A bunch of new things in AI at Google

A bunch of new things in AI at Google
A bunch of new things in AI at Google

Google held its annual I/O conference. The technology giant has unveiled many new features around its AI models including Gemini, including image and video generation tools. As well as displaying AI-created summaries in its search engine results.

Google held its annual I/O developer conference. From its headquarters in Mountain View, California, the tech giant has highlighted a whole series of new products and features. The company has particularly focused on AI.

“We are fully in our Gemini phase,” Google CEO Sundar Pichai said in his opening remarks. Gemini, Google’s AI model renamed in the spring, is getting a new variant. Gemini 1.5 Pro will be made available in the Gemini Advanced program and is expected to have the longest pop-up window for a chatbot available to individuals. The pop-up length of one million tokens allows Gemini 1.5 to process multiple large documents of 1500 pages in total or summarize 100 emails. In the future, the model should also be able to process an hour of video content or Source texts of more than 30,000 lines.

Version 1.5 of Gemini Flash is also one of the announcements made during I/O. Flash is Google’s fastest model to date and is optimized for high-volume, high-frequency tasks at scale.

Personalized Gems

Gemini-Advanced users will also have the option in the future to create personalized versions of Gemini – called Gems. Gems can be freely configured, for example as fitness coach, sous chef or programming partner.

Gems are created using prompts. Gemini takes these instructions and extends them to create a Gem that meets the user’s specific needs, like the custom GPTs in competitor OpenAI’s ChatGPT. Google does not give a specific date for the introduction of Gems.

Integration into Calendar, Notes and Tasks

Gemini is penetrating deeper and deeper into Google’s product line. At I/O, Google announced new extensions for its calendar, to-do list, and notes apps. For example, it will be possible to photograph your child’s schedule and order Gemini to generate entries in the diary.

AI summaries for Google searches

Google’s search function is also getting more AI capabilities. In the future, the Google search engine will create an AI summary on a topic and display it above regular search results. This overview combines the contents of the search results and links to additional sources.

AI summaries are also based on a Gemini model. It will thus be possible to ask complex questions at several levels. The search will use the model to find information thematically. Instead of a list of links, content will then be grouped under AI-generated keywords.

Initially, AI summaries will only work for English-speaking search queries originating from the United States. However, deployment in other countries should follow.



PREV Summers expected to warm faster than winters with climate change
NEXT Android’s number one problem will be fixed by Google’s artificial intelligence