Google I/O 2024: From Gemini 1.5 Flash to new AI options in Android, try the highest bulletins

Listed here are prime bulletins made by Google I/O occasion:

1) Gemini 1.5: 

Google rolled out a brand new Gemini 1.5 Flash language mannequin on Tuesday, which it touted to be extra light-weight than the favored Gemini 1.5 Professional mannequin whereas being quick and environment friendly. The brand new AI mannequin is able to multimodal reasoning and comes with a bigger context window of 1 million tokens.

Google says Gemini 1.5 Flash mannequin has been optimized for dealing with high-frequency duties the place response time issues probably the most. 

2) Gemma: 

Google additionally unveiled the subsequent era of Gemma AI mannequin with claimed enhancements in efficiency and effectivity. The brand new mannequin has been optimized by TPUs and GPUs and comes with 27 billion parameters. 

Google introduced its first imaginative and prescient language mannequin, PaLI-3, to the Gemma household of fashions. The brand new fashions will likely be launched in June. 

3) Ask Images: 

Google is offering a Gemini AI-powered improve to Google Images which can permit customers to recall a particular image based mostly on a single immediate by the customers. The brand new characteristic referred to as ‘Ask Images’ takes benefit of Gemini’s multimodal talents to know the context and topic object of an image and furnish the picture requested by the consumer.

For example, customers can ask Gemini questions associated to a particular occasion, object or particular person and the AI assistant will immediately scan by way of the consumer’s gallery to scan for the corresponding photographs.

4) AI in Android: 

Whereas Sundar Pichai and Co. did not reveal a lot details about what’s coming within the Android 15 replace, Google did affirm it’s bringing various AI capabilities to Android because of Gemini. Circle to Search, the favored AI characteristic seen on Pixel and Samsung gadgets, will now assist college students full their homework by offering them with step-by-step directions to resolve a spread of physics and math issues. 

The corporate additionally introduced that Gemini Nano, it is light-weight on-device giant language mannequin, will get multimodal capabilities which means it is going to be in a position to perceive data in sights, sounds and spoken language, other than simply textual content. Gemini Nano can also be getting a brand new security characteristic that may present actual time alerts throughout a name if the AI detects a dialog sample generally related to scams.

Furthermore, Google can also be offering Gemini a significant enhance by permitting the AI assistant to overlay on  prime of various apps on the display. The brand new characteristic will permit customers to make use of Gemini for duties operating on their display. For example, customers will in a position to ask particular questions on a YouTube video or drag Gemini generated photographs immediately into Gmail.

5) VEO, the Sora different we wished:

Google challenged the dominance of OpenAI’s Sora with the introduction of its newest text-to-video era mannequin, Veo. The brand new mannequin is succesful to producing 1080p decision movies with a spread of cinematic and visible kinds. Veo will even present customers with a larger stage of management owing to its understanding of cinematic phrases like ‘timelapse’ or ‘aerial photographs of panorama’. 

6) AI search: 

Google can also be revamping its search expertise with AI, because of a brand new custom-made Gemini mannequin made particularly for this goal. The search big will now present fast summaries on any specific search matter with AI overviews, a characteristic which was in experimental phases until now with Google Search Generative Expertise (SGE). 

The corporate will even roll out a brand new AI-organized search outcomes web page that may categorize the outcomes beneath ‘distinctive’ AI-generated headlines. The brand new replace will likely be out there to customers within the US from at present and the brand new search web page will initially be rolled out for eating and recipe searches, adopted by others like motion pictures, music, books, motels and purchasing.

7) Mission Astra, Google’s reply to GPT-4o: 

Google went toe to toe with OpenAI on Tuesday with the introduction of Mission Astra, a brand new AI agent referred to as Mission Astra. In a video showcased by Google, the AI assistant was in a position to establish the objects within the room, acknowledge and clarify a particular a part of the code, determine its precise location by wanting exterior the window, discover the consumer’s glasses and even discover inventive names for a canine. Furthermore, Google showcased the usage of Mission Astra by way of a smartphone or sensible glasses, suggesting that there might be a significant Gemini-powered revamp to Google Lens down the highway.

Milestone Alert!
Livemint tops charts because the quickest rising information web site on the earth 🌏 Click on right here to know extra.

Unlock a world of Advantages! From insightful newsletters to real-time inventory monitoring, breaking information and a customized newsfeed – it is all right here, only a click on away! Login Now!

Printed: 15 Could 2024, 10:25 AM IST