gen‑ai.news
← Back
Multimodal

Let us filter AI slop, you cowards

Let us filter AI slop, you cowards

Over the past year, platforms including YouTube, Instagram, and TikTok have introduced or expanded content authentication systems designed to flag AI-generated material. Many now apply automatic labels to images, videos, and audio that have been identified as synthetically produced. The intent is transparency - letting audiences know what they are looking at - but the practical effect on how content is surfaced and consumed has been limited.

The core argument gaining traction among users and commentators is that labeling and filtering are two different things. A label tells you what something is after you have already encountered it. A filter lets you decide in advance whether you want to encounter it at all. For people who find heavily processed AI imagery distracting, low quality, or simply unwanted, no amount of labeling changes the fact that the content still fills their feed.

The technical groundwork for user-side filtering already exists in a partial form. Platforms have built classification pipelines capable of identifying AI-generated content with enough confidence to attach a label - the same signal that could, in principle, be used to deprioritize or hide that content for users who opt in to such a preference. The gap is not infrastructure so much as a product decision about whether to offer that level of control.

The broader context is that AI-generated content has increased substantially in volume across social platforms, and quality varies widely. Some creators use generative tools thoughtfully as part of a defined workflow, while a separate category of low-effort, algorithmically optimized imagery - sometimes called slop - is produced primarily to generate engagement with minimal creative input. Critics point out that lumping both under a single label, with no filtering option, puts the burden entirely on the viewer rather than giving them any practical recourse.

Enjoy this story? Get the next one in your inbox.

Twice a week: the most important stories in generative image and video AI, distilled into a 2-minute read.

Free. Unsubscribe any time. No spam, ever.

Your next read

No image
Multimodal

Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers

NVIDIA and Hugging Face have joined forces to bring large-scale fine-tuning of image and video diffusion models into the NeMo Automodel framework, integrated with the Diffusers library. The collaboration aims to make distributed training more accessible for teams working with models that would otherwise be difficult to fine-tune on limited hardware. The result is a more streamlined path from a pretrained diffusion model to a customized one, without requiring deep infrastructure expertise.

No image
Multimodal

Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort

Thinking Machines Lab has released Inkling, a 975B-parameter open-weights multimodal model built on a Mixture-of-Experts architecture that keeps only 41B parameters active at any given time. Licensed under Apache 2.0, it accepts text, image, and audio inputs and offers a 1M-token context window. Rather than competing for top benchmark rankings, the model is positioned as a customizable base with adjustable reasoning depth.