At the beginning of this year, Microsoft's gaming division, instead of focusing on solving its internal issues in game development and publishing, decided to join the race for leadership in generative artificial intelligence. The company introduced an AI model called Muse, which was intended to simplify the process of creating new projects and help preserve older games. However, despite bold claims, experts who examined the technology concluded that it is already outdated and lacks real practical value in its stated applications.

From the moment of its announcement, Muse failed to generate enthusiasm among representatives of the gaming industry. An additional point of criticism was that the model was trained on the failed multiplayer action game Bleeding Edge by Ninja Theory, which quickly lost popularity and support. After a detailed analysis, the concerns of specialists were confirmed.

Although Microsoft claimed that Muse was "the first of its kind," similar technologies have existed for a long time. For example, in 2024, the Israeli company Decart introduced Oasis—a model capable of generating game worlds in real time. That same year, Google DeepMind announced Genie 2, which can create virtual spaces based on text or images, simulating object behavior and lighting. Nvidia's GameGAN even managed to recreate the game Pac-Man without using an engine or source code. Against the backdrop of these developments, Muse appears not as a breakthrough but as a belated attempt to catch up with competitors.

Journalists from Wired discussed the announcement of Muse with developers from major studios, and all of them expressed extreme skepticism. Overall, it boils down to the fact that instead of trying to retain valuable employees who are leaving, the company stubbornly continues to develop its AI direction. Moreover, it seems to be done more for investors than for ordinary developers. If the prototyping process is simplified and automated, young developers may simply miss out on acquiring many key skills. Additionally, Muse has proven incapable of recreating lost mechanics or adapting classic projects for new platforms.

Main image: wall.alphacoders.com

Comment on this