Why Epic getting its way could be bad for Android. Plus, Android apps on Windows live on, Android phones and tablets now work better together, and why Gemini Nano is more than hype.
Welcome to Android Decoded, a newsletter devoted to explaining the significance of each week’s news on the Android ecosystem. This newsletter is written by Mishaal Rahman, a veteran tech reporter devoted to the Android beat.
If you were wondering why you didn’t receive last week’s newsletter, it’s because there wasn’t one. I spent my Memorial Day weekend with my family, and since there wasn’t a lot of news to cover last week, I decided to skip it. With two weeks worth of news to cover now, there’s a LOT to work about. Here are the stories I’ll be focusing on in this edition:
Earlier this year, Microsoft announced that they were ending support for the Windows Subsystem for Android (WSA) after March 5, 2025. WSA, if you weren’t aware, allowed for running Android apps and games on Windows PCs. Microsoft partnered with Amazon to integrate the latter’s Appstore into the Microsoft Store. The Amazon Appstore, however, has significantly fewer apps than the Google Play Store. Plus, the Amazon Appstore on Windows was only made available in the U.S., parts of Europe, and a few other countries. It was never made available in any Asian countries apart from Japan. Combine that with the fact that it had to compete against Google Play Games on PC with its significantly better catalog, it’s no surprise that Microsoft shuttered the program due to a lack of profitability.
The Amazon Appstore on Windows. Image: Microsoft.
But Microsoft isn’t done with Android just yet. The company announced a partnership with Chinese gaming giant Tencent to integrate the Tencent App Store into the Microsoft Store. The Tencent App Store already exists on Windows, and it utilizes Intel’s Project Celadon and Bridge technologies. Celadon allows for running Android apps in containers or virtual machines, while Bridge allows for Android apps with ARM native code to execute on x86 Intel CPUs. Like with WSA, Android apps installed from the Tencent App Store via the Microsoft Store will be treated like native Windows apps in terms of windowing, search, and launching.
Microsoft shifting its focus to supporting Android apps in markets like China makes a lot of sense. The Chinese market is filled with highly lucrative mobile games and gamers who want to play these games across devices. I’m not sure if Microsoft will be taking a cut of revenue from Tencent, but even if they don’t, they stand to greatly grow the Microsoft Store’s presence in China. (Microsoft lets developers of non-gaming apps keep 100% of their revenue if they use their own commerce platform, which sounds like it’d apply to the Tencent App Store.) Furthermore, since Tencent’s emulator will be the one utilizing Intel Bridge, Microsoft might not be on the hook this time for the licensing fees that entails.
Google has long been trying to push users towards its own ecosystem consisting of an Android phone (ideally a Pixel but any Android phone with Google services is fine), a Wear OS smartwatch, and a Chromebook. They know that making these devices work better together will go a long way towards locking people into this ecosystem. In fact, Google has a whole initiative internally called Better Together to make their platforms more synergistic. Better Together began with the development of Chrome OS’ “Phone Hub” and has since been responsible for the creation of many new features. The latest project to be born from Better Together is Android’s Cross-Device Services, and it’s one of the initiative’s most promising in recent memory.
In case you missed it, Google announced a slew of updates to its suite of Android apps the other day. These updates include two new features called “Call Casting” and “Instant Hotspot.” Call Casting lets you switch between your phone, tablet, or web browser when in a Google Meet call, while Instant Hotspot connects your tablet or Chromebook to your phone’s hotspot with a single tap.
By themselves, Call Casting and Instant Hotspot aren’t that exciting. A version of these features has already existed on other platforms or devices for years. What’s really exciting is that Google is baking these features into Play Services, which means they’ll be available on phones from every Android brand (with one exception).
One of the ways that Google differentiates its ecosystem from Apple is that they point out how Android has a lot of freedom in what hardware you want. With Cross-Device Services, Google is ensuring you don’t have to buy a phone, tablet, and Chromebook from the same brand to enjoy these features. Cross-Device Services won’t eliminate brand-exclusive features, but it’ll bring even more hardware freedom to the Android ecosystem. (Of course, I have to note that it also has the effect of making Android devices without Google services less enticing.)
I’m sure many of you are sick of hearing about AI. I, too, believe AI is overhyped. I also believe we’re in a period where every company is just throwing things at the wall and seeing what sticks. At Google I/O, we got a teaser of some genuinely useful features powered by Gemini Nano, the mobile-optimized version of Google’s Gemini large language model.
For starters, Google revealed some of the companies that are working to integrate Gemini Nano support into their apps. These companies include Adobe, Grammarly, and Patreon. Of these three, I think Grammarly is the one that made the best case for its use of on-device generative AI. The company plans to power on-device grammar suggestions using Gemini Nano. Doing so will reduce the time it takes to generate grammar suggestions as well as improve privacy.
Third-party apps may very well make the best use of Gemini Nano, but sadly, its use is locked behind Google’s AI Edge SDK for Android, which is still in private preview. For now, third-party app developers who want to leverage a LLM like Nano on-device will have to seek an alternative, like Google’s MediaPipe LLM Inference API, which requires bundling or having users download the multi-gigabyte LLM.
An upgraded version of Gemini Nano, called Gemini Nano with multimodality, will debut on the Pixel 9 series later this year. I think Google did a great job at pitching the new model: Its audio and image processing capabilities will power on-device image captioning and scam warnings. The former will be huge for people who have trouble seeing and have to deal with the myriad of images on the Internet that lack alt. text (which I’m guilty of contributing to). The latter could save less tech savvy folks from getting scammed. The former is something that can already be offloaded to the cloud, but doing so harms the user experience by forcing users to wait for the cloud-generated caption to be returned. The latter, however, is something that can only really be done on-device, for safety and privacy reasons.
Gemini Nano is one of the main reasons I’m pretty optimistic about Android’s place in Google’s new AI-centric vision. Not only does Android provide Google with the platform they need to bring their AI-powered products and services to the masses, but it also gives them a competitive advantage in what kinds of AI-powered features they can offer.
For example, I can foresee a future where Gemini Nano becomes even more powerful and contextual compared to what Google showed off at I/O, such as being upgraded to take actions directly within Android apps. Google could take screenshots of apps or read the screen contents using the Accessibility API and then use Nano to process the screen and take action according to the user’s voice input. (Apple is rumored to be doing exactly this in iOS 18, for what it’s worth.)
I don’t think Gemini Nano gives Google AI a huge advantage right now since the model is only available on a handful of Android devices, but give it some time and we’ll likely see availability expand significantly.
Google originally announced its Find My Device network (FMDN) back in May of 2023. The FMDN is an item tracking network that relies on the billions of Android devices worldwide to crowdsource location reporting. The rollout of the network was delayed to give Apple time to implement unwanted tracker alert detection, but on April 8 of this year, Google finally announced that its FMDN had begun rolling out globally.
However, for many people, the real rollout was last week. I, as well as many others, finally got access to the FMDN, right in time to test the FMDN-compatible item trackers from Pebblebee and Chipolo. I’ll be reviewing the FMDN and these item trackers in the weeks to come, but if you have any questions for me right now, feel free to shoot me a message.
Very quickly, I wanted to highlight some details I noted about Android 14 for TV, the latest version of the Android TV platform Google announced at I/O.
First, the picture-in-picture mode feature that Google says is finally coming to Google TV? Yes, it’s already been available on select Android TV devices for a few years now, but the feature is entirely new to Google TV. However, Google won’t allow media content to be shown in a PiP window.
Second, you may have heard Google specifically highlight 2W of power draw as the maximum amount of energy consumed when using Android 14 for TV’s new low energy mode or optimized energy mode. Why 2W? Because starting May 9, 2025, the EU will require that most networked devices consume 2W or less energy when in networked standby mode. The platform update also brings a slew of other performance improvements you can read about here.
Lastly, Google TV is adding support for a Find My Remote feature, but you’ll need to have a remote with a built-in speaker to support this.
It’s wild to think it’s been almost 4 years since Epic Games began its crusade against Apple and Google. Regardless of Epic’s intentions in the matter, it’s hard to argue with them that certain aspects of Apple’s App Store and Google’s Play Store businesses are unfair. Epic’s lawsuit has already spurred both Apple and Google to make numerous developer-friendly changes to their app store businesses, but these changes have mostly been beneficial to smaller app development teams rather than Epic, who seeks to open their own competing app store.
That’s why when Epic won its antitrust case against Google last December, they asked the court to force Google to not only put third-party app stores on equal footing with the Play Store but also block Google from entering into any agreement that might favor its app store over others. If approved, Epic’s list of proposed changes, detailed in an extensive 16-page document, would dramatically overhaul Google’s Play Store business and relationship to Android. Some of the proposed changes are reasonable, such as reducing the friction of downloading apps from third-party stores and eliminating the anti-steering provisions, while others are bonkers.
Particularly the changes listed in page 8, which would not only require Google Play to distribute the Epic Games Store (as well as other third-party app stores) but also force them to provide third-parties access to their entire catalog of apps. Requiring Google Play to host and distribute other app stores is a bit outlandish but would not be unprecedented. The Microsoft Store in late 2021 opened up its catalog to include third-party stores, but I see Microsoft’s announcement more as a PR move as their store has been on the back foot for years. (In a jab at Apple, for example, Microsoft also announced at the time that third-party browsers that use their own engine would be allowed.)
Downloading the Epic Games Store from the Microsoft Store. Image: Mishaal Rahman
The latter change is far more problematic for Google and, frankly, unreasonable. Epic is essentially asking the court to force Google to bootstrap other competing app stores. With access to Google Play’s entire catalog of apps, third-party stores can surface apps that aren’t available on their own platform. That, obviously, wouldn’t help the third-party store by itself, which is why Epic is also demanding that Google Play then download and install such apps discovered on third-party stores in the background.
These apps would still be subject to Google Play’s distribution terms and thus platform fees, but Epic is also demanding that Google implement a mechanism that lets users transfer their Play Store library to a third-party store. Google already implemented a feature in Android 14 that lets app stores enforce update ownership, but what Epic is seeking is a simple solution that gives third-party stores a list of apps the user installed from the Play Store. Third-party stores could then take that list, see which of those apps are available on their own platform, and then prompt the user to transfer update ownership.
Google, of course, thinks every single change (not just the two mentioned above) proposed by Epic is bad. The judge, understandably, doesn’t buy this argument but does think some of Epic’s proposed changes are too open-ended and vague. What’s really surprising, though, is that the judge is actually considering the two changes listed on page 8. These two changes, if enacted, would significantly improve the competitiveness of third-party stores, but in a way I feel is unfair for Google.
The reason I’m wary is that, while yes Google does frankly make an absurd sum of money from its app store business, this profit is a large part of why Android is where it’s at right now. If Google’s revenue from its Play Store business were to substantially decline, how would that impact the company’s future interest in and development of Android? Would the company scale back investments in certain areas, shift resources to other parts of the business, or (God forbid) lay off some employees?
This might sound dramatic, but Google is in a precarious position because they aren’t the ones in control of what happens here: the U.S. judiciary is. Google has to convince the same judge who ruled them a monopolist to not totally upend their business model. Epic largely lost its court battle with Apple, giving the latter the chance to dictate how (not) to ease up on third-party app stores. The outcome of the judge’s decision could vastly alter the future of Android and its competitiveness with iOS, so there’s a lot at stake here.
Recommended reading
Thanks for reading this edition of Android Decoded! Apologies for this going out so late, but as you can see, there was a lot to cover. Next week’s edition should hopefully come out sooner, though since I’m anticipating the release of Android 14 QPR3, I suspect I’ll have a lot to talk about.
If you have any feedback to share about this newsletter, feel free to contact me through any of my available channels!
