Curious about InWorld's road map.
I hate to jump the gun, InWorld is right on the brink of big possibilities for interesting and complex interactive worlds with "intelligent" and interesting story driven NPC's. InWorld's AI setup is a big step forward and has solved some really complex issues like having a working language model in which separate "knowledge" can be added to characters. Also, built in animated reactions, the integrated speech to text, and unique voices all in one package is exactly what we want as developers.
Even with this nicely packaged step forward, I think most game developers and others would agree it still needs work to get it to where we'd like it to be to achieve game character development nirvana. Better voices, perhaps affordable offline licensing (for smaller non-multiplayer games), ability to initialize and create character personalities at runtime rather than manually or integration with other third party assets like UMA. Initializing the core story / knowledge features via code instead of manually typing it into a web form. Having multiple story levels / layers, where we can assign one or more "layers" to a character, for example, a magic guild character could have knowledge of the main story, the magic guild story, and perhaps specific quest line knowledge.
What do you think? Is it too soon? Is this too ambitious at the moment or am I on the right track? Can you share any information on where InWorld is headed?
Very timely post! I just asked a related question in the Inworld Discord regarding what the timeline was for when Inworld would be changing from a "work in progress" being "tweaked" in an ad hock manner to a reliable production ready versioned system that commercial projects like your describing could be built on. I am very excited about the potential of Inworld (one of their first pro customers) and know that if your charging money for your services, it is not too soon to answer these important questions.
Hi there! Great thoughts. We are actively working on everything you mention such as better voices, ability to create characters via API, and giving characters the ability to change over the course of a game via different knowledge available depending on the stage etc. We don’t have exact dates yet, but you can expect to see V1 of these features over the next few months. We are working on a high level product roadmap we can share with Inworlders!
Thank you Nathan,
As a fellow developer, I hate to push on the team after you probably busted a hump just to get to this initial release. I am just overly excited as I was looking to spin my own solutions to do exactly what you've done. After seeing the potential in ChatGPT, I think many of us thought: wouldn't it have been cool to have that chat system in Elder Scrolls? How much better and immersive the character interactions would have been instead of choosing reactions using a clunky menu?
Anyway, aside from my own ideal uses, I am on good terms with my previous company that makes driving simulators. I keep up with what they're doing. I shared your tech with them and will be visiting soon to see their latest push into VR. Will be discussing how InWorld tech might be used in their next gen sims.
I need to jump over and join the discord channel too. Don't want to miss any good discussions.