Enabling Technologies

Large Language models or AI:

ChatGPT4 or similar. Maybe use the LLama model (open source) and even VLLM (https://vllm.AI).

Voice to text:

We will generate AI voices and also license users voices.

Zero Knowledge Machine Learning (ZKML) :

Game developers share their in-game data with us without explicitly sharing user data. In the context of gaming, ZKML can be applied to train models that make personalised recommendations or predictions for players, while ensuring the privacy of their gaming data. It enables game developers to analyse and learn from user data without accessing or exposing the specific details of individual players. This approach enhances privacy protection while still allowing for the benefits of machine learning and personalised experiences in gaming.


Layer 2 scaling solutions like ImmutableX will be used to store assets created by users, including robots, upgrade data, and properties. The on-chain storage will manage the metadata and form experiences of robots.

Users will be able to access their robots through the Plasm Dashboard, which will feature a wallet to store their robots in the form of NFTs. While users will be creating assets on our software products or acquiring robots through the Marketplace, all assets are stored as NFTs to ensure decentralised ownership and facilitate management. All members have the ability to move their assets using the token standard.


Our platform will have the option for NPC’s to be issued on the blockchain. This is interesting for Web3 companies that want to issue NFT’s in their game.


The use of AI in voice synthesis and its application in gaming is an exciting frontier. This concept, wherein users record their voices, which AI then learns to synthesise, has the potential to personalise the gaming experience and also provides an avenue for users to monetise their contributions.

Users could record various speech samples, which would then be used to train a machine learning model to generate a synthetic voice that sounds just like them. This synthesized voice could then be used to voice non-player characters (NPCs) or even the player's own character in a video game, adding a personal touch to the gaming experience. Users would also have the option to share or sell the rights to their synthesised voices to game developers, thereby earning money for their contribution.

This concept could significantly enrich the gaming industry. It would open up a new form of user-generated content, with players contributing not just their voices but also their unique expressions and dialects. It could make games more immersive and personalised, as characters speak in familiar or player-specific voices.

Moreover, it could democratize the process of voice acting in games. Anyone with a unique voice or an engaging way of speaking could contribute to a game's soundscape, regardless of their location or professional background.


Plasm AI is envisioning a groundbreaking application of artificial intelligence for the gaming industry: real-time voice localisation. This ambitious goal aims to use AI to instantly adapt in-game voice interactions to different languages, significantly enhancing the gaming experience for players worldwide.

At the heart of this endeavour lies a sophisticated AI model, trained on extensive ultilingual datasets. The plan is for the AI to translate spoken language accurately into a target language in real-time and generate the corresponding voice output. The resulting localized voice is intended to maintain the original tone and emotional intent, offering an immersive, linguistically tailored gaming experience.

This transformative technology could profoundly impact non-playable character (NPC) interactions within games. Rather than relying on text translation, players could potentially hear NPCs speaking their language fluently, increasing the engagement and accessibility for a global audience.

Moreover, the potential of Plasm AI's technology extends beyond pre-recorded game dialogues. It could also apply to real-time player interactions in multiplayer games, offering on-the-fly translation of voice communications. This would facilitate smoother cooperation and communication between players from different linguistic backgrounds, thereby enhancing the multiplayer experience.

It's important to note that while Plasm AI's vision for real-time voice localisation is incredibly promising, it remains a future goal. The technical complexities involved in such an endeavour are significant, and achieving this level of instantaneous, accurate voice translation is a challenging feat. However, if realised, it would set a new standard for immersion and inclusivity in the gaming world, reinforcing the universal appeal of gaming by ensuring no player is left out because of language barriers.

Last updated