Generative AI needs improvement
I UNDERSTAND AI IS DEVELOPING, AND WANT TO POINT OUT A PROBLEM. I AM A SCREENWRITER. I USED THE NEW ADOBE GENERATIVE AI IN ACROBAT TO ANALYZE A PORTION OF A SCREENPLAY I WROTE. WHEN I ASKED A QUESTION ABOUT A DETECTIVE ANALYZING PICTURES OF A CRIME SCENE, THE RESPONSE WAS "SOME CONTENT IN THE DOCUMENT, QUESTON, AND/OR ANSWER VIOLATES THE USER GUIDELINES"
THE GUIDELINES AT
https://www.adobe.com/legal/licenses-terms/adobe-gen-ai-user-guidelines.html THREATEN THAT "We may take action on your Adobe account if we discover content or behavior that violates these Guidelines."
FIRST: THE AI SHOULD RECOGNIZE CREATIVE WRITING, SUCH AS A SCREENPLAY. SECOND: A WRITER SHOULD NOT BE THREATENED WITH "ACTION"
THIRD: THE AI THEN STOPPED ALLOWING ME TO ASK QUESTIONS ABOUT MY SCREENPLAY.
I'll point out more people are writing screenplays, novels, and news articles than are plotting race wars and mass murder, so this consequence is affecting the wrong people.
Michael R. Barnard
shared this idea
-
Jen Le
commented
Generative AI in Acrobat is still in beta, and we are working on developing responsible AI that conforms with Adobe’s AI Ethics principles while helping users unlock value and insights from their documents.