Google I/O is on us next week, and as usual, there is sure to have an avalanche of new products, features and upgrades between Hupla on the stage. Google Mithun and many of its AI offshoots will definitely be a part of them. Rumors of preview and more attractive, but there are some that I am especially eager to hear. I have placed a set of highlights coming from Google together as it shares the next stage of its plans for Gemini.
AI agent of Project Meriner
Project Meriner is the answer to Google which is for increasing proliferation of AI agents such as manus and browser’s use. Instead of only surfasting the link, the merrinner is designed to interact with the web, the way a human does, to fill the form using an invisible mouse and keyboard, find things within the websites, and click on the button required to complete the tasks.
Meriners can fill your tax form, book a trip, and send it to any complaint that you are in the company’s customer service. Although not strictly Gemini is not strictly, the construction of Google Deepmind is a lot of Gemini’s story that helps to automatically automatically make digital activities for people. In fact, merrinner is believed to have advanced and integrated with Google Chrome. This will be particularly impressive for those who manage repetitive administrators, navigate government or insurance websites, or simply want a more efficient way to handle online tasks.
Gemini’s personal memory
Constant memory is a stable, but usually generative AI assistants feel an incomplete feeling. Google is expected to unveil the upgrade in Gemini’s memory, which means that now there is no need to remind your priorities AI. Gemini may remember that you dislike morning meetings, prefer metric units, or always book isal seats on flights.
Like the memory system of Chatgpt, Gemini is expected to remember things from interaction with both of you and also set a custom instruction, where you can manually add things you want to remember. Of course, Google is likely to assurance to users that the frequent memory feature is opt-in and involves users to see, edit and remove the controls that recall Gemini.
Imagene 4 and VO3
Imagene and VO are Google’s generic AI image and video creation tools respectively. Google is expected to start the latest versions of both on I/O. Imagene 4 is considered much better in photorolic images and matched actual signals. Whatever style you request, you should also be better to be consistent. The Veo 3 is also going for a more consistent style from clip to clip. They will be integrated with Gemini for easy access to material creators, students and anyone really want a quick picture or video.
Share gemstone
Gemini gems, customized and concentrated Gemini models that any user can make is useful for all types of activities. You can create an art critic for your own motivational coach, food-plan nutritionist, or your latest pictures. What you can’t do right now is sharing them with other people.
The gems are basically like Custom GPTS available from Chatgpt, except GPTS are shared and discovered in the GPT store. Google is expected to coincide and start allowing users to share their gems with others. You can see everything from a class-specific tuition gemstone, a group of gems designed to recommend coding for various outlets, or recommended films. And a gem market is not just a benefit for users. Google would prefer to build a community around Gemini as it is with apps on the Play Store. Shared gemstone can be the best entrance for that type of community-managed ecosystem.
