

But let’s not get too carried away; Google isn’t exactly reinventing the wheel by adding magnets and an updated processor. Some of the AI features on board the 10 Pro are new and actually useful, though — starting with Magic Cue. This is the context-aware AI that’s designed to deliver helpful information when you need it, without you having to go look for it. That might sound vague and unserious, but it’s the kind of thing we’ve been promised ever since companies started talking about AI on our phones. And it actually does work. Was it magic? Hardly, but what I’ve seen so far makes me more optimistic about AI than anything else I’ve seen it do on phones.

Magic Cue is, by its nature, just kind of floating around in the background while you do stuff on your phone. It runs completely on device — it’s not offloading anything to the cloud — and only works in a handful of Google apps. But they’re important ones, including Gmail, Messages, Calendar, and the phone app as well as last year’s addition: Screenshots. If someone messages you and asks for information about a date or reservation, Magic Cue will check your calendar or your inbox and you’ll see the details pop up above the keyboard as a suggestion. You can long press to check where Magic Cue got the information or just tap to drop it into the conversation. I tried it out in a quick staged-but-plausible test with my colleague Vee Song by having her ask about a concert I’d already put on my calendar and it was honestly cool as heck.

PreviousNext

1/2The Magic Cue suggestion pops up at the bottom of our conversation.

Magic Cue is also supposed to help you search for stuff. It does this by paying attention to what’s on your screen, detecting the kind of app you’re opening up — it knows Amazon is a shopping app, for example — and suggesting text might want to paste into the search bar, like the name of a product you were just looking at in Chrome. Not quite as exciting as saving me a trip to my calendar, but I can see it being something I’d get used to using. Is the idea that my phone is monitoring what’s on my screen a little weird? For sure. But the fact that this information doesn’t leave the device makes me feel better, or at least as good as I can feel, knowing that Google knows everything about me anyway.

I saw Magic Cue most often offer to save calendar events based on my conversations. As I was figuring out a time to meet for coffee, Magic Cue offered a link to check the appropriate day on my calendar. When we landed on a time, it let me add the event with a tap. None of this is life-changing, and “magic” seems like a strong descriptor, though I’ll allow it. Mostly, this seems like a really handy feature that will take just a little bit of the friction out of using your phone. Remember the first time you saw your phone autofill a one-time 2FA code from a text? It’s like that. Ultimately, I think that’s what AI on our phones will become. Something that just happens, saving us a little time here and there, that fades into the background once you get used to it. I guess there is a little magic to that.

I heard the translations in English — and in a voice that kind of sounded like Vee

The on-device AI extends all the way to the phone app where it provides real-time voice translations on calls — a familiar feature, this time with a twist: it mimics the voice of the speaker. I once again called on my colleague Vee Song (literally) who was also using a Pixel 10. She enabled voice translations on her side and spoke in Japanese. I heard the translations in English — and in a voice that kind of sounded like Vee. According to Vee, the feature’s translation of my English into Japanese was pretty good, though it struggled a bit with what she calls her textbook Japanese. It got the point across, and I guess it was nicer than listening to a neutral-sounding robot.

But as Vee observed, this feature would likely work best for a tourist in a foreign country trying to make a dinner reservation, not for family members trying to catch up. And if that’s the case, the personalization of the spoken voice feels kind of unnecessary. It makes for a cool party trick, though.

If voice translation is halfway between a helpful feature and a gimmick, then Daily Hub leans even further into gimmick territory. It’s a lot like the Now Brief Samsung introduced on its S25 phones, and it’s supposed to act as a quick digest for your day as well as a place to find some inspiration based on your recent activity. It does the first part of that job just fine; it’s maybe a more longwinded version of Google’s At a Glance widget, which gives you a heads up on the weather and upcoming calendar events. But it also misconstrued some of my recent Google search history in puzzling and hilarious ways. I looked up the schedule for our recycling service, provided by Waste Management, and it took that to mean that I’m interested in learning more about waste management generally. Uh, not quite.

Thanks Daily Hub, but I’m all good on embracing tech-fueled adventures.

I had a similar experience with the new Journal app, which uses on-device AI to generate prompts and reflections based on what you write. In one entry I mentioned that my kid was sad because it was his friend’s last day at school. The app reassured me that it was okay to feel sad about her passing. To be clear, she was just moving to another school.

Maybe knowing that you’re going to get some kind of feedback at the end is motivation for people to journal when they wouldn’t have otherwise? And I guess I did make a point to write an entry every night before bed, which is not a habit I’m in currently. But aside from that, I can’t say I found the prompts or bits of reflection helpful.

AI shows up in yet another place: the camera app. And not just as a tool for adding wild stuff to your photos — it’s in the actual camera. On the 10 Pro and 10 Pro XL, there’s a new feature called Pro Res Zoom that aims to make digitally zoomed photos look a little less like garbage. Once you get past 30x zoom and up to the maximum of 100x, Pro Res Zoom will kick in and use a diffusion model to try and clean up your image.

This happens after the fact; the original image is retained, and it all runs on your device. It also doesn’t attempt to enhance any people it identifies in an image, which is for the better. The results depend a lot on your subject, the conditions, and how far you’re pushing the zoom range. With enough light, a predictable subject, and moderate zoom, the results can be really good.

Before Pro Res Zoom (left) and after (right). You gotta admit that Gen AI does a pretty good job.

Pro Res Zoom has a harder time with writing. If you take a picture of a sign in the distance, you’ll see the telltale signs of generative AI — something similar to real writing that’s actually an alien language when you look closer. And all the way at 100x zoom there’s only so much even AI can do to give you a usable photo.

In a shot I took of a crane wading in a pond, it didn’t know what to do with the leaves and bits of debris on the surface of the water and turned them into little white points, like dozens of tiny sailboats. Kind of poetic but not what I had in mind. And this goes without saying, but Pro Res Zoom isn’t going to give you anything that looks as good as an optical zoom lens. Trust me. I lugged around a Nikon Coolpix P1100 with a 3000mm equivalent lens just to be sure.

1/3Taken with Pixel 10 Pro at 100x zoom before Pro Res Zoom processing

Pro Res Zoom photos are tagged with C2PA metadata that identifies them as captured with a camera and edited using generative AI. In fact, all photos taken with the Pixel 10 are tagged to reflect whether or not AI was involved, which might seem like overkill but feels increasingly necessary in a world with accessible, capable gen AI editing tools.

Does it still creep me out a little? Yeah. Is a picture I took with Pro Res Zoom still a photo, or is it something else? I’m not so sure. But I don’t think this is the last we’ll hear of diffusion models in phone cameras, so we’re all going to have to find our own levels of comfort with this kind of thing.

Elsewhere in the camera there’s some good news: portrait mode is much improved. The Pixel camera had some catching up to do here, and it’s not perfect, but subject isolation is a lot better on the Pixel 10 series than on the 9. Reynolds tells me that this revamped portrait mode pays “particular attention to hair,” which is great news for me, personally, because my kid’s hair is ungovernable. In photos taken with the 10 Pro, I can see where the camera retained individual strands of hair rather than just blurring them into the background the way the 9 Pro tends to. It goes a long way to making that photo look more convincing.

A good tool for the job.

The Pixel is in kind of a funny spot, especially in the US. Year after year its market share is comparatively small in our iPhone and Galaxy-dominated landscape. In fact, Counterpoint Research counts its percentage of the market in the single digits and lumps Google into the “others” category, far behind Apple, Samsung, and Motorola. But what Google is doing on Pixel phones matters more than its sales figures suggest. The Pixel is a showcase for the Android ecosystem, particularly in the last few years as Google has pushed it into proper flagship territory. The 10 Pro feels like that symbol of what’s possible on Android more than ever.

The Pixel 10 series represents the first phones from a major OEM to get full Qi2 support. They’re the first phones to put C2PA content credentials on photos taken with the camera. They offer a glimpse of what AI can actually do to save us some time and effort tapping around on our phones. It still feels like AI is being shoved into corners of the device where it doesn’t really need to be, but for the first time it feels like there’s a kind of connective tissue between the useful bits.

There was a time, particularly in the Tensor era, where using a Pixel phone felt a little bit like being an early adopter, and not in a good way. The prices were lower, but the hardware felt cheaper, software bugs persisted, and the chipsets ran hot. But the Pixel series has evolved into something better, something worthy of the title “flagship.” If the 10 Pro represents the best of what Android can do, then there’s a lot for Android fans to look forward to — whether it’s on a Pixel or not.

Photography by Allison Johnson / The Verge

Agree to Continue: Google Pixel 10 Pro

Every smart device now requires you to agree to a series of terms and conditions before you can use it — contracts that no one actually reads. It’s impossible for us to read and analyze every single one of these agreements. But we started counting exactly how many times you have to hit “agree” to use devices when we review them since these are agreements most people don’t read and definitely can’t negotiate.

To use a Pixel 10 series phone, you must agree to:

The following agreements are optional:

  • Provide anonymous location data for Google’s services
  • “Allow apps and services to scan for Wi-Fi networks and nearby devices at any time, even when Wi-Fi or Bluetooth is off.”
  • Send usage and diagnostic data to Google
  • Talk to Google hands-free: “If you agree, Google Assistant will wait in standby mode to detect ‘Hey Google’ and certain quick phrases.”
  • Allow Assistant on lock screen

Additionally, if you want to use Google Assistant, you must agree to let Google collect app info and contact info from your devices. Other features like Google Wallet may require additional agreements.

Final tally: five mandatory agreements and at least five optional agreements

0 CommentsFollow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

  • Close

    Allison Johnson

    Posts from this author will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All by Allison Johnson

  • AndroidClose

    Android

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Android

  • GoogleClose

    Google

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Google

  • Google PixelClose

    Google Pixel

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Google Pixel

  • MobileClose

    Mobile

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Mobile

  • Phone ReviewsClose

    Phone Reviews

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Phone Reviews

  • ReviewsClose

    Reviews

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Reviews

  • TechClose

    Tech

    Posts from this topic will be added to your daily email digest and your homepage feed.

    PlusFollow

    See All Tech



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google Doubles Down on AI: Veo 3, Imagen 4 and Gemini Diffusion Push Creative Boundaries
GameFi Guides

Google Reveals Layer-1 ‘Universal Ledger’ Plans as Circle, Stripe Prep Rival Chains

by admin



In brief

  • Rich Widmann, Google Cloud’s head of Web3 strategy, confirmed that the Universal Ledger is a layer-1 blockchain.
  • The system uses Python for smart contracts, diverging from industry standards like Solidity and Rust.
  • Analysts question Google’s neutrality as it competes with Stripe and Circle for institutional blockchain infrastructure.

Over five months after Google Cloud announced a partnership with CME Group, Rich Widmann, the tech giant’s head of Web3 strategy, confirmed Tuesday that the company’s Universal Ledger is indeed a layer-1 blockchain.

“All this talk of layer-1 blockchains has brought Google’s own layer-1 into focus,” Widmann wrote on LinkedIn. “If you’re building a layer-1, it has to be differentiated.”

Widmann’s statement follows CME Group’s March 25 announcement that it has completed the first phase of integration and testing for the project. At the time, details were sparse on whether it was public or private, as well as if it was a layer-1 chain.

A layer-1 or L1 blockchain is a foundational network that runs independently, handling transactions and security directly. Unlike layer-2 or L2 chains, it doesn’t rely on another chain for validation or settlement, though those can extend and improve a chain’s efficiency.



Decrypt reached out separately to Widmann and Google, but did not receive an immediate response.

Why Python?

Dubbed the Google Cloud Universal Ledger (GCUL), Widmann described it as a base layer enabling Python-based smart contracts, setting a programmable, distributed ledger for wholesale payments and asset tokenization.

The choice of programming language sets Google’s L1 apart from those typically used and accepted as standard in the crypto industry, such as Solidity for Ethereum-compatible chains and Rust for chains like Solana, Aptos, and Sui.

Choosing Python is “pragmatic” because it “lowers the barrier for enterprises and fintech developers who already use it for data, finance, and machine learning,” Christine Erispe, a developer advocate at Ethereum Philippines, told Decrypt.

With Python, the upcoming L1 could “accelerate experimentation,” but may also “silo developers” unless Google makes efforts to provide “strong tooling, auditing, and interoperability bridges,” Erispe said.

That move is “a contrarian bet,” because “instead of being EVM-compatible, it leans on Google’s scale, financial institution reach, and a differentiated programming model,” she added.

Credibly neutral?

Unlike other upcoming layer-1 chains such as Stripe’s Tempo or Circle’s Arc, Google’s network is positioned as open infrastructure, with Widmann describing it as a “performant, credibly neutral” chain that “any financial institution” can build on.

While Stripe and Circle are “building chains that fit directly into their existing businesses,” Google is “playing a different game: scale and neutrality,” Aharon Miller, co-founder and COO of crypto payments gateway Oobit, told Decrypt.

As a centralized tech giant, Google “already runs half of the internet’s infrastructure, but the real test is whether institutions believe they’ll stay neutral in the long term,” Miller said.

However, Dr. Sean Yang, chief technology officer at OORT—a data cloud for decentralized AI—argued that Google’s neutrality claim may be “more marketing than reality.”

Google has “massive conflicts of interest across payments, cloud services, and advertising,” Yang told Decrypt.

Asked about the differences between the three L1s underway, Yang said Google is “going broad” while “Circle is going deep,” and “Stripe is targeting developers and payment companies.”

While not in direct competition, the three are “carving out different segments of institutional blockchain infrastructure,” Yang said.

Daily Debrief Newsletter

Start every day with the top news stories right now, plus original features, a podcast, videos and more.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google Cloud sparks backlash with ‘private and permissioned’ L1
Crypto Trends

Google Cloud sparks backlash with ‘private and permissioned’ L1

by admin



Google’s new Layer1 platform faces heavy criticism from the crypto community. Dubbed GCUL, the platform is meant to facilitate cross-border payments and asset settlements through a distributed ledger.

Summary

  • Users on X criticized Google’s upcoming L1 for being a permissioned and private system.
  • The platform plans to be more open in the future and aims to simplify cross-border payments and asset settlements through a distributed ledger.

In its main blogpost, Google Cloud’s blockchain is described as a “private and permissioned system” which leverages Google’s technology. These are principles that go against the decentralized and permissionless values the crypto community is built upon.

“GCUL offers significant benefits to both clients and financial institutions. Clients experience near-instant transactions, especially for cross-border payments, along with low fees, 24/7 availability, and payment automation,” wrote the company.

Even though the announcement claims it plans to make the blockchain more open overtime, traders on X criticized the company for launching a blockchain that seems to oppose the core principles of crypto.

“It’s a permissioned chain, ran by an American corporation with close ties to the government. I don’t think these people understand what “credibly neutral” means in the context of blockchains,” said one trader.

“Fully centralized? Then they shouldn’t even call it blockchain,” said another X user.

The polarizing project invited comments from the CEO of StarkWare Industries Eli Ben-Sasson as well as co-founder of crypto investment firm DBA, Jon Charbonneau who had first heard of the project.

“I think corporations doing L1s is ngmi [not gonna make it]. Including base. I know this is a contentious opinion. But reminding you I said the same about Diem,” said Ben-Sasson.

Details surrounding Google Cloud’s upcoming L1

In a recent LinkedIn post, Head of Web3 strategy at Cloud, Rich Widmann, dropped details about the project that is still in its private testing phase. According to Widmann, the Layer1 blockchain is designed to enable Python-based smart contracts.

“As a product leader in crypto, you know that if you’re building a Layer 1 it has to be differentiated,” said Widmann in his post, which compared the ledger to the likes of Stripe and Circle.

In addition, the platform claims to be “credibly neutral” as to enable any financial institution to make use of it to build on-chain infrastructure.

At the moment, the L1 blockchain is still in its testing phase. However, its announcement implies that it will be opened to the full Google user base and “100s of institutional partners.”

According to the blogpost, GCUL will be aimed at simplifying the management of commercial bank accounts and facilitating cross-border transactions through a distributed ledger. It plans to integrate multiple currencies and assets, eliminating the need to build additional infrastructure.

Another highlighted feature it claimed to have was transaction fees that remain stable and are invoiced to the user on a monthly basis. All the while, it criticized other crypto protocols for having “volatile upfront crypto gas fees.”

At the moment, the company has yet to reveal a set date for when the project goes public.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google Will Make All Android App Developers Verify Their Identity Starting Next Year
Product Reviews

Google Will Make All Android App Developers Verify Their Identity Starting Next Year

by admin


Android’s open nature set it apart from the iPhone as the era of touchscreen smartphones began nearly two decades ago. Little by little, Google has traded some of that openness for security, and its next security initiative could make the biggest concessions yet in the name of blocking bad apps.

Google has announced plans to begin verifying the identities of all Android app developers, and not just those publishing on the Play Store. Google intends to verify developer identities no matter where they offer their content, and apps without verification won’t work on most Android devices in the coming years.

Google used to do very little curation of the Play Store (or Android Market, if you go back far enough), but it has long sought to improve the platform’s reputation as being less secure than the Apple App Store. Years ago, you could publish actual exploits in the official store to gain root access on phones, but now there are multiple reviews and detection mechanisms to reduce the prevalence of malware and banned content. While the Play Store is still not perfect, Google claims apps sideloaded from outside its store are 50 times more likely to contain malware.

This, we are led to believe, is the impetus for Google’s new developer verification system. The company describes it like an “ID check at the airport.” Since requiring all Google Play app developers to verify their identities in 2023, it has seen a precipitous drop in malware and fraud. Bad actors in Google Play leveraged anonymity to distribute malicious apps, so it stands to reason that verifying app developers outside of Google Play could also enhance security.

However, making that happen outside of its app store will require Google to take a page from Apple’s playbook and flex its muscle in a way many Android users and developers could find intrusive. Google plans to create a streamlined Android Developer Console, which devs will use if they plan to distribute apps outside of the Play Store. After verifying their identities, developers will have to register the package name and signing keys of their apps. Google won’t check the content or functionality of the apps, though.

Google says that only apps with verified identities will be installable on certified Android devices, which is virtually every Android-based device—if it has Google services on it, it’s a certified device. If you have a non-Google build of Android on your phone, none of this applies. However, that’s a vanishingly small fraction of the Android ecosystem outside of China.

Google plans to begin testing this system with early access in October of this year. In March 2026, all developers will have access to the new console to get verified. In September 2026, Google plans to launch this feature in Brazil, Indonesia, Singapore, and Thailand. The next step is still hazy, but Google is targeting 2027 to expand the verification requirements globally.

A Seismic Shift

This plan comes at a major crossroads for Android. The ongoing Google Play antitrust case brought by Epic Games may finally force changes to Google Play in the coming months. Google lost its appeal of the verdict several weeks ago, and while it plans to appeal the case to the US Supreme Court, the company will have to begin altering its app distribution scheme, barring further legal maneuvering.

Among other things, the court has ordered that Google must distribute third-party app stores and allow Play Store content to be rehosted in other storefronts. Giving people more ways to get apps could increase choice, which is what Epic and other developers wanted. However, third-party sources won’t have the deep system integration of the Play Store, which means users will be sideloading these apps without Google’s layers of security.

It’s hard to say how much of a genuine security problem this is. On one hand, it makes sense Google would be concerned—most of the major malware threats to Android devices spread via third-party app repositories. However, enforcing an installation whitelist across almost all Android devices is heavy handed. This requires everyone making Android apps to satisfy Google’s requirements before virtually anyone will be able to install their apps, which could help Google retain control as the app market opens up. While the requirements may be minimal right now, there’s no guarantee they will stay that way.

The documentation currently available doesn’t explain what will happen if you try to install a non-verified app, nor how phones will check for verification status. Presumably, Google will distribute this whitelist in Play Services as the implementation date approaches. We’ve reached out for details on that front and will report if we hear anything.

This story originally appeared on Ars Technica.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Google Doubles Down on AI: Veo 3, Imagen 4 and Gemini Diffusion Push Creative Boundaries
Crypto Trends

Google Boosts Gemini AI Image Capabilities in Latest Salvo Against ChatGPT

by admin



In brief

  • Google upgraded Gemini with 2.5 Flash Image, taking aim at OpenAI’s dominance.
  • Developers can remix template apps in AI Studio and deploy custom image projects instantly.
  • Google has expanded access through OpenRouter and fal.ai, widening distribution to coders worldwide.

Google launched Gemini 2.5 Flash Image on Tuesday, delivering a new AI model that generates and edits images with more precision and character consistency than previous tools—attempting to close the gap with OpenAI’s ChatGPT.

The tech giant’s push to integrate advanced image editing into Gemini reflects a broader push among AI platforms to include image generation as a must-have feature. The new tool, now available across Gemini apps and platforms, lets users edit visuals using natural language—handling complex tasks like pose changes or multi-image fusion without distorting faces or scenes.

In a blog post, Google said the model allows users to “place the same character into different environments, [and] showcase a single product from multiple angles… all while preserving the subject.”

🍌 nano banana is here → gemini-2.5-flash-image-preview

– SOTA image generation and editing

– incredible character consistency

– lightning fast

available in preview in AI Studio and the Gemini API pic.twitter.com/eKx9lwWc9j

— Google AI Studio (@googleaistudio) August 26, 2025

The model first appeared under the pseudonym “nano-banana” on crowdsourced testing site LMArena, where it drew attention for its seamless editing. Google confirmed Tuesday it was behind the tool.

Google said the system can fuse multiple images, maintain character consistency for storytelling or branding, and integrate “world knowledge” to interpret diagrams or combine reference materials—all within a single prompt.



The model costs $30 per million output tokens—about four cents per image—on Google Cloud. It’s also being distributed via OpenRouter and fal.ai.

OpenAI introduced the GPT-4o model in May 2024 and added image generation in March 2025, which helped push ChatGPT’s usage above 700 million weekly active users. Google reported 400 million monthly active Gemini users in August 2025, which would indicate weekly usage that considerably trails OpenAI.

Google said all outputs will include an invisible SynthID watermark and metadata tag to mark them as AI-generated to address concerns around misuse and authenticity.

Generally Intelligent Newsletter

A weekly AI journey narrated by Gen, a generative AI model.





Source link

0 comments
0 FacebookTwitterPinterestEmail
Decrypt logo
NFT Gaming

Brothers Seek to Block Google Search History in $25M Crypto Heist Case

by admin



In brief

  • Brothers Anton and James Peraire-Bueno filed motions Friday to exclude Google search evidence, claiming prosecutors want to use searches for legal terms to prove criminal intent in their alleged 12-second Ethereum blockchain exploit.
  • The defense says the searches occurred during attorney consultations after being threatened by “anonymous sandwich attackers” and would force them to waive privilege to explain the context.
  • If convicted, the brothers face up to 20 years in prison per count, in what prosecutors call the first criminal case over MEV-boost blockchain exploitation.

Two MIT-educated brothers accused of allegedly stealing $25 million in crypto through a blockchain exploit are fighting to keep their Google search history out of court, saying federal prosecutors want to unfairly use searches for “top crypto lawyers” and “wire fraud statute of limitations” to prove criminal intent.

Anton and James Peraire-Bueno filed the motion in Manhattan federal court on Friday, claiming the searches are “unfairly prejudicial” and occurred during privileged attorney consultations following their alleged April 2023 heist.

U.S. District Judge Jessica G.L. Clarke must now decide whether searches conducted after the alleged crime can demonstrate consciousness of guilt or simply reflect prudent legal consultation during the investigation.



The brothers were arrested in May 2024 on conspiracy, wire fraud, and money laundering charges, with prosecutors calling it a “first-of-its-kind manipulation of the Ethereum blockchain.” 

Authorities allege they used their “specialized skills and education” to exploit Ethereum’s MEV-boost system in April 2023, fraudulently intercepting private transactions and diverting $25 million in just 12 seconds.

Court documents reveal they retained counsel immediately after being “threatened by anonymous sandwich attackers” who demanded the return of the allegedly stolen funds.

Defense attorneys provided detailed privilege logs showing Google searches coincided precisely with attorney communications.

A search for “top crypto lawyers” occurred the same day as “communications with potential counsel seeking legal representation,” according to court filings.

“For the government to argue its preferred inference (i.e., consciousness of guilt of the alleged crimes), the government would first need to establish that any given search was connected to this case,” the brothers said in the motion. “But the contents of the searches themselves do not show that.”

The defense claims prosecutors lack witnesses who can provide context for the searches, making any criminal inference “purely speculative.”

“Google search histories can be used as hints, but they’re context-dependent,” Even Alex Chandra, partner at IGNOS Law Alliance, told Decrypt. “The mere fact that someone googled something isn’t automatic proof of intent or guilt.”

“Post-conduct searches are weaker evidence,” he said, compared to searches conducted before alleged crimes, which can show planning or intent.

“It still needs corroborating evidence showing that the searches align with criminal intent,” he added. “Since it would be dangerous if Google searches are determinative alone.”

The brothers also moved to exclude news articles as hearsay with “inflammatory descriptions,” and to block a Twitter screenshot of their alleged “false signature,” saying prosecutors cannot authenticate an image from pseudonymous researcher samczsun’s tweet.

Each brother faces up to 20 years in prison per count if convicted.

Daily Debrief Newsletter

Start every day with the top news stories right now, plus original features, a podcast, videos and more.



Source link

0 comments
0 FacebookTwitterPinterestEmail
Close