yea that’s just playing on semantics. Of course machine learning will keep assisting many different workflows, thank you Tim.
Go ahead with your ML-assisted procedural animations, your ML-enhanced denoising, your ML-powered stochastic mesh pruning.
What people don’t want is generated visuals/music that try and pass off as art. I’d love to summon debility to explain that the ruling class doesn’t get it, but they do -it’s a convenient way for them to save on human labor that is also… just too tempting to use for replacing art as well.
AI generated visuals and music are one thing, but one use of AI in that category is giving NPCs actually interesting things to say.
Currently, without AI, they have one or two lines about the weather, or the general state of affairs, but if you can pass an LLM the current state of the world and the player’s recent actions, then run the output through a tts model tuned to a distinct voice to, you’d be making the world feel way deeper than it is now with much more insightful interactions. You could maybe go even further and add infinite mini-quests in a similar manner, but it’s better to start off small.
AI doesn’t have to replace, and the best studios will use it to enhance and add to their art, the average ones will avoid it, and the worst ones will use it to replace art.
yea that’s just playing on semantics. Of course machine learning will keep assisting many different workflows, thank you Tim.
Go ahead with your ML-assisted procedural animations, your ML-enhanced denoising, your ML-powered stochastic mesh pruning.
What people don’t want is generated visuals/music that try and pass off as art. I’d love to summon debility to explain that the ruling class doesn’t get it, but they do -it’s a convenient way for them to save on human labor that is also… just too tempting to use for replacing art as well.
AI generated visuals and music are one thing, but one use of AI in that category is giving NPCs actually interesting things to say.
Currently, without AI, they have one or two lines about the weather, or the general state of affairs, but if you can pass an LLM the current state of the world and the player’s recent actions, then run the output through a tts model tuned to a distinct voice to, you’d be making the world feel way deeper than it is now with much more insightful interactions. You could maybe go even further and add infinite mini-quests in a similar manner, but it’s better to start off small.
AI doesn’t have to replace, and the best studios will use it to enhance and add to their art, the average ones will avoid it, and the worst ones will use it to replace art.