causal
Foundational models democratized deepfake creation
The emergence of general-purpose, text-prompted foundational models (Stable Diffusion, Midjourney) trained on internet-scale data has shifted deepfake creation from a niche skill requiring VFX expertise and post-production to a democratized capability where anyone can type a prompt and generate images of anybody or anything, as seen in the Midjourney v5 Trump arraignment images and the Pope in a Balenciaga jacket.
causalpending
Speaker
Nina SchickEvidence Quote
“they have nlp tied into them so that when we use them how do we get them to create something but we prompt them we just type what they want what we want them to create”
Created: 6/14/2026, 2:18:00 AM
Your notes
Loading notes...