Stealing back the AI crown starts with Google reinventing its most important product: Search. Now — or soon, as it's rolling out over time — at the top of the screen, you'll find a generative AI-powered area called AI Overview that offers summarized responses to queries along with links to the sources.
It's a major departure for Google and could impact everything from the company's advertisers, its main source of revenue, to websites that rely on the firm for search traffic, such as travel platforms and media organizations.
While the Search changes are easily the biggest updates to Google's public-facing products, they weren't the only consequential announcements at the show. Google also showed off a new AI-powered prototype assistant called Project Astra.
During a pre-recorded stage demo, the platform, which views content streamed through your smartphone camera, was able to identify objects, solve programming problems, and more.
Unlike Google Assistant and Apple's Siri, Project Astra offers a more creative, conversational tone that sounds more human-like. It's even able to understand when you interrupt it, stopping to listen without getting mad. Not human-like!
Google co-founder Sergey Brin, who was touring the I/O demo areas, said the technology behind Project Astra would have been helpful for the company's canceled Google Glass, adding that the headset was 10 years ahead of its time.
But Google's announcement came a day after OpenAI debuted its own GPT-4o model, which, based on pre-recorded and live-streamed demos, is capable of many of the same tasks as Project Astra. What's more, OpenAI is making parts of its GPT-4o model available to users for free, while Google says it will charge for access to Project Astra.
Google also revealed AI updates to its Workspace productivity suite, including bringing Gemini to the Gmail mobile app. The feature allows users to quickly summarize emails and write contextual replies based on messages.
Android, Google's mobile operating system, also received its share of generative AI improvements, including scam call detection and the ability to add AI-generated images and more to text messages and Gmail.
As of now, that puts Google far ahead of Apple, which has yet to introduce generative AI features to its iOS operating system for the iPhone. But that is largely expected to change when the company hosts its annual WWDC event in June.
Even if Apple debuts new generative AI offerings, it could still be a boon for Google since, according to Bloomberg, Apple is working to use Google's Gemini to power the iPhone's AI capabilities. But Apple is also in talks with OpenAI. Folks, we've got an AI bake-off.
For now, Google appears to be back on top. While GPT-4o cooled some of the excitement around Project Astra, Google's AI Overview for Search will expose far more people to its generative AI technologies than Microsoft or Apple when the latter shows its hand.
But with Microsoft preparing to host its Build developer conference next week, Google's lead could be short-lived. |