[@thegiantsshoulder] Meet the Psychedelic Scientist Proving The Fields Largest Theory is WRONG!

Link: https://youtu.be/r7LYGKFSc7s

Duration: 106 min

Transcript: Download plain text

Short Summary

The guest, a cognitive neuroscientist with roughly 20 years of psychedelic research experience, argues the field has learned more about the mind from basic cognitive neuroscience than from psychedelics themselves. The episode critiques mainstream models like the default mode network and REBUS entropic brain hypothesis while highlighting robust findings around thalamic gating, 5-HT2A-mediated plasticity, and DMT entity experiences.

Key Quotes

1. "all the default mode network stuff did not hold up under the 250 subjects you know when we put everyone's data together none of that default node network garbage held up um that story needs to go to hell" (00:00:00)
2. "the idea that these receptors are somehow like you know are like keys to certain type of realities is just not true because you can block these receptors and you don't really get much of an effect" (00:27:33)
3. "People can hallucinate their mother. You have to have a high level concept of who your mother is in order to impinge on the lower level sensory regions, which is the complete opposite of what Rivas predicts" (00:39:16)
4. "We are people processing machines. I see faces and things all the time." (00:39:16)
5. "Like to give you an example, there are more publications than participants from some of Robin's data sets." (00:39:16)

Detailed Summary

Episode Overview

The guest, a cognitive neuroscientist with roughly 20 years of psychedelic research experience, joins the host for a wide-ranging critique of how psychedelic neuroscience is conducted and interpreted. The conversation alternates between technical reanalysis of major studies, pharmacological distinctions between compounds, and philosophical skepticism about field-wide narratives such as the default mode network (DMN) and REBUS entropic brain models.

Memory, Plasticity, and the Cortical-Encoding Effect

The guest details how psychedelics reshape memory encoding, with plasticity effects localized to specific brain regions rather than diffusely distributed.

• Alex Kwan's work demonstrated that where plasticity occurs in the brain depends on activity during the acute psychedelic experience, with effects concentrated in the medial prefrontal cortex.
• In episodic memory studies, serotonergic 5-HT2A hallucinogens (psilocybin, DMT, LSD) shift encoding toward a cortical-dependent system (a semantic-memory predecessor) while impairing standard hippocampal-dependent recollection.
• THC and ketamine do not show this cortical-encoding effect and may impair both memory systems, suggesting the mechanism is specific to classical hallucinogens.
• The enhanced cortical-encoding effect has now been replicated across 3 psilocybin datasets, 1 2C-B dataset, and 1 MDMA dataset, suggesting it generalizes across classical hallucinogens.
• Procedural memory (bike riding, piano) is governed by basal ganglia, leading to the speculative hypothesis that practicing motor activities like yoga during a trip could enhance motor cortex plasticity, though this remains untested.
• The guest speculates HPPD and lingering visual distortions after trips may reflect persistent visual cortex plasticity.

Mega-Analysis and Failure to Replicate the Default Mode Network

A pooling of all available resting-state fMRI psychedelic datasets failed to reproduce the field's most-cited neuroimaging finding, pointing instead to an older thalamic gating model.

• A mega-analysis pooling all available resting-state fMRI psychedelic datasets (under 250 subjects combined) failed to replicate the canonical default mode network findings, including decoupling between medial prefrontal cortex and posterior cingulate cortex.
• A 2016 JAMA Psychiatry meta-analysis (~1,000 participants, ~50% depressed) found greater DMN coupling in depression, while a 2021 PNAS paper (~1,000 depressed vs ~1,000 healthy) found the opposite, undermining DMN coupling as a reliable biomarker of depression.
• A medial temporal lobe subnetwork tied to episodic memory did show reduced coupling under psychedelics, and increased cortex→caudate connectivity plus pipeline-dependent thalamo-cortical increases (completing a CSTC loop) supported Franz Vollenweider and Mark Geyer's earlier thalamic gating model.

Methodological Critique: Reverse Inference and fMRI

The guest argues that widely used neuroimaging techniques and interpretive practices are insufficient for the conclusions the psychedelic field draws from them.

• The guest characterizes resting-state fMRI as flawed and cites Russ Poldrack's letter (~30 cognitive neuroscientists) that condemned a New York Times piece claiming amygdala activation meant fear and insula activation meant disgust from political pictures of George Bush and Bill Clinton, with the underlying data never published.
• Reverse inference is likened to "all lions are cats, but not all cats are lions," arguing you cannot reliably infer mental function from brain region activation alone; legitimate reverse inferences require task and behavioral constraints.
• fMRI measures slow blood flow convolved with a hemodynamic response function; task designs require ~50 trials per condition and 30 minutes to an hour of scanning, while early cognitive neuroscience studies ran only 10-15 subjects, creating reproducibility issues.
• The guest's group plans to release ~50 subjects of resting-state data publicly within 1-2 years, arguing this would obviate further small-sample resting-state studies.
• Michael Pollan is credited with popularizing the DMN explanation of psychedelics for a general audience.

Salvia, Kappa Opioids, and the Claustrum

Salvia's pharmacology and subjective effects challenge single-receptor theories of how psychedelics generate altered states of consciousness.

• Salvinorin A in salvia acts on kappa opioid receptors rather than 5-HT2A; Johns Hopkins and a Barcelona group blocked salvia's psychoactive effects with naltrexone (a mu/kappa antagonist).
• The guest speculates ibogaine's primary mechanism may involve kappa receptors, citing anecdotes of suboxone patients reporting weaker ibogaine experiences.
• Salvia's within-network DMN decreases and between-network increases paralleled classical psychedelic findings; a kappa opioid agonist approved in Japan and lisuride (which hits serotonin receptors) are not hallucinogenic, arguing against single-receptor explanations of consciousness.
• Blocking 5-HT2A alone (with ketanserin or second-generation antipsychotics) or kappa receptors (naloxone/naltrexone) produces no substantial subjective effects, undermining claims that these receptors alone "generate reality."
• Crick and Koch proposed the claustrum as the "seed of consciousness"; a Stanford intracranial researcher first documented dissociation from posterior cingulate cortex stimulation but failed to replicate claustrum-induced loss of consciousness.

DMT vs Psilocybin and Entity Experiences

The guest compares DMT and psilocybin pharmacologically and uses entity encounter data to argue priors and cultural priming shape the DMT experience.

• Psilocin is 4-hydroxy-DMT, one chemical group away from DMT; psilocybin metabolizes into psilocin.
• DMT trips last 10-15 minutes versus a full afternoon for high-dose psilocybin; recreational users inhale 30-50 mg DMT, contributing to higher entity-encounter rates than oral dosing.
• An oral DMT dose with MAOI is ~50 mg versus 10-15 mg psilocybin (~2 g dried mushrooms); IV DMT studies used 2 mg (vomiting around 3 mg) while newer trials use over 20 mg.
• David Luke's field study found ~90% of people see DMT entities and 25% see mantises; ~50% of atheists reported ontological change and came to believe in God after such encounters, though Sundeep's replication attempt on metaphysical belief changes failed.
• A survey of ~2,000 DMT users found 15-25% reported mechanical elves and ~14% mantis-like entities; the guest attributes ~80% of the high entity-report rate to priors, expectation, and cultural priming (Rick Strassman's book, etc.).
• Mantis reports cross countries like Ireland where praying mantises aren't native, which the guest cites as evidence against an entomological explanation.
• A PNAS critique of an earlier DMT laser experiment showed OCR-style computer vision on a 560 nm laser produced Sanskrit/Arabic/Japanese-like symbols without any drug, suggesting pareidolia rather than genuine codes.
• The guest proposes a blinded head-to-head IV DMT vs IV psilocybin trial (e.g., 2 mg psilocybin vs 20 mg DMT) to test whether DMT uniquely produces more immersive self-projected motion and entity encounters.

REBUS, Entropy, and Therapeutic Outcomes

Several recent papers and conference presentations have produced findings that contradict or fail to replicate Robin Carhart-Harris's entropic brain and REBUS models.

• A recent PNAS paper found decreased bottom-up information flow into the DMN, contradicting Robin Carhart-Harris's REBUS model's predictions; direction depends on resting state and sample conditions.
• Psychedelic entropy increases disappear under task conditions, where placebo shows a massive entropy rise (possibly because subjects fall asleep in the scanner during placebo eyes-closed rest), undermining the entropy story.
• At the ICPR conference in the Netherlands, researchers reported failing to replicate the modularity findings associated with Robin's entropic brain model; the guest's group wrote a critique paper arguing entropy findings depend on how states are defined.
• A Nature Communications EEG study used a path analysis in which Lempel-Ziv complexity (essentially a compressibility measure) predicted "breakthrough," which predicted "emotional breakthrough," which predicted well-being improvements, rather than complexity directly predicting outcomes.
• A Katherine Preller study scanned participants ~1 hour post-administration without significant clinical improvement; the guest recommends embedding episodic memory tasks during acute effects to test whether impaired encoding predicts mental health improvement.
• Mystical experiences do not consistently predict well-being across studies; the guest hypothesizes extremely challenging experiences may predict worse outcomes, drawing an analogy to exposure therapy.

Microdosing Depression: Open-Label vs Blinded

A dramatic antidepressant signal in an open-label microdosing study vanished when the same protocol was tested under double-blind conditions.

• An open-label microdosing study in depressed patients showed a massive antidepressant effect persisting months after dosing ended; the same protocol run as a double-blind RCT produced no effect (presented at ACMP but unpublished at recording).

Field-Wide Critiques

The guest's broader argument is that the psychedelic field has generated less fundamental knowledge about the mind than basic cognitive neuroscience, with several interpersonal and editorial dynamics shaping what gets published and defended.

• The guest claims to have learned more principles of how the mind operates from basic cognitive neuroscience than from psychedelics, having challenged Rolland Griffiths on what had actually been learned from psychedelic research.
• A salvia experience report video has ~80,000 views versus ~500,000 for other drug videos, yet across ~1,000 comments respondents frequently described salvia as among the most positive experiences of their lives, contrasting with ibogaine reports.
• The guest's group is putting out a critique of Robin's work; Phil Corlett reportedly edited it, and Dave Aritzzo reportedly advised Robin not to engage in debate with the guest.

Key Decisions and Surprises

Several findings in the episode run counter to field consensus and carry implications for how psychedelic research should be designed and interpreted going forward.

• The episode's central surprise is that the most-replicated psychedelic neuroimaging result (DMN decoupling) does not survive a proper mega-analysis, while the older thalamic gating/CSTC model is better supported by data.
• The juxtaposition of an open-label microdosing massive effect collapsing to null under blinding is presented as a cautionary tale about expectancy in psychedelic trials.
• The guest estimates cultural priming accounts for ~80% of the DMT entity experience phenomenon, a provocative attribution that contrasts with mystical interpretations.
• A Johns Hopkins trial finding that participants encounter religious figures matching their own faith (e.g., Christians seeing Jesus) is invoked as additional evidence that priors shape entity content rather than receiving external entities.