Google has bet on the launch of
a new AI model that promises to exceed human expectations by being able to perform and understand multitasking processes.
Most technology companies have launched their own offerings, wishing to be leaders in most European households. They want to generate interesting products that are used on an ongoing basis. That is why they try to attract our attention by launching novelties and updates to present themselves as the most successful.
What is Gemini?
It is an AI project, as mentioned above, which can be defined as multimodal. Among the tasks it can perform are word processing, creation of written content, generation of computer code, creation of images as well as audio tracks and even videos.
Eli Collins, the vice president of Google DeepMind has already confirmed that there will be an “Ultra” version that will be available from 2024. Some of the features of this much more refined model will be that it will outperform humans in “massive multitasking language comprehension”. It will generate comprehensive knowledge based on learning in 57 science, engineering, mathematics, social sciences and humanities and technology subjects.
According to Collins,
is the most capable AI model at the moment. He added that: “it is inspired by the way people understand the world,” relying on the way people reason things out, interacting with the world around us based on the conclusions we draw. Collins defined
as “a useful collaborator.
Here’s what Gemini has done so far
Gemini has been able to identify and distinguish geometric shapes. He has also made a study on formulations to find their areas. At the presentation, he delighted the audience with an organized talk in which he also found mathematical errors in the formulas he had been given, providing much more appropriate solutions.
He was surprised to identify drawings through representations of scattered dots. It also generated debate based on images, audios and proposed texts. The good thing about Gemini is that it also charmed people by being able to provide suggestions, demonstrating a lot of autonomy and effectiveness in resolving any questions, doubts or conflicts.
This new AI will have several versions
Gemini will operate from Google Bard servers, one of ChatGPT‘s main competitors. From there it will send information to browsers or Android devices as normal. Its developers have created 3 versions of Gemini that will be available in a staggered manner.
For example, its Nano version can now be used from any smart device with Android. The Pro can only be used from December 13 onwards. As we mentioned before, the last one to be released will be the Ultra version, which will arrive at the beginning of the new year. The exact date has not yet been published and is still a secret, so we will have to wait to see it.
will also be available internationally to more than 180 countries, primarily in English. As the developers update other languages, these will be integrated with English so that all users around the world can enjoy the benefits of interacting with Gemini.
. Collins explained that
has been created with a variety of programming sources. Something that helps it to understand all kinds of problems and debates, with greater ease than other existing AI models.