AI was at the forefront of Google I/O, but so was Android

With Google I/O underway, you might have been one of the many who tuned into the conference’s keynote presentation, which highlighted Google’s current projects into AI including Gemini, Deepmind, and more. While Android was briefly mentioned during the event, it wasn’t totally pushed to the sidelines – in fact, Google revealed some interesting updates to AI for Android devices.

First up is Circle to Search, which can now help students with their homework. The feature was originally introduced as a more streamlined method of doing image searches, but can now also handle math equations, graphs, diagrams and more. Google attributes this upgrade to the “LearnLM” AI model.

Google also announced new contextual awareness to Gemini on Android – the assistant will now be able to better sense when you’re in the middle of doing something. For example, users can summon Gemini’s overlay on top of the app they’re using – this can allow users to drop generated images into messaging apps, find specific information from a YouTube video, and even “ask” a PDF file to quickly get answers.

For Gemini Nano on Android, Google is planning to bring multimodality to Gemini Nano on Pixel devices later this year, which will allow phones to process text input, and understand more information in context such as sights, sounds and spoken language.

There’s also TalkBack, which will now be able to fill in missing information such as elements in a photo or images found while online shopping. This will work offline, thanks to on-device processing. Google is also testing a feature that will enable Gemini Nano to give users realtime alerts should they receive suspected scam calls, more of which will be revealed later this year.

With all that said, Google is undoubtedly going all-in on AI, although given that Android is among their largest software products, it’s interesting to see how these new prospects will evolve within the year and beyond.