US EditionPublished • loading... • Updated
Doodles Trained an AI Model on Nothing But Its Own Art. Now the Company Wants to Make a Feature Film with It
Doodles AI, trained exclusively on its own art, creates studio-grade images 84,000 times faster and aims to produce the first AI-native feature film using proprietary intellectual property.
- On Tuesday, March 17, 2026, entertainment company Doodles launched Doodles AI, a proprietary creative tool powered by Prism 1.0 that generates studio-grade imagery without using third-party training data.
- Doodles CEO Scott Martin, aka Burnt Toast, created the tool to address industry concerns regarding copyright and consent, utilizing designs developed in-house over five years without incorporating external artists' work.
- Powered by SYSTMS, the Prism 1.0 model generates outputs 84,000 times faster than traditional illustration workflows, with the company describing this framework as "Artistic Intelligence" designed to expand creative expression.
- Doodles plans to produce the first feature-length animated film created entirely through its own AI systems, building on the success of the Pharrell Williams-executive produced animated special Dullsville and the Doodleverse.
- Throughout 2026, the company will introduce prompt-to-image and image-to-video generation capabilities while refining outputs through supervised fine-tuning, enabling fans and brand partners to create content faithful to Doodles' original artistic vision.
Insights by Ground AI
16 Articles
16 Articles
The Daily News - Jacksonville, NC+14 Reposted by 14 other sources
Doodles Trained an AI Model on Nothing But Its Own Art. Now the Company Wants to Make a Feature Film with it
The entertainment company behind the TIFF-premiered Dullsville and the Doodleverse launches Doodles AI, a creative tool powered by a model with no scraped training data and no third-party IP
Coverage Details
Total News Sources16
Leaning Left4Leaning Right2Center4Last UpdatedBias Distribution40% Left, 40% Center
Bias Distribution
- 40% of the sources lean Left, 40% of the sources are Center
40% Center
L 40%
C 40%
R 20%
Factuality
To view factuality data please Upgrade to Premium
