I'm curious about how the back-end of this works. Are you just making API calls to OpenAI's servers? Do you have your own server running? This isn't running locally as is doesn't work without an internet connection.
Our game is built upon the ChatGPT framework (turbo3.5) by OpenAI. Now the internet connection is a must for our game. But there's no need for players to provide a ChatGPT account; the game utilizes the underlying technology to create a unique and immersive gameplay experience.
Got it. Thank you. It would be cool if there was a possibility to use models running locally to allow for trying some of the newer LLMs that are both powerful and less restrictive. Additionally, it would lower your operating costs, if the development was not too time consuming.
