Google launches fully AI-powered search engine in ‘reimagining’ of web tool

Google is to begin offering a fully artificial intelligence-powered version of its search engine in what it calls a “total reimagining” of one of the web’s foundational features.

The US tech giant will begin the public rollout of AI Mode in Search in the United States on Tuesday, the company confirmed at its annual I/O developer conference.

The new feature, which will appear as a tab at the top of the search engine, will allow users to ask longer, more complex queries as well as follow-up questions to dive deeper into a topic, Google said.

It builds on the AI Overviews the company already places at the top of many search results, which are AI-generated summaries in response to a query, alongside links to sources.

The firm’s chief executive Sundar Pichai said the update was a “total reimagining of search” that would use “advanced reasoning” to think before it answered a user’s query.

Mr Pichai said early testers had been asking queries “two to three times, sometimes as long as five time the length of traditional searches”.

“It’s been a pretty exciting moment for Search,” he said.

“People are excited. It’s made the web itself more exciting. I think people are engaging a lot more across the board and so it feels like a very positive moment for us.”

The announcement came amid a flurry of updates from the company during its I/O conference, with tools utilising the firm’s Gemini AI dominating the new products and services unveiled.

This is despite Google and others in the tech sector having a number of high-profile issues with previous AI products over the last year, with generative tools returning inaccurate or misleading results on a number of occasions.

Concerns also remain about the impact of the technology on the jobs market because of its potential to replace human workers, the data privacy implications of AI being trained on data scrapped from the public web and safety concerns around AI’s ability to supercharge misinformation, help cybercriminals become more sophisticated or being used to create more dangerous weapons.

Google’s latest product revelations also come as the company faces growing questions about the future of its online search business, with some reports suggesting search engines are now being used less as more users turn to AI chatbots – such as OpenAI’s ChatGPT – instead.

Other notable updates at Google’s event included real-time translation between English and Spanish being introduced to Google Meet video calls, with AI dubbing the translation over the top of the speaker so the other person can understand what is being said instantly.

In addition, Google announced Project Mariner, an AI agent which can be instructed to carry out tasks on the web, such as online shopping or restaurant booking, on behalf of the user.

AI agents, tools which can are capable of carrying out specific tasks autonomously, have been earmarked by many in the industry as the next step in AI tool evolution.

During its I/O keynote, Google also announced it was launching new personal context tools, which use AI to lift relevant data from other Google apps – such as Docs or Calendar – and apply it to its existing Smart Replies tool in Gmail to create personalised automated replies that sound like the user’s style of writing.

“What all this progress tells me is that we are now entering a new phase of the AI platform shift,” Mr Pichai said.

“We are starting to bring agent capabilities to Chrome, Search and the Gemini app.

“It’s a new and emerging area, but we want to get it in the hands of people so that we can explore how to bring the benefits of agents to as many users as possible and get feedback from the broader ecosystem.

“We want to make it as useful as possible, to fit your reality. That’s why we are working on something called personal context with your permission.

“Gemini models can use relevant context from across your Google apps in a way that’s private, transparent and fully under your control. You can choose to connect what you want and turn off what you don’t want.

“One example we are introducing is personalised smart replies in Gmail. It takes Smart Reply a step further, pulling relevant information from Google Docs and past emails, it matches your tone and style, and generated the email as you would have written it, personalised.”

The array of announcements also included a video AI-generated video and audio tool, as well as a new product to help filmmakers called Flow.

These new products come despite ongoing concern from the creative industries about the potential impact of AI-generated content on the the film, TV and music sectors.

Elsewhere, the company also unveiled a new video conferencing device called Google Beam, which uses AI and six cameras to create a 3D video experience which Google said feels closer to natural, in-person conversations.