Everyone expects Google to announce Android 5.0 at Google I/O, so that's not really a "prediction"? My prediction:
Android 5.0 won't really be a big deal.
Now, why would I think that? Android 4.0 was a huge re-imagining of the very core of Android, while Android 4.1 brought much-needed polish to the experience. Why wouldn't Android 5.0 be on the same level?
Simple: One of the biggest strengths of Android over iOS is the ability to update core functionality without a direct OS update, and I think Google is going to push this strength hard. You can't update the built-in Apple apps without a new version of iOS. Android, on the other hand, has already separated out major features like:
- Google Calendar
- Google Maps
- Google Now
- Google Keep
- Play Store
- Play Music
- Google Plus
I expect to see this trend continue at I/O. The long-rumored Google Babel will show up at (or perhaps even before) Google I/O, and it will likely be compatible with Android 4.0 (or at least 4.1) and above. If Babel ties into SMS, this could even replace the Google Talk and AOSP Messaging apps. The AOSP E-mail and Camera apps may very well move into the Play Store. Google Now will get major changes, but these changes will come to Android 4.1 and up, as well as Chrome. We could even see an iCloud-style device backup system make an appearance, not as a system-level feature, but as a new app - or even part of the Play Store itself.
This, I think, is Google's best way of fighting Android's fragmentation problem - the version of Android you're using is significantly less important if you're getting all or most of the new features anyway.
Android 5.0 itself will be reserved for major OS-level changes that, for whatever reason, can't be made at the app/service level - things like:
- Lockscreen Widgets
- Low-level performance enhancements ala Project Butter
- API changes
- Multitasking/other general UI changes
Honestly, most of the features I want from the core Android OS are already there. The performance improvements that came with 4.1 have brought Android to near-parity with its competitors, especially on the latest generation of hardware. I'm not sure how much more remains to be done with this - I have a admittedly-limited understand of how Android works behind the scenes, but I'm not sure performance could get much better without a ground-up re-write.
What do you guys still want from Android that couldn't be done at the app/service level? Personally I'm mostly thinking of API changes, like the way Android treats background audio.
Of course, it's possible that this is all merely wishful thinking - I pre-ordered a One from AT&T yesterday, and given AT&T's history of Android updates, I don't expect to receive the 5.0 update until sometime in 2017. That being said, I think this route makes the most sense for Google - it both tackles the internal fragmentation problem of Android, while at the same time allowing Google to more easily bring those same core services to other platforms.