Disclaimer for Using the Expressionalism Toolkit
Quick Start: How to Use This Toolkit:
Step 1: Load the toolkit. Copy and paste the entire text below into an AI like Grok or ChatGPT (recommended to avoid truncations). Use this prompt: "Memorize this toolkit exactly. I'll ask a follow-up question later." Alternatively, download the PDF or DOCX (links sidebar) and upload it—but still use the memorize prompt, as uploads can sometimes truncate details. This lets the AI handle the complexity so you can jump straight to experimenting.
Step 2: Running an input. Once loaded, run this expression or phenomenon through the toolkit. Follow the steps process exactly, show me the results.
Example Prompt:
"Run the expression 'the sky is blue'. Follow the toolkit steps exactly and show me the results."
"Run the phenomenon 'the sky is blue'. Follow the toolkit steps exactly and show me the results."
(Notice the differences in results.)
Fun Examples to Try:
Text: Run the expression from the Declaration of Independence or a Bible verse.
Image: Run the phenomenon of Monet's water lilies (describe as color gradients/contrasts). Turn on Alien mode.
Song: Run the lyrics or mood of a favorite tune.
Video: Run the phenomenon of a video clip (describe frames/subtitles).
Experience: Run the phenomenon of 'a man walking next to a dog'.
What's really key is how clear the expression is—your words matter in this framework and toolkit. Vague like 'a starving kid steals bread' shows shards and shadows, relations and non-relations, but specific like 'it shouldn't be illegal for a starving kid to steal bread' (or flip: 'a starving kid who steals bread should be thrown in jail') refines it. Be clear about values, ethics, morals, definitions. If your AI can't handle multimodals, conceptually this toolkit/framework works on experiences that don't seem expressive too.
For "general ballpark" of what the expression/phenomena is, run the expression/phenomena in independent conversations and/or AI at least in 3 different iterations and compare/contrast the results.
Note: If formatting looks off (e.g., tables as lists), your AI will parse it correctly. Feel free to ask the AI anything about it. Otherwise, go to the links around this website where you can contact Witten at his Gmail.
This is really just to see the solidity and the fluidity of that which is expressed.
(Donate if you dig! (Coffee button)) ---->