Google launches creation tools and “transforms” search with personalized AI

Thank you for reading this post, don't forget to subscribe!

Last Tuesday (14), the Google presented several new features involving generative artificial intelligence, during Google I/O, an event taking place in San José, California. In terms of creation, the highlights are Veo, a model that allows the creation of realistic videos from scratch, and MusicFX with AI Sandbox, which allows the creation of music.

Other important announcements that should impact the way we use Google and Gemini generative AI were also released. Now, AI will officially be in Google Search; previously, it was necessary to enable the tool to gain access. Check out the main new features of creation and search tools.

VideoFX with Veo – video creation

Google announced VideoFX, a tool that creates realistic videos from text commands, powered by DeepMind’s Veo. The tool competes directly with Sora, from OpenAI. The maker of ChatGPT announced the artificial intelligence model, which can create realistic videos of up to 60 seconds with text commands, this year. For now, US users can join a waiting list to use VideoFX.

ImageFX with Imagen 3: realistic image generator

About two months ago, Google stopped generating Gemini images after distortions and inappropriate images were generated. In a statement, the company apologized and guaranteed to do better in the future.

Now, the company has announced Imagen 3, available in Google Labs’ ImageFX. It is possible to generate completely new images using a text command, something that is now more common and available on platforms like MidJourney, for example.

MusicFX with AI Sandbox: Songs

Along with YouTube, Google announced a generative AI tool for creating music. The tool competes with companies like Moises.ai, which uses AI to separate tracks of the same song and was created by Brazilian Geraldo Ramos, who lives in the United States. The company also sells “voice packs” that allow users to apply an artist’s timbre to different performances while maintaining the original user’s natural accent and expression.

New search from Google with AI

The company presented its new way of connecting with the world, with the official arrival of generative AI, Gemini, in Search. This is a new look for the entire service, with the presentation of even more personalized information and improved by the intersection of the user’s consumption and their interests. It’s a new feature that will change the way you search on the internet forever, with automatically generated summaries.

The news comes one day after the OpenAI introduce GPT-4o, a model that is faster than previous models and programmed to appear “chatty”, and sometimes even seductive in its responses to prompts. The new version reads and discusses images, translates languages and identifies emotions from visual expressions. Google showed something similar, but still in the experimental phase, called project Astra.

Now, instead of multiple questions, you can ask the most complex questions, with all the ideas you have in mind, all at once. This is all possible thanks to a new custom Gemini model. It brings together Gemini’s advanced capabilities, including multi-step reasoning, planning, and multimodality, with search systems.

For now, the new way of searching is only available in the United States and will soon be available in other countries. Check out the resources that will be available.

Planning

It will be possible to use Gemini in Search to plan something. Meal and travel planning, for example, allows AI to create an itinerary based on your interests, flight time and other details. Later this year, Google will add personalization features and more categories like parties, date nights, and workouts.

Chat with any app

Gemini is now available on the vast majority of Google products. Android users can now enable the generative AI assistant on their smartphone.

In Google Photos, it is now possible to write what you want through the chatbot. Ask, for example, what your car’s license plate is or when was the last time you took photos in the pool, and receive an answer immediately;

Gmail users can request a summary of the main emails from a specific account. It is also possible to summarize meetings that take place via Google Meet;

Made for students, Google Notebook will work with an AI-assisted tutor (one of the new features announced yesterday by OpenAI). Users can send questions via audio;

Through Google Chat, it will be possible to create a generative AI assistant that automatically answers the questions asked. It is available, for example, in group conversations, allowing any employee on a team to see the response in real time;

Soon, the Google Maps platform will offer summaries of establishments, generated with the help of AI from comments left by users. It will also be possible to find summaries of a region or neighborhood and ask for specific recommendations;