←Index



11/22 Interface ## Week 11: In this week, we learned how to ... Interface and Application Programming # Show and Tell - using h-bridge to make audio # Class notes --- ## Idea # Assignment # 1) Rhythmic objects This week's assignment overlaped nicely with the class "Tangible Interfaces". With my group, we want to mitigate awkward dinner situations with music and invite people to play. We attached accelorometer and capacitive sensors to chopsticks, bowls and a bottle and sonified the data. Would you synchronise your movements and rhythm with the people on a dinner table when you hear a sonic feedback or see a visual?  Specifically, I worked on the accelorometer.  # 2) Tangible Prompting Furthermore, we explored how we can use everyday objects to communicate with GenAI Researchers found that being polite to Chat GPT can give you more accurate answers. What if we could give Chat GPT a physical high five? Can we have tangible prompting and do so using existing everyday objects?  I wrote a server client system for GenAI to read serial data input. It is made out of 3 parts. 1) An accelerometer is attached to a bottle. The movement of the bottle is read in through python. The data is mapped to predefined prompts. For fast changing movements are mapped to 'positive feedback prompts' e.g. "It's evident that you're continuously learning and improving. Well done!". Slow changing movements are mapped to more accurate feedback e.g. "Could you consider alternative perspectives or viewpoints in your response?" 2) The physical prompts are then send to OpenAI's Assistant API. 3) The answer is then displayed on a locally hosted website.  ## Next Steps - --- # Resources - files - [accelerometer ino](files/rotation_vector_it_works_for_L.ino) - [Tangibe Prompt API](files/TI_api.py) - [Tangibe Prompt Assistant](files/TI_assistant.py) - [preprompt](files/pre_prompts.py) --- ↳About