Google will overhaul the service that built the company into a $2.1 trillion technology juggernaut — its online search. The curtain has just fallen on a two-hour keynote with high-profile executives at Google I/O, the annual developer conference held by the company to announce new software features and hardware.
And this year was all about Artifiticial Intelligence (AI).
Google unloaded a barrage array of new AI-powered products, including new video and image creation tools. But the biggest announcement of the night was confirmation that Google would start to rebuild its search engine around generative AI, powered by its controversial Gemini AI model.
Google has integrated AI into its search engine on laptops and desktop PCs, as well as mobile phones and tablets to generate recipes, meal plans, travel itineraries, and written explanations based on a single or multi-faceted query
GOOGLE PRESS OFFICE
"This is a moment of growth and opportunity," CEO Sundar Pichai told reporters when asked whether bringing AI into search posed a risk to Google’s hugely profitable business.
Google I/O is being held for two-days at the Shoreline Amphitheatre in Mountain View, California — a stone's throw from Google's headquarters. The opening keynote kickstarted less than 24 hours after a shorter showcase from AI rival OpenAI. The company, which has enjoyed an eye-watering $10 billion investment from Microsoft to use its AI models in its paid-for Microsoft Copilot feature in Windows 10 and Windows 11, OpenAI demonstrated how ChatGPT could voice answers with human-like intonation to any written or visual prompt.
The startup's CEO, Sam Altman, wrote that OpenAI had delivered software that "feels like AI from the movies," with many comparing ChatGPT-4o's ability to flirt and banter to Academy Award-winning film Her.
During the keynote, which is available to stream on YouTube, Google confirmed it would expand a small test conducted in the UK that brought generative AI responses and suggestions to its search results. For the first time, this AI-powered search would be coming to users in the United States.
Google says the feature should “take more of the legwork out of searching”.
Rather than presenting a list of links to webpages that might have the answer to your query, the overhauled Google search will use AI to attempt to present a fully-formed answer to your query — pulling from resources across the internet to do so.
For example, if you search "how to clean a fabric sofa," you'll be presented with a block of text with a number of different tips — with Google Gemini presenting links to the webpages with the original suggestions at the bottom of the new panel, which the US company refers to as an AI Overview.
Google's latest AI model will put together a comprehensive answer, drawing from resources across the web, which will be shown in a carouse for further reading. Scrolling further down the page, you'll finally come across the list of 10 blue links that has characterised Google search since its launch in the late 90s
GOOGLE PRESS OFFICE
If you continue to scroll down the page, you'll find Google Shopping results, the People Also Ask... prompts, and only then, the list of blue links that have characterised Google search since its launch back in September 1998.
Users have the chance to tweak just how much information is shown in the AI Overview, toggling between a dense text-heavy response with plenty of detail, or something snappier for those short on time.
Google's AI-enhanced new search engine is capable of breaking down longer, multi-part queries — so you're not limited to asking a single question. In one example, a Google executive asked the search engine to find gyms within a 10-minute walk, organised by the best reviewed, and showing all intro offers. The AI model was able to break down all parts of this multi-faceted question in a single search result.
Google lets you choose how much information is delivered in the AI Overviews that appear at the top fo search results, toggling between Original and Simpler
GOOGLE PRESS OFFICE
In the future, Google said users would be able to submit search queries via a video.
In a demo of this still-in-progress feature, Google showed how it could help diagnose a broken record player when the owner uploaded a short video demonstrating the problem they were experiencing. The company said these new innovations represented the start of a “new era for search”, powered by generative AI, with chief executive Sundar Pichai calling it the “most exciting era of search yet”.
In a blog post about the rebooted Google search, Vice President of Google Search Liz Reid wrote about a number of examples that showcased how you might be able to use this new AI-enhanced search engine. Reid wrote: "With planning capabilities directly in Search, you can get help creating plans for whatever you need, starting with meals and vacations. Search for something like 'create a 3 day meal plan for a group that’s easy to prepare,' and you’ll get a starting point with a wide range of recipes from across the web.
"If you want to change anything, you can easily ask for whatever adjustments you need, like swapping dinner to a vegetarian dish. Just like that, Search will customize your meal plan. You’ll be able to quickly export your meal plan to Docs or Gmail."
"This is just a glimpse of how we’re reimagining Google Search — combining the very best of Search today with the Gemini model’s advanced capabilities. Soon, Google will do the searching, simplifying, researching, planning, brainstorming and so much more," the Google executive promises.
The Californian company will start rolling out AI Overviews to all users in the USA later this week.
Elsewhere, Google I/O saw the US firm unveil a new video creation tool called Veo, capable of turning text prompts into longer-form videos, as well as Imagen 3, an AI image creator.
As part of its creative tools development, Google said it was working with musicians including Wyclef Jean and songwriter Justin Tranter, who had created new music with help from Google’s Music AI Sandbox to create new sounds, as well as with filmmaker Donald Glover who had been using the firm’s text-to-video AI tools.
The new creation tools come amid ongoing concerns around AI-generated content, particularly deepfake images and videos which have been used to spread misinformation.
In a blog post on the new tools, Google said: “We’re mindful about not only advancing the state of the art, but doing so responsibly. So we’re taking measures to address the challenges raised by generative technologies and helping enable people and organisations to responsibly work with AI-generated content.
“For each of these technologies, we’ve been working with the creative community and other external stakeholders, gathering insights and listening to feedback to help us improve and deploy our technologies in safe and responsible ways.
“We’ve been conducting safety tests, applying filters, setting guardrails, and putting our safety teams at the centre of development. Our teams are also pioneering tools, such as SynthID, which can embed imperceptible digital watermarks into AI-generated images, audio, text and video. And starting today, all videos generated by Veo on VideoFX will be watermarked by SynthID.
“The creative potential for generative AI is immense and we can’t wait to see how people around the world will bring their ideas to life with our new models and tools.”
LATEST DEVELOPMENTS
Google also used the opening keynote at I/O to unveil the latest addition to its family of Gemini 1.5 AI models, known as Gemini 1.5 Flash that's much faster and cost-effective to run, as well as a prototype system called Project Astra that's capable of recognising almost anything in the viewfinder of your smartphone — allowing users to chat with AI about the objects captured in real-time.
Finally, Google CEO Sundar Pichai showcased a number of new ways Google was integrating Gemini into its popular apps, including a new Ask Photos tool for the Google Photos app, which used text prompts from users to find specific images or create photo collections, or asking Gemini to create a summary of all recent emails from a specific sender or on a topic if a user needed to catch up.
