Link to the code: brain-emulation GitHub repository

'Conscious' (2026): What Dementia Teaches Us About the Nature of Mind

Whole brain emulation research typically asks what to copy: the connectome, the synaptic weights, the firing patterns, the molecular states. It is a question about completeness of capture. Suki Chan’s documentary Conscious, which had its world premiere at the Copenhagen International Documentary Festival (CPH:DOX) on March 14, 2026, asks a different question: what is actually there to copy?

By examining consciousness through the lens of dementia — through people whose minds are gradually and systematically dissolving — the film reveals the fragility and complexity of what makes someone themselves. And in doing so, it provides an unexpected perspective on the mind uploading question that no neuroscience paper quite captures.

What the Film Does

Conscious is not a science documentary in the conventional sense. Suki Chan’s background is in fine art — her practice deals with light, space, and the phenomenology of perception. Her previous work has explored altered states of consciousness through immersive installation. The documentary extends this sensibility to a human subject: a family member living with dementia, filmed over an extended period as the disease progresses.

The film interweaves the personal with the scientific. Chan interviews neuroscientists who study consciousness and memory, clinicians who work with dementia patients, and philosophers who have thought carefully about personal identity. But these interviews are grounded in the specific, irreducibly individual story of a person whose mind is changing in ways that can be described clinically but not fully explained.

CPH:DOX — the Copenhagen Documentary Film Festival, one of Europe’s leading documentary festivals — selected Conscious for its main program. The film has since begun a festival circuit that will bring it to audiences across North America and Europe through 2026.

What Dementia Reveals About Consciousness

The dementia context is philosophically productive in ways that laboratory consciousness research often is not.

Standard consciousness experiments work with healthy subjects in controlled conditions. They measure response to stimuli, self-report of experience, neural correlates of specific perceptual events. These experiments reveal the mechanics of conscious processing. They are less informative about the nature of the self — the long-term, narrative, socially embedded identity that persists across time and gives consciousness its felt character of being someone in particular.

Dementia attacks that self directly. Early-stage Alzheimer’s disease produces episodic memory loss: events from the recent past disappear first, while remote memories remain relatively intact. Procedural memory (how to do things) is typically preserved longer than declarative memory (facts and events). Personality and emotional reactivity change in ways that depend on which brain regions are affected.

What dementia patients demonstrate is that the self is not a unified thing that either exists or doesn’t. It is assembled from components — memory, personality, social recognition, narrative coherence, emotional resonance — that can degrade independently. A person with advanced dementia may recognize family members by feeling (comfort, familiarity) without being able to name them. They may lose their autobiography while retaining their characteristic humor. They may become, in some respects, a different person while remaining, in other respects, the same one.

This is important for mind uploading. The standard WBE framework treats the target of preservation as the brain’s information state at a moment in time: snapshot the connectome, simulate the dynamics, preserve the mind. But Conscious asks: which mind? At which stage of its development and dissolution? And which components of selfhood matter most?

The Temporal Problem

Conscious connects, obliquely but productively, to a line of 2026 consciousness research that examines temporal dynamics. Bennett’s AAAI 2026 paper (covered earlier on this blog) argued that sequential digital computation cannot host consciousness because conscious experience requires genuine simultaneity — the “chord” of experience cannot be sequentially constructed without losing its character as a chord.

The dementia case raises a related but distinct temporal question: how much temporal extension does a self require? A healthy mind is not just an instantaneous information state. It is a history: a narrative that connects the current moment to a specific past and anticipates a specific future. The dissolution that dementia produces is not just the loss of information but the unraveling of temporal continuity — the patient’s sense that the present moment is embedded in a personal story.

If what matters about personal identity is psychological continuity through time — the connectedness of memories, personality, and narrative across moments — then a mind upload that captures a single snapshot, however complete, misses the temporal dimension of selfhood. A snapshot is a chord without the music before and after it.

What Needs to Be Preserved That a Connectome Doesn’t Capture

The dementia perspective highlights several components of selfhood that standard WBE frameworks treat inadequately:

Emotional memory and attachment. Some dementia patients lose declarative memory of family members while retaining emotional warmth toward them. This suggests that emotional attachment is stored through mechanisms (limbic system, body-based memories) that may be partially independent of cortical episodic memory systems. A connectome capture focused on cortical architecture might miss these.

Procedural and implicit memory. Skills, habits, and bodily routines are often preserved in dementia long after explicit memories are gone. These are not in the neocortical areas that most WBE proposals focus on. The bandwidth bottleneck analysis has already shown that even episodic memory encoding (in hippocampal circuits) resists BCI-based readout; procedural and implicit systems are even less understood.

The social substrate of self. Family members often report that dementia patients’ personalities can stabilize or partially revive in familiar social contexts. The presence of specific people, specific places, specific routines activates aspects of selfhood that seem dormant in unfamiliar environments. This is consistent with the 4E cognition perspective — self is not purely internal to the brain but is maintained in the interaction between brain, body, and social environment.

The Emotional Dimension of the Mind Uploading Argument

Conscious also accomplishes something that scientific papers and philosophical arguments do not: it makes the stakes feel real.

The abstract question — “is continuity of personal identity preserved in a digital transfer?” — becomes concrete when it is grounded in the experience of watching someone you know lose themselves. The dementia narrative makes visible what would actually be lost if a mind uploading procedure failed to preserve psychological continuity — not just the theoretical wrong of identity discontinuity, but the specific loss of a specific person.

This emotional grounding does not resolve the philosophical questions. But it changes the quality of attention with which those questions deserve to be asked. For audiences who encounter the mind uploading discussion through Chan’s film rather than through the academic literature, the framing is appropriate: this matters because specific people matter, and the question of what preserves them is not abstract.

Official Sources