GPT-2 trained on one artist’s essays becomes a surreal creative muse
Merzmensch · x · 2026-07-21
The author says training GPT-2 on their essays gave them back the “spark” of inspiration — as an AI muse that produces neologisms, chaotic structures, absurdist scenes, and surreal worlds.
In the follow-up, they describe the visual/text generation as a kind of collision of meanings, comparing it to Max Ernst, Tristan Tzara, and Daniil Kharms. The attached image shows a retro-style generative artwork interface labeled reMERZ, framed as a dreamlike “glimpse into the past” and “future.”
Related event: GPT-2 Fine-Tuned on Essays Becomes a Surreal Muse(2 posts)→
More from Multimodal
- HeyGen adds a media-sourcing skill for coding agents with 75k images and 10k tracks — HeyGen · 2026-07-22
- Reddit user seeks ComfyUI NSFW text-to-image and image-to-video workflows under 20 GB VRAM — hobbyist2020 · 2026-07-22
- Krea 2 users recommend a two-pass Clownshark sampler setup for sharper image details — listopalafoto · 2026-07-22
- Gemini Omni Flash turns a boat cabin into a cave in Flow by Google — chrisfirst · 2026-07-22
- A simple workflow to turn a photo into an image prompt using Gemini, Grok, or GPT Image — harshitagu72595 · 2026-07-22
- A Reddit user proposes a consistency LoRA to keep anime and game scenes visually stable — ThirdWorldBoy21 · 2026-07-22