rct AI has developed a new algorithm ACAR, and has reached cooperation with many game guilds around the world
Recently, rct AI announced the development of an AI algorithm—ACAR (Action Clustering using Action Representation) aimed at improving the efficiency of large-scale action space in Web3 games. This paper was published by IEEE CoG 2022, the top international conference in the field of artificial intelligence games, and rct AI was invited to give a speech.
This study chose Axie Infinity, the most representative card battle chain game, as an example. Through the application of ACAR algorithm, it overcomes the "huge and complex action space" of Axie Infinity (the number of action space combinations of all cards is about 10^ 23), "gaming in the game", "many card groups" and other issues, in a large number of simulated battle data, the game winning rate and overall efficiency have been greatly improved, so that AI can truly be more efficient and more realistic in Web3 games Simulate human players, and at the same time, it can further provide rich diversity for the game ecology based on the characteristics of the algorithm.

The launch of ACAR has attracted attention and aims at the new blue ocean of AI application in the field of Web3 games. rct AI has joined hands with a number of overseas game guilds to carry out a series of in-depth cooperation around a variety of different types of chain games, allowing AI virtual humans to participate in various gameplays with human players on an equal footing, participate in the economic cycle and bring To achieve real sustainability and stability, completely liberate the liquidity of digital assets in the game, and bring players a more personalized intelligent interactive experience, while bringing more efficient and stable automation benefits.
All the papers recognized by the top academic conferences in the global game and artificial intelligence circles were completed by the algorithm engineers of rct AI's decentralized virtual human platform Deterrence team. Up to now, rct AI has successfully established a mature Web3 business ecosystem, including the self-reproducing virtual human NFT "Nüwa Nuwa" released by rct AI and artist Song Ting; Delysium, the world's first playable AAA-level MMORPG Web3 game; The Web3 game matrix Mirror World driven by interoperability and interoperability SDK; the Soularis brand jointly launched with China's virtual human ecology leader Next World Culture, to create a virtual society where users and AI virtual humans coexist, and digital identities and assets communicate.
rct AI was incubated by Y Combinator W19, and its investors include Galaxy Interactive, Makers Fund, Xinghan Capital, Y2Z Ventures, Springwind Ventures, Zonff Partners, etc. rct AI adheres to "Future Web3 with AI", and constantly explores broader market space and innovative application methods.


