News & Media

We are delighted to announce that Fred Grinstein has joined Starling Lab as a 2026 Fellow, where he continues his research at the intersection of documentary practice, generative AI, and media authenticity.

AI is now embedded across the non-fiction content ecosystem: in historical documentary broadcast series, festival films, independent shorts, and throughout the contemporary creator economy spanning YouTube, Instagram, TikTok, X, and other social video platforms. What was once speculative has become common practice. The question has shifted from whether these tools belong in documentary to how they are being used, by whom, and with what kinds of disclosure and care.

The discomfort many of us are experiencing now is not simply about fake videos. It reflects a growing awareness that the infrastructure we assumed was holding non-fiction media together may never have fully existed. What sustained trust instead was a fragile system of consensus built from institutions, conventions, and shared habits of interpretation. It went unexamined because it worked well enough.

I fed a 1930s Dust Bowl photograph of migrant workers by Dorothea Lange into the latest iteration of Google DeepMind’s Project Genie and within seconds I was walking around inside it. A shack lay ahead. I turned around and the scene behind me (a scene that doesn't exist in the original photograph) was already there, rendered in dusty period-accurate detail, all in keeping with the source photo.

I've been experimenting with 3D reconstruction from archival photographs for the past year at Spatial Lab, and when we started in early 2025, I wasn't sure high-quality single-image-to-3D was even possible. Gradually I developed processes with diffusion models that created convincing 3D scenes from 2D images, but they were painstaking manual processes that took hours to complete. Now, with Apple's SHARP, there is a fast, opensource, image-to-3D process freely available.

At Starling Lab, we've been prototyping what we call a "nutrition label" for synthetic media, intended to be a standardized way of communicating what audiences are actually looking at. After months of experiments and dead ends, here's where our thinking has landed.

For over 180 years, photographs have served as evidence. A photograph was understood to be a mechanical recording of light that existed in a specific place at a specific time. In other words, proof of the existence of an object or event. This evidentiary assumption underpins journalism, law enforcement, insurance claims, scientific research, and countless other domains. With the advent of AI-generated synthetic imagery, that assumption is now obsolete.