US Large Model Data Collection

US Large Model Usage Data Research Project Background This research is a preliminary preparation for building a real multimodal dialogue dataset for large model training. We require a large volume of authentic ChatGPT and Gemini historical dialogue data with images and videos. We are now launching a part-time research task to sort out eligible user information and corresponding data quotation standards. Research Objectives To count the number and gender distribution of qualified US-based users, as well as the expected unit price of a single valid multimodal dialogue data. The valid data refers to dialogue records (questions with images/videos plus AI responses) generated within the past year. Target User Groups Two types of US-resident users who regularly use ChatGPT/Gemini are included: 1. Native US users; 2. Chinese international students residing in the US for more than 1 year with a TOEFL score of 100+ and proficient English skills. Core Research Content 1. Count the total number of qualified users and their gender ratio; 2. Confirm the fixed unit price for data collection: $2 per valid data piece; 3. Verify the practical feasibility of large-scale data collection by target users. Work Requirements Ensure all researched information is true, accurate and valid. Statistic user information and data quotations truthfully and submit complete research results on time. This is an early-stage research project, and the subsequent large-scale data collection will be carried out based on the final research results. Contact Feel free to contact us if you have any questions. Thank you for your cooperation and support!