Psilocybin doesn’t scramble the brain — it aligns it with context

The standard story about the psychedelic brain is that it becomes desynchronized, more “entropic,” more disordered. A new open-access Nature paper from Adeel Razi’s group at Monash (Stoliker et al., with co-authors including Anil Seth, Robin Carhart-Harris and Gustavo Deco) argues that this is mostly an artifact of averaging over time. When you look at how brain activity unfolds moment to moment, psilocybin shows a hidden order: brain activity becomes more tightly locked to whatever the person is experiencing at that moment.

The study. Sixty-two adults with no previous psychedelic experience were scanned with fMRI and EEG before psilocybin and again after a 19 mg dose. Each session had four conditions: eyes-closed rest, guided meditation, and music, followed by an eyes-open silent movie of drifting clouds. Half the participants later ranked the experience among the most meaningful of their lives.

What the conventional measures show:

• Eyes-closed conditions:
  • Integration of sensory (especially visual) regions with the rest of cortex fell.
  • Integration of associative regions rose.
• Eyes-closed vs. eyes-open: the normally sharp difference between these states largely collapsed.
  • In the visual network, the eyes-closed/eyes-open gap in global connectivity shrank by 85%.
  • On EEG, the eyes-open/eyes-closed difference in alpha power fell by about half.
• Modularity: functional modularity, meaning how segregated the networks are, decreased in every condition.

What the machine-learning view adds. The authors used a contrastive embedding method (CEBRA) to compress each person’s whole-brain activity into a low-dimensional trajectory. They then asked how well a classifier could tell which context (rest, meditation, music, movie) each moment belonged to.

• Context alignment: under psilocybin, trajectories formed tight, distinct clusters for each context. In strong responders, classification rose from about 62% to about 98%.
• Link to experience:
  • Classification accuracy correlated with mystical-experience scores (r = 0.64).
  • It also correlated with next-day “mindset change” (r = 0.40).
  • The strongest links were with positively felt self- and boundary-dissolving effects such as unity, bliss and spirituality.
  • Anxiety and impaired cognition showed no link.
• Why averaging misses it: time-averaged modularity predicted neither subjective experience nor later change. The meaningful organization lives in the temporal dynamics that averaging throws away.
• Where the effect comes from: a perturbation analysis found that the default mode network and the visual network together account for nearly half of the gain. These are the two ends of the brain’s principal internal-to-external processing gradient. When both are altered, the functional boundary between self-model and world loosens.

Embeddedness. The authors call the experiential correlate of this alignment embeddedness: feeling continuous with the environment rather than connected to it as a separate entity. One participant described melding with the MRI machine, the floor, the walls and the air. One month later, the group showed gains in death acceptance, personal meaning and nature-relatedness.

An 8-week mindfulness course before dosing made no detectable difference.

Caveats. The design was open-label, the conditions ran in a fixed order, and the brain–behaviour links are correlational.

The closing paragraph will resonate with anyone who thinks in predictive-processing terms. The authors suggest the felt boundary between self and world is not a given but something the brain actively constructs and maintains. Embeddedness, they write, “may thus reveal less about what psychedelics add to consciousness than about what ordinary neural dynamics keep apart.”

(The above is a Claude Opus 5.5 summary of the article) 

Three reasons why AI is not conscious and perhaps never will be

In the October 22 issue of Nature, Anil Seth, Ned Block, and Giulio Tononi (Tononi from my old UW–Madison home) pool their decades of work in experimental neuroscience, theoretical neurobiology, and philosophy of mind into a pointed Comment. They argue that current AI systems are not conscious and that, on present trajectories, they are unlikely to become so. The piece is partly a response to recent claims from figures such as Richard Dawkins and Geoffrey Hinton, and to Anthropic’s 2026 preprint reporting computational signatures of a form of consciousness in its Claude model.

The authors frame the stakes in practical terms. If AI systems really are conscious, they might suffer, and proliferating them raises serious moral questions. If they only seem conscious, then granting them rights or welfare protections could hobble oversight just when it is most needed, and could divert concern from beings whose sentience is not in doubt. They reject a precautionary stance: uncertainty does not make all possibilities equally likely, and in their judgment false positives (attributing consciousness where there is none) are far more likely than false negatives.

They offer three reasons for skepticism.

1. Intelligence is not consciousness. Consciousness is what disappears in dreamless sleep or under anesthesia, what Nagel described as there being “something that it is like” to be an organism. Intelligence concerns doing, while consciousness concerns being and feeling. The cerebellum makes the point. It holds roughly four-fifths of the brain’s neurons and does sophisticated, adaptive motor computation, all outside awareness, and losing it does not directly affect consciousness. Our tendency to merge the two reflects anthropocentric and anthropomorphic biases, amplified by the seductive power of language. LLMs, trained on vast amounts of human writing about inner life, act as mirrors of our collective digitized past. As the authors put it, “Just because something is intelligent does not mean that it is conscious.” The contrast case is AlphaFold. It is also a powerful neural network on silicon, but because it predicts protein structures instead of chatting with us, almost no one wonders whether it is conscious.

2. Brains are not computers made of meat. The case for conscious AI rests on computational functionalism, the assumption that the right computation suffices regardless of substrate. The authors trace this idea to Turing’s 1936 universal machine and to McCulloch and Pitts’ 1943 logical neurons. They argue the brain-as-computer metaphor has been overextended. In biological brains, what the system does cannot be cleanly separated from what it is. There is no software/hardware divide in living neural tissue. Evidence that substrate matters is therefore evidence against functionalism, and against consciousness in silicon.

3. Leading theories of consciousness are non-computational.

• Integrated information theory (IIT) holds that consciousness requires a physical system whose components are interconnected in specific ways that satisfy the essential properties of experience. Cortical regions meet these requirements. The cerebellum does not, and neither do the hardware architectures that run deep learning, however capable those systems are.
• Biological naturalism proposes that consciousness depends on particular material features of living systems. Candidates include electrochemical synapses, cortex-wide electromagnetic fields, metabolism, and autopoiesis (self-regeneration).

Their conclusion is that the research community’s default position should be that silicon-based AI is not conscious and should not be treated as such. Such systems are, in their phrase, objects rather than subjects, “smart tools, not smart people.”

Readers of MindBlog will recognize this as one side of a debate I have followed for years. The predictive-processing and Agüera y Arcas perspectives I have often featured lean toward functionalism, and the authors’ substrate argument is the crux on which they part company.