Also I have a feeling that the NPC narratives when enriched by all player interactions would tend to go towards some average, i.e. become more and more generic and dull. The truly rich experiences have always been on the fringes in all art for me. I'm interested to see a counter example to this though.
Obviously we don't want experience averaging NPCs.
> Mad Hatter’s first goal is to ask the player to humor him with a joke.
This is ok for simple stylized, small world, single storyline games, but not for open worlds.
For that, NPCs need their own motivations, so they essentially play the game too. With needs like maintain their smithy, so they can build quality armor, so they can make customers happy, so they can make money, so they can feed their family. I.e. not just being props.
And flexible in how they adjust subgoals to meet their core goals, relative to player interaction: such as being convinced to go on a trek as an armorer, or on a search to find and extract rare materials for a magical shield. Willing to fight in revenge for their home town's sacking.
Westworld got NPC dynamics right.
As c048 said best (https://news.ycombinator.com/item?id=40316308), good art needs direction and focus, and AI NPCs being able to do anything they want in an open world doesn't make for a good story. It could make for a good sandbox with emergent gameplay, sure. But my point was that there are people who yearn for a good story, and a good story doesn't have many options in it, otherwise the story's beats would become meaningless.
Hmm. Maybe with this the problem Rockstar has with their linear missions mixed with open world could be fixed a bit - i.e. the open world NPCs could react in story missions to what you've done in free roam. This would lessen the cognitive dissonance the NPCs seem to have when you can now blow a town into pieces and then start the story mission and no one cares.
Subgoals that are resource, constraint defined, could make very flexible solutions. For instance, if you need help with some subgoal, it is going to play out very different depending on what NPCs you have established credibility with before, and their skills and dynamics.
If you need money, then how you get that money is also going to depend on your history, knowledge of a city, previous connections, etc.
That could provide an overall story arc, but with a very open world and fully functional NPC experience.
So: organized series and trees of goals. Open ended solutions.
So you want a system like rimworld, and a LLM doesn't really help on that front. It can power the dialogues, but that's it, all the logic you describe would need to be encoded somewhere else.
a) LLMs are excellent at keeping a "linear enough" storyline without being linear. They'll let you do outlandish things, but given the assignment of "tell a cohesive story" they manage to corral the story back to something sensible unless the player intentionally keeps pushing at the boundary (in which case they probably do want things to go off the rails)
b) LLMs can do delightfully colorful dialogue, they just need to be grounded in a character. Everyone thinks of factual grounding, but given enough description of speech patterns, character motivations, etc. they're capable of dialogue that's lively and completely rid of "GPT-isms", which are what tend to break immersion
I actually trained an open model [2] on the task of grounding LLMs in characters and actions as opposed to factual things like RAG, and eventually I want to build a game demo out of it
[1] https://www.tryspellbound.com [2] https://huggingface.co/hf-100/mistral-spellbound-research
Didn't go through the registering phase cause I couldn't find any info on what info you store and on the pricing. Could you provide that?
The pricing page is also visible when signed in and in a story (no dedicated pricing page yet but it's in the backlog)
Since existing worlds are (currently) where most of the stories are set, it's worth it to use a closed source models and wrangle their issues with dialogue.
To it's credit though, Llama 3 is the first OSS model trained on enough tokens to not feel lost for most worlds, so I've started routing some traffic to it for free users
The output format the site uses is also really really hard for most models to follow without fine-tuning, but fine-tuning then causes them to pick up the vocabulary of whichever model they were fine tuned on, which is a bit unfortunate
In Skyrim it took me multiple years to finally complete the main story line. I would always look for books, store potion ingredients, walk in the forest...
There's a big role play component in this. The game could learn from the players RP and improve.
So your subscription pays for generations, but also fills out the persistent narrative and lore.
I actually like the 3 choices dialogue system and don't want to come up with my own NPC small talk on my couch, but the idea that any NPC that's been talked to by any player gradually has a whole catalogue of interactions to pull from means there would be compounding depth to the world as more players interact.
Like Minecraft but where it's collective world building by exploration and interaction rather than actual building it yourself.
Humans tend to be pretty predictable, so I feel like the universe would quickly feel like it's insanely deep after just a short while with a somewhat decent player base.