Reproduce “Cost Transparency of Enterprise AI Adoption” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Soogand Alavi, Salar Nozari, Andrea Luangrath
Researchers from the Tippie College of Business at the University of Iowa found that incorporating politeness into user prompts for Large Language Models significantly reduces the number of generated output tokens, thereby lowering operational costs for enterprises without compromising response quality. This effect translates into potential millions in cost savings for businesses and challenges current token-based API pricing models.
Key insights
- Politeness in user prompts is not merely a social nicety but a quantifiable factor that causally reduces LLM output token generation, creating an unexpected economic lever for enterprises.
- The reduction in output tokens due to politeness does not result in a loss of semantic meaning or perceived quality in the LLM's response.
- Current token-based LLM API pricing models create a lack of cost transparency and predictability, as end-user linguistic styles can significantly influence enterprise expenses beyond direct control.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.