Google Live in Paris: LaMDA-Powered Search Features Unveiled

At Google’s ‘Live in Paris‘ event, the tech giant revealed a series of AI innovations. Google has incorporated AI into everything from maps to museums to “reimagining how people search for, explore and interact with information”.

Let us take a look at the key highlights of the event-

Google Search & Lens

Google adds its LaMDA-based AI chatbot Bard to search. Now you can get a summary of answers to your queries, similar to ChatGPT. 

“Search is our biggest moonshot”, said Prabhakar Raghavan, senior vice president at Google. 

Download our Mobile App

Bard has opened up to  ‘trusted testers’ for feedback before being expanded to the public. 

Google Lens reached a record of enlisting over 10 billion times every month. In the coming months, you can use Google Lens to search what you see in photos or videos across websites and apps on your phone. The multi-search feature is now globally available in 70 languages. Currently being rolled out for image search results, you can use it to find real-world objects in different colours.

Translate

Over 1 billion people use Google’s translate feature. With a new ‘Zero Shot Machine Translation’ method, over 24 new languages are added. Google also said many Ukrainians seeking refuge had used Translate to fathom new places. 

By utilising AI and Google Lens to organically translate words for common household objects, the firm claims to be actively working to preserve endangered languages.

Maps

Google introduced several new AI features to Maps. Maps with the ‘Immersive View’, are rolling out now in over 1000 cities like LA, NY, and London.

With its improved “Search with Live View” feature, which blends AI and AR, you can use your phone’s camera to visually locate objects in your area, such as restaurants, ATMs, and transit hubs.

The ‘Indoor Live View’ is now being used for airports, shopping malls, and train stations. It uses AR arrows to indicate important spots like baggage claim, elevators etc, in limited cities right now, but with its “largest expansion”, it will be available in more than 1000 cities “in the coming months”.

Another new AI feature is added for electronic vehicles whereby Maps can instantly filter charging stops and give you the “best charging stop” and also “very fast charging” stops.

Read more: Google, the Tech-Savvy People Person

Arts and Culture

Google unveiled a number of new AI-powered arts and culture tools, including one that lets you look up well-known artwork by thousands of artists and showcase its deep details. Blob Opera also made a special entry. 

“Although we are 25 years into the search, I dare say that our story has just begun”, said Prabhakar reiterating the company’s motto towards “Responsible AI”. 

The company is also planning to hire more developers to advance ‘Generative AI’ in the coming months.

No mention of Anthropic was made. Although the event did not have any huge announcements, and it mostly seemed like a recap, viewers were quick to spot an error in Bard’s results during the live demo. 

During the demo, Google tweeted a brief GIF clip showing Bard in action and called the Bard a “launchpad for curiosity” that would help make difficult subjects simple. Prompt: “What recent findings from the James Webb Space Telescope (JWST) can I share with my 9-year-old? Bard offered several responses, one of which claims that the JWST was used to capture the first images of exoplanets or worlds outside the solar system. That is incorrect. According to NASA, the Very Large Telescope (VLT) of the European Southern Observatory captured the first images of exoplanets in 2004.

Microsoft held a close door conference last day where it released an upgraded AI-powered Bing search and Edge web browser, based on a new LLM but seeking findings from OpenAI’s GPT 3.5.  

Read more: The Battle for AI Supremacy Begins