There was a lot of talk about artificial intelligence at today’s Google I/O keynote address, with company CEO Sundar Pichai diving straight into the Mountain View tech giant’s prospects with all things AI. Google’s headstrong charge into the world of generative AI and large language models over the past couple of years has seen no slowdowns, and we finally know why the Pixel 8a was launched so discreetly this year – Google wanted to focus solely on software during the event.
With that in mind, we wanted to list five big highlights with Google’s new announcements for AI this year. There’s admittedly a ton of them mentioned throughout the two-hour event stream (which you can check out here in full), so we’re diving into some of the more impressive AI announcements for Google I/O 2024. Let’s take a look!
AI is in Full Swing
Google announced a ton of updates to its Gemini family of large language models, including a new lightweight model called 1.5 Flash, which according to the company is optimized for speed and efficiency. Meanwhile, Gemini 1.5 Pro, the company’s most powerful model for general performance, has also been improved with a longer context window and better performance on a variety of tasks. In addition to all of this, Google also announced Gemma 2, the next generation of its open models for responsible AI innovation, in addition to news on Project Astra, its effort to build universal AI agents.
AI Overview Lands in North America
Google also announced that Search is getting the new AI Overview feature, which will be available for users worldwide starting in North America this week. The AI Overview function lets users ask complex, multi-part questions and get faster results powered by Google’s AI model, Gemini. This includes anything from asking Search to find a very specific restaurant with detailed questions such as travel time, pricing and facilities, as well as the ability to help users plan trips, recipes, and more. The feature plays nicely with Workspace, featuring integration for Docs and Gmail for example.
“Ask Photos” is Here
The Google Photos app is getting a new AI-powered feature called “Ask Photos” that uses Google’s Gemini model to understand user queries and requests. You can ask questions about your photos, like finding the best picture from each national park you’ve visited, or figuring out forgotten details from past events by asking about things like birthday party themes based on decorations in the photos. Google assures users their privacy is protected, although it should be noted that this feature is still experimental at the moment.
Circle to… Study?
When Google announced Circle to Search several months back, it introduced a more streamlined and convenient way of performing image searches at anywhere and anytime on your phone. With that said, the company has added new functionality to the feature. More specifically, Circle to Search will now work with mathematical equations, and will even be able to handle graphs and charts – homework has never been easier, kids.
Taking Spam Prevention One Step Further
Pixel phones are known for their built-in ability to screen and detect unwanted spam calls, and Google seems to be building on this standout feature with one of the announcements made during I/O 2024. Using Gemini’s AI capabilities, supported Android devices will soon be able to warn users if a caller might be scamming them, based on certain word prompts and speech patterns.
All in all, Google I/O 2024’s opening was very interesting to see, and we should expect more updates from Google in the coming weeks and months with their AI prospects.
