Google to start embedding AI chatbot in makeover of its search engine
The tech giant is changing the way people get information and curtailing the flow of Internet traffic to websites.
The next phase outlined at Google's annual developers conference includes releasing a new "AI mode" option in the United States.
The feature makes interacting with Google's search engine more like having a conversation with an expert capable of answering questions on just about any topic imaginable.
AI mode is being offered to all consumers in the US just two-and-a-half-months after the company began testing with a limited Labs division audience.
Google is also feeding its latest AI model, Gemini 2.5, into its search algorithms, and will soon begin testing other AI features, such as the ability to automatically buy concert tickets and conduct searches through live video feeds.
In another example of Google's all-in approach to AI, the company revealed it is planning to leverage the technology to re-enter the smart glasses market with a new pair of Android XR-powered spectacles.
The preview of the forthcoming device, which includes a hands-free camera and a voice-powered AI assistant, comes 13 years after the debut of "Google Glass," a product that the company scrapped after a public backlash over privacy concerns.
Google didn't say when its Android XR glasses will be available or how much they will cost, but disclosed they will be designed in partnership with Gentle Monster and Warby Parker.
The glasses will compete against a similar product already on the market from Facebook parent Meta Platforms and Ray-Ban.
The expansion builds upon a transformation that Google began a year ago with the introduction of conversational summaries called "AI overviews" that have been increasingly appearing at the top of its results page and eclipsing its traditional rankings of web links.
About 1.5 billion people now regularly engage with "AI overviews," according to Google, and most users are now entering longer and more complex queries.
"What all this progress means is that we are in a new phase of the AI platform shift, where decades of research are now becoming reality for people all over the world," Google CEO Sundar Pichai said before a packed crowd in an amphitheater near the company's Mountain View, California, headquarters.
Although Pichai and other Google executives predicted AI overviews would trigger more searches and ultimately more clicks to other sites, it hasn't worked out that way so far, according to the findings of search optimisation firm BrightEdge.
Clickthrough rates from Google's search results have declined by nearly 30 per cent during the past year, according to BrightEdge's recently released study, which attributed the decrease to people becoming increasingly satisfied with AI overviews.
The decision to make AI mode broadly available after a relatively short test period reflects Google's confidence that the technology won't habitually spew misinformation that tarnishes its brand's reputation, and acknowledges the growing competition from other AI-powered search options from the likes of ChatGPT and Perplexity.
The rapid rise of AI alternatives emerged as a recurring theme in legal proceedings that could force Google to dismantle parts of its internet empire after a US federal judge declared its search engine to be an illegal monopoly.
In testimony during a trial earlier this month, longtime Apple executive Eddy Cue said Google searches done through the iPhone maker's Safari browser have been declining because more people are leaning on AI-powered alternatives.
And Google has cited the upheaval being caused by AI's rise as one of the main reasons that it should only be required to make relatively minor changes to the way it operates its search engine because technology already is changing the competitive landscape.
But Google's reliance on more AI so far appears to be enabling its search engine to maintain its mantle as the Internet's main gateway - a position that's main reason its corporate parent, Alphabet Inc, boasts a market value of $2 trillion (€1.8 trillion).
During the year ending in March, Google received 136 billion monthly visits, 34 times more than ChatGPT's average of 4 billion monthly visits, according to data compiled by onelittleweb.com.
Even Google's own AI mode acknowledged that the company's search engine seems unlikely to be significantly hurt by the shift to AI technology when a journalist from The Associated Press asked whether its introduction would make the company even more powerful.
"Yes, it is highly likely that Google's AI mode will make Google more powerful, particularly in the realm of information access and online influence," the AI mode responded. The feature also warns that web publishers should be concerned about AI mode reducing the traffic they get from search results.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
France 24
10 hours ago
- France 24
The shrill is gone: AOL to shut down dial-up internet
The raucous sounds of modems establishing their connection to distant servers marked a generation of internet users and still pop up in memes to this day. It also made AOL one of the biggest tech firms of the era, but in the following decades the dial-up connections were steadily replaced with much faster ADSL and then fibre-optic lines. "AOL routinely evaluates its products and services and has decided to discontinue dial-up internet," the pioneering internet service provider (ISP) said in a blog post. AOL did not say how many users would be affected by the September 30 shutoff. CNBC reported that numbers using dial-up had fallen from around 2.1 million in 2015 to just a few thousand in 2021. AOL merged with Time Warner in a 2001 mega-deal worth $162 billion at the peak of the dotcom bubble. After splitting off again, it was sold to Verizon in 2015 for a far humbler $4.4 billion. AOL was merged with another early internet heavyweight, Yahoo, and sold to the investment fund Apollo Global Management for $5 billion in 2021. Its once widely used chat programme AIM, launched in 1997 and beloved of early-2000s teens, was shut down in 2017.
AFP
17 hours ago
- AFP
AI-generated visuals depicting daughters of Thai soldiers killed in border clash spread online
"Don't scroll past this unless you are heartless. Please extend encouragement to the students and daughters of brave soldiers who defended the land," reads part of a Thai-language Facebook post shared on August 5, 2025. The post includes two images of girls donned in student uniforms holding framed pictures of a soldier and a scene depicting a funeral, showing a flag-draped coffin. Image Screenshot of the false Facebook post, taken on August 7, 2025, with a red X added by AFP The post quickly gained traction, with over 800,000 likes and more than 12,000 shares. These posts appeared after Thailand's Chulalongkorn University announced on August 4 that it would provide free scholarships to children and spouses of Thai soldiers, border patrol police, and civilians killed during a military confrontation with Cambodia (archived here and here). Bangkok Post also reported that the Mahasarakham University of the Northeast also offered a similar scholarship while the Ramathibodi School of Nursing said it will waive entrance examination requirements for "Pimook" -- an orphaned daughter of a Thai soldier, according to Thairath News (archived here and here). At least 43 people were killed in the conflict that ended on July 29 after a long-standing dispute over border temples erupted into violence (archived link). A truce was brokered by Malaysian Prime Minister Anwar Ibrahim -- chair of the ASEAN regional bloc -- after cajoling by US President Donald Trump and a team of Chinese mediators. Social media posts shared the same images alongside similar claims on X, Facebook, TikTok and Threads. But the images were made with AI. A reverse image search led to more false posts, but Google's "About this image" feature identified the video as having been made with AI. The ability to detect AI-generated images is based on Google's SynthID technology, which was launched by its DeepMind AI lab in 2023 (archived here and here). Image Screenshot of the Google page indicating the video was made using AI A closer inspection revealed tell-tale visual anomalies of images generated by AI -- including unreadable name tag on the student's uniform and a mismatched flag. The student's uniform features a malformed name badge, a common issue in AI-generated images where supposed legible languages are replaced with gibberish text. Secondly, the flag shown in the funeral photo does not match the Thai national flag -- but instead resembles the flag of Costa Rica, which features similar colors with a different stripe arrangement (archived link). Image AI-generated photo falsely shared by social media posts, with elements made by AFP to highlight visual anomalies AFP has also debunked other misinformation related to the conflict between Cambodia and Thailand.
Euronews
20 hours ago
- Euronews
Google's newest AI model acts like a satellite to track climate change
Satellites orbiting Earth to gather images and measurements of land, forests, cities, and coastal waters have helped scientists understand our planet. However, since these images come from many different sources, it can be difficult to combine them into a single picture. Google's artificial intelligence (AI) unit, DeepMind, recently announced an AI model called AlphaEarth Foundations that can create highly detailed maps of our world in nearly real time. AlphaEarth Foundations functions like a 'virtual satellite' that maps the world 'at any place and time,' Christopher Brown, a research engineer at Google DeepMind, said at a press briefing in July. 'Whether they are monitoring crop health, tracking deforestation, or observing new construction, [researchers] no longer have to rely on a single satellite passing overhead. They now have a new kind of foundation for geospatial data,' Google DeepMind wrote in a statement last month. The system combines trillions of images from dozens of public sources, including satellite images, radar scans, laser-based 3D mapping, and climate simulations. It maps the entire planet's terrestrial land and coastal waters. Google said the model can generate accurate enough data about an ecosystem down to an area of 10 square metres. Data from AlphaEarth Foundations takes up far less storage space than similar AI systems, which makes large-scale analysis more practical, the company said. During AlphaEarth Foundations's initial testing of data from 2017 to 2024, it beat similar AI models in identifying land use and estimating surface properties, with an average error rate that was 24 per cent lower, according to a paper published by DeepMind. Google hopes it will help researchers study changes across the planet for food security, deforestation, urban expansion, and water resources. Why scientists want this level of detail and how AI can help AlphaEarth Foundations is part of a growing trend in environmental science where AI turns the constant stream of satellite observations into practical tools for studying the Earth. High-resolution, regularly updated data help researchers measure environmental changes precisely and understand what is driving them. They can be used to track the effects of climate change, plan conservation, and manage resources like water and farmland. For example, in 2020, scientists at NASA and the University of Copenhagen mapped 1.8 billion individual tree canopies in the Sahel and Sahara regions of west Africa, using AI trained to recognise trees in satellite images. Without AI, this would have taken millions of people years to complete, according to the study authors. Since 2022, NASA's Surface Water and Ocean Topography (SWOT) satellite has been taking high-definition measurements of oceans, lakes, reservoirs, and rivers over 90 per cent of the world's surface. NASA's Jet Propulsion Laboratory (JPL) said it could observe water on Earth's surface with 10 times greater resolution than existing technologies. Meanwhile, the European Space Agency (ESA)'s EarthCARE satellite launched in 2024. It is studying how clouds and airborne particles in the atmosphere impact the Earth's temperature. Much of Google DeepMind's data also comes from long-running missions from NASA and ESA, such as the Landsat and Moderate Resolution Imaging Spectroradiometer (MODIS) satellites and the Sentinel fleet, all of which monitor vegetation, coastlines, water bodies, snow, and ice. Google says its model has already been tested by more than 50 organisations around the world for ecosystem monitoring and urban planning. For example, an environmental initiative in Brazil known as MapBiomas is using AlphaEarth Foundations' data to better understand agricultural and environmental changes, including in the Amazon rainforest. The model's annual datasets have given the team 'new options to make maps that are more accurate, precise and fast to produce, something we would have never been able to do before,' said Tasso Azevedo, the founder of MapBiomas, in Google's statement. Google says it is releasing the dataset through Google Earth Engine, Google's environment data platform, to encourage further research.
