Discussion about this post

User's avatar
Neural Foundry's avatar

The prompt theft thing is peak irony but also kind of predictable. Once people invest time engineering prompts that actually work, they start treating them like IP even if the underlying model scraped millions of artists without asking. I spent hours trying to get consistent results for a project last year and honestly did feel annoyed when someone copy-pasted my exact phrasing and got better outputs. But then I remembered the whole thing runs on unconsented training data so like... maybe we're all just operating in a moral grey area here. The real question is wether this creates a new category of labor or if it's just gatekeeping with extra steps.

No posts

Ready for more?