When Google was launched, I see 3, millions of users around the world were amazed by what they have found. The quality of the videos generated by this model is brutal, but What happens if we tell you that it was just the beginning?
Just a couple of months later, Google presented Genie 3 and to which generates realistic and interactive virtual worlds It will make you fly in your head. In addition, it does so by increasing the duration of the simulations, being able to change them and are even able to remember them. Continue reading so don’t miss any detail.
Genie 3 creates realistic simulations in which you can interact with the environment and change everything in real time
Genie 3 comes to revolutionize everything we knew about the creation of virtual environments, left in diapers to what has been seen in Genie 2. The model developed by Deepmind (Google) improves everything, being able to do it Create complex simulations of any environment with incredible graphic quality.
The new model It works with a resolution of 720p and a tariff of 24 images per second. It looks very little, but it is double the resolution of Genie 2. Apart from, Genie 3 is capable Generates several minutes of simulations And not a few seconds, like their predecessor.
And if this was not enough justification, Genie 3 works in real time and beyond Allows you to navigate on stage, also interacts with the environment. Yes, everything you do within the simulation will have a consequenceWe are not talking about static recreation.
Also the new model Solve two of the major problems of Genie 2. The first are the limits of the simulation, because in Genie 2 you reached a point and it was like crossing with an invisible wall that did not allow you to go further. The seconds are Physical inconsistencieswho appear when you try to go back and note that there are obvious changes in the simulation that were not before before.
In Genie 3 this does not happen, because You can go back to a minute ago and the entire environment will remain consistent With what you have seen and done. Yes, you may find some minor imperfections, Google does not deny it, but the consistency will be there. Moreover, While continuing to navigate in the simulation, Genie 3 will create an environment that you can travel and you will not come across an invisible wall.
How is it possible? Because Genie 3 works with Author and generation in real timewhich allows you to create worlds. have a little memory of what is happening.
Is there more? Yes, because Genie 3 too It allows you to add elements and events to the royal simulationthrough the first text. Do you want a corridor to appear in your morning career? Make it to Genie 3 and add it. Do you want, suddenly, a meteorite falls from the sky and caused chaos in the city while you were in silent traffic? Go ahead, Genie 3 will do it, even if initially it was not programmed in the simulation.
What can Genie 3 be used and where is it available?
Unlike Genie 2, designed for 3D entertainment environments, Genie 3 goes far further. According to Google, it can be used not only for video games, but also for Simulate the functioning of the body and movement In different scenarios, to create training simulations for emergency and safety personnel, as well as for Trains robots and agents to virtual environments Before bringing them to real life scenarios.
With Genie 3 the limit practically puts itBecause you can even travel to other eras in our history with him. And let’s say “practically”, because there are some limits, although they seem little for everything it can do.
Some examples? It is unable to recreate places of the real world with absolute geographical precision, does not allow multiple agents to interact with each other In the same virtual environment, it continues to have problems with the first generation of texts and the interaction lasts only a few minutes.
Now, surely they made you want to try it immediately, but we have bad news for you: Genie 3 will not be available for the public. At the moment, Google will only offer companies and governments. However, they plan to offer it soon to researchers and developers of the AI.
Fuente | Google
