Google reveals upcoming AI features while Bard makes a blunder

9 Feb 2023

Image: © Heidi/Stock.adobe.com

Google has shared plans to use AI in various features such as Search, Translate and Maps, but a mistake from the Bard chatbot has shaken up company shares.

Google revealed more of its hand in AI and showcased how it will be integrated into various services in the coming months.

At a live event in Paris yesterday (8 February), the tech giant shared a new Immersive View maps feature that has already rolled out to five cities. Google said this feature combines Street View and aerial images to create a digital model of a location.

This model also adds information such as weather, traffic and how busy certain locations are. Google said an advanced AI technique called neural radiance fields is used to transform the original images into 3D representations.

This feature is available in London, several US cities and Tokyo, with plans to roll out to more cities including Dublin in the coming months.

Google also plans to update its popular Translate service, which lets users type in one language to receive a rough translation in another.

The company said AI will allow the Translate feature to provide more contextual translation options, such as if a user wants to order bass for dinner, compared to playing bass at a music event.

With machine learning, Google said users will soon be able to use its Lens image recognition software to blend translated text into complex images in real time.

Searching with images and multisearch

Earlier this week, Google shared plans to add AI into its Search engine, in a similar move to Microsoft with Bing.

Now, Google has revealed new AI-powered features it will launch to improve its dominant service. The tech giant plans to give users of Lens a bonus feature by letting them instantly search from an image on their screen.

For example, if a user is sent a video or photo from a friend in Paris, they will be able to click on a landmark to search more about it. Google plans to launch this feature in the coming months.

Another update that has been already launched is Multisearch, which lets users search with an image and text at the same time. For example, you can use an image of a circular table and add the word “rectangle” to find similar versions in a different shape, to fine tune the search.

This feature is available globally on mobile, in all languages and countries where Lens is available.

“We’re creating search experiences that are more natural and visual – but we’ve only scratched the surface,” said Google Search VP Elizabeth Reid  “In the future, with the help of AI, the possibilities will be endless.”

Google plans to expand the Multisearch feature to have the ability to use it “near me”, to let users find a specific type of location or service that is nearby. This is available in the US and will be rolled out globally in the coming months.

The company also plans to make Multisearch available for any image a user sees on the search results page on mobile devices.

Bard plays the wrong note

Despite the large list of upcoming features Google has revealed, a blunder of its recently announced chatbot, Bard, has impacted its market value.

A promotional video of Bard in action contained a clip where the chatbot said the James Webb Space Telescope was the first satellite to take pictures of a planet outside our solar system. This mistake was first spotted by Reuters.

Google’s parent company Alphabet lost roughly $100m in market value after its shares dropped by around 9pc, Reuters reports.

The blunder marks a bad start for the upcoming chatbot, as it seeks to challenge the sudden popularity of OpenAI’s ChatGPT.

This advanced chatbot has been reported to make errors of its own, however. A popular Q&A site banned ChatGPT answers from its site last year due to a wave of incorrect answers allegedly flooding the site.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Leigh Mc Gowran is a journalist with Silicon Republic

editorial@siliconrepublic.com