Google unveils new search bar, smart glasses as it ramps up the AI wars
Things are moving at breakneck speed in the race to create AI products that people will actually find useful. Although Google was once seen as the slow bearer having to drag the old business of advertising and search into the new age, it has recently stepped up to what appears to be a decisive leader.
The company leveraged its Android business, close relationship with Apple, massive user base, and access to sensitive personal information through existing apps and services to bring Gemini AI to every aspect of digital life. It showed that advertising revenue in the last quarter actually increased as a result of artificial intelligence. At Google I/O, the developer conference held on Wednesday morning, it was stated that the number of Gemini users doubled in a year and reached 900 million. Here are some of the ways he recommends continuing to expand.
new models
The most far-reaching new announcement at I/O was Gemini 3.5, Google’s new family of pioneering AI models; The web giant also promised to increase speed and efficiency while empowering more autonomous use for AI agents beyond text chat and rendering.
The first model of the new family, the 3.5 Flash, is out now and currently powers the Gemini app and AI Mode in Search. Google showed benchmarks showing that the model is four times faster than other leading models, but in terms of encoding it is still more powerful than heavyweight models such as the Gemini 3.1 Pro, which was released in February.
Some research He noted that average spending on enterprise AI is growing much faster than expected, with some companies exhausting their annual token budgets in less than half a year. Google CEO Sundar Pichai said that the lower cost of 3.5 Flash, which is powerful enough for almost all tasks, will attract the attention of these companies.
“The top companies process around 1 trillion tokens per day. If they shift 80 percent of their workloads from other leading models to 3.5 Flash, they will save more than US$1 billion ($1.4 billion) per year,” he said.
Also introduced at I/O was the Gemini Omni, a new family of models designed to accept any combination of input modes (e.g. text, audio, code, image, video) and output in any mode. For now, Omni Flash will only output video; A Google rendering shows a user requesting a clip that uses the camera style of one clip and the visual style of the other, along with a character created from a photo uploaded. The mockup will only be available to paying subscribers on the Gemini app, but it will be free with an update to YouTube Shorts this week.
Artificial Intelligence Search
Google claims it has made the biggest change to its web search box in 25 years, as it is now formatted to take advantage of 3.5 Flash speeds. The box now expands as you type, encouraging you to ask long and detailed questions and making suggestions as you go. You can add images, files, or videos to reference, or point to an open tab in Chrome. You’ll still get a list of links as web results below the AI chat and output, but it will be more relevant due to context, the company said. This change is now live.
By the end of this year, the search box will also be able to code its own visualizations, tables, and graphs to explain concepts interactively. Paying subscribers will also be able to create mini-apps that they can return to directly in Search; Google said this is designed to eliminate repeat searches. For example, you can create a mini-app that always shows you what movies are showing at a particular theater or creates a custom exercise routine taking your local weather into account.
Users can now also choose to connect AI Mode in Search to Personal Intelligence, Google’s platform that allows its AI to crawl your data in Gmail, Google Photos, and more.
Content credentials
To counter the consequences of ever more powerful image and video rendering models, Google has introduced some new tools for transparency. The SynthID watermark, an invisible piece of data that can be read by machines to determine whether something was done by AI, has been placed on more than 100 billion images and videos, as well as audio assets that are 60,000 years old, he said. It has worked to encourage other companies to embed SynthID into assets created by its tools, and announced at I/O that OpenAI has also agreed to use it.
The next step is something Google calls Content Credentials verification. It and many other companies use the C2PA credential standard when creating media; For example, when you look at a photo in Google Photos, it will be able to show you the make and model of the camera that took it. Google said it is rolling out a feature to Gemini, Search and Chrome that allows users to ask questions about the supply of any media they see and get information about how that media was captured or created and whether it was edited by AI. Google said it is advocating global standards that mean images taken by a phone or camera (without editing by AI) will be easy to verify wherever they are published.
smarter glasses
Last year, Google announced Android XR, a new platform developed with Samsung and Qualcomm that will put Google AI on your face through smart glasses. This year the company showed a little more of what that would actually look like.
The first wave of eyewear to be released in the coming months will be launched by fashion brands Warby Parker and Gentle Monster. They’re similar to Meta’s original Ray-Ban smartglasses in that they have speakers, microphones, and cameras, but no screens. Users will be able to tap the frames to summon Gemini and ask questions about everything they see. The glasses will also give directions to Google Maps, take photos and videos, play music and podcasts, answer calls, transcribe messages and notifications, and perform live translations.
Later this year, Google plans to release Project Aura, a product developed together with Xreal. These extended reality glasses have built-in displays that will layer up to five applications simultaneously over your real-world view. Like Apple’s Vision Pro, these are connected by a cable to a phone-sized processor box that you can put in your pocket or hang around your neck with a lanyard. However, unlike the Vision Pro, they look like sunglasses. Project Aura even lets you connect external devices to use the glasses as a monitor; so you can mirror any phone, laptop, or even a gaming console like a Steam Deck to the screens.
Your own representative
While Google promises a future where everyone will have an army of AI agents at their beck and call, performing digital tasks on the internet in the background, the most evocative real-world version of this will initially be available only in the US and only to users with the most expensive Google subscription.
Gemini Spark is an artificial intelligence agent that runs online 24 hours a day on dedicated servers. You talk to it through the Gemini app, but it can still work when you lock your phone or close your laptop. It will be able to connect to Google apps like Gmail and Drive, and in the future, it will be able to browse the web and connect to other apps you use. Google cited examples such as the agent periodically reviewing emails to provide a daily summary of important dates from a large number of messages from an elementary school, or reviewing credit card statements to flag any irregularities.
Spark is designed to ask for your permission before performing “high-risk” actions on your behalf, the company said. He showed off an upcoming Android feature that will give your agent a spot at the top of your phone so you can see what he’s doing and if he needs anything. While Spark is only available for US Ultra subscribers, it plans to eventually roll out more widely.
Google also showed off agents in Search, which will be available to subscribers later this year. You program your agents by defining the types of searches you want to make (for example, homes coming on the market in a certain area at a certain price range or new sneakers from a certain brand), and the agent will go over the issue in the background and keep you updated.
Media documentation
AI is already deeply embedded in Google Workspace, which has 4 billion users. But at I/O, the company showed off some of the ways it’s trying to fundamentally reimagine the way its users create on the platform. The most impressive demonstration was Docs Live, a new way to draft entire documents by verbally brain dumping them into Gemini. The outline changes as you continue talking, whether you have new ideas for how you want to fix something the AI has done or pull information from the web and your personal data if you’ve given permission. A similar Live feature is coming to Gmail and Google Keep, too, pulling relevant information from your inbox or turning your shower thoughts into lists and reminders, respectively.
Less useful was Pics, a new app that lets you create and edit images using AI. It’s positioned as an easy way to replace visual assets, remove elements, or edit text for use in apps like Slides and Drive. Like other workspace tools, it also allows teams to collaborate on projects together.
Artificial Intelligence Shopping
Finally, Google laid the foundations for how online shopping could work in an internet where we navigate by talking to agents rather than clicking on links. Universal Cart is a tool that can appear in apps like Search, Gemini, Gmail, and YouTube, allowing you to add any product you see to your list. The cart finds purchasing options and can notify you about restocks or discounts. Google’s demo even showed the cart notifying users if an item is at its lowest price in 60 days or if added computer parts are incompatible with each other.
Users can remain in Google’s cart to checkout and let a representative handle their purchase, or you can transfer the items to your chosen retailers’ website. Universal Cart will be released in the US later this year.
Get news and reviews on tech, gadgets and games in our Tech newsletter every Friday. Sign up here.
Tim Biggs is a writer covering consumer technology, gadgets, and video games.Connect with: X or email.
