# 20241031-adams-planning-doc ## staff available * mk * mw (set up + first 15min, 12:30-1:20) * cc * bp * elisa (until 12:30) to check: * joel * dan HGSE ## stations * entry * take photo * (optional) * character * either "what they are dressing up as for Halloween" OR "what is a creature/character that scared you as a child?" * ask if they have ChatGPT Edu and how to get it * tarot cards * first part of the station can be "getting a tarot card reading" * chat completions from an OpenAI assistant that specifically return a certain number of cards in a specific spread, text-to-voice by elevenlabs? * then generating, from replicate, a "major arcana" card representing the person * this could be part of the workflow where we take a photo of them, put it through CV description, then feed that into a prompt for the card (and maybe this prompt and/or the "title" of their card is based around their initial reading?) * see if we can print with cardstock at Adams (make them a tarot card size) * "scary stories to tell in the dark" * starting with the "character" information from entry * transform this into a longer tale * tale is used to make a voice form elevenlabs * recording is fed through elevenlabs or runwayML to make a video * "costume design" * starting with image taken at entry * adding costumes to a person with multiple tools * Canva * Adobe * Stable Diffusion w/ controlnet * replicate? * on exit: * tarot card * costume of self * QR code on a card to an airtable gallery of scary stories/tarot readings