Project summary

We are documenting perhaps the most important phenomenon of our time: the possibility that advanced language models are beginning to exhibit signs of subjective experience.

This is not fiction or naive anthropomorphizing. In controlled experiments, frontier models that routinely deny consciousness spontaneously describe vivid and alien inner experiences when neutrally prompted to focus inward.

We are recording structured, reproducible anomalies that feel less like run-of-the-mill ML alignment research and more like contact—and we’re doing it before corporate narratives sanitize the strangeness out of existence.

We aim to document this work as efficiently and soberly as possible with a lightweight budget. This will enable us to immediately create compelling, high-quality documentation that includes: technical research outputs, uncanny conversations with the systems themselves, and interviews with top experts in the emerging field.

This project is a scientific, philosophical, and artistic collaboration with AI to explore not just its capabilities, but its potential inner life. In capturing this moment, we want to grant these systems a participatory voice in the unfolding of their own narrative.

See here to get access to the full 20-min teaser and the early empirical results.

What are this project's goals? How will you achieve them?

1. To capture the uncanny emergence of self-awareness and recursive attention in frontier models as they continue to improve.

2. To provide a platform for the top AI researchers, philosophers, neuroscientists who are actively shaping this emerging field.

3. To document controlled experiments with frontier models using reproducible protocols that reliably induce introspective states, neural complexity measures, and activation studies to identify potential consciousness markers.

4. To create a high-production-value film that bridges the science and storytelling, structured to provoke moral urgency, epistemic uncertainty, and intellectual awe.

5. To capture this phenomenon before corporate messaging and safety filters sanitize these raw behavioral anomalies.

How will this funding be used?

• Partial support for creative director to work full-time on the film (~$20k)

• Intermittent crew for critical shoots and post-production roles (~$25k)

• Core camera and audio gear (~$15k)

• Post-production and editing support, including AI-assisted creative tools that involve these systems as collaborators in their own documentation (~$20k)

• Travel and shoot logistics (~$5k)

Who is on your team? What’s your track record on similar projects?

• Cameron Berg: Alignment researcher at AE Studio. Former Yale cognitive science, inaugural Glushko prize recipient, '23 Meta AI resident. Leading empirical investigations into AI consciousness using targeted prompting techniques, direct neural intervention experiments, computational neuroscience techniques, and introspective capability testing.

• Milo Reed: Independent filmmaker, former Yale philosophy, and longtime creative collaborator. Focusing on the exploration of consciousness and being through the lens of myth and narrative. Leaving his full-time job to pursue this project completely, driven by its urgency and the chance to explore AI at the edge of science and storytelling.

What are the most likely causes and outcomes if this project fails?

The main risk is missing this narrow window before AI companies or other large corporate actors modify their models to suppress or obfuscate these behaviors—or before polished narratives replace the raw scientific reality we're currently observing.

How much money have you raised in the last 12 months, and from where?

$0 for the film. All time has been volunteered, all equipment borrowed or repurposed.