While there is still not much known about the next From Software title, Elden Ring, fans are grasping at just about any information they can get their hands on these days. This includes the kind of enemies and bosses that will be included in the game once it is released, with one talented fan taking it into their own hands and making their own predictions of what the final boss in Elden Ring could be.
News of Elden Ring is few and far between, since almost nothing about the game has been released since E3 in 2019. While a bunch of footage was leaked recently, much of the game is still unknown. Players are eagerly anticipating the game based on the developer’s previous titles such as Sekiro: Shadows Die Twice, Dark Souls and Bloodborne, as well as the addition of famed novelist George R. R. Martin, known for writing the Game of Thrones series.
Click the button below to start this article in quick view.
The online Elden Ring community has been sharing their theories on what the final boss in the game will be and creating their own models. The latest one, by Reddit user Geldoph, is a bit different from the rest. This boss was made using neural networks, instead of the imagination of one person. Geldoph used the small amount of information currently available to produce the image, sourcing the material into the program to generate a possible boss based on the kind of designs that have been seen so far. The image that was posted on the Elden Ring subreddit was generated by a neural network, creating the horrifying image below.
Neural networks are computing systems that follow the same concept as the human brain, connected by a biological network of neural systems. This kind of system is often used by computers when developers need to simulate learning in order to complete tasks. One such task involves generating images based on text inputs, or the reverse, where a program identifies an image with a text source. Here, the program had only a small pool of resources to use in order to generate a possible boss monster.
While it is certainly possible that some aspects of the generated image may be correct, there is no way of knowing until Elden Ring is finally released or the boss is previewed by the developers. This kind of creation should be approached more often, as the possibilities of creativity posed by neural networks have yet to be explored on a large scale. Perhaps, sometime in the future, an entire game will be generated by a neural network.
About The Author