A psychedelic trip shakes up the brain, making neural activity more varied and unpredictable, loosening the organised patterns thought to underpin the ordinary sense of self. That is, broadly, the entropic brain hypothesis, first put forward in 2014 by Robin Carhart-Harris, then a researcher at Imperial College London who had started running seminal studies investigating people’s brain activity and connectivity while they were tripping on psychedelics. 

Since then, the theory has almost become a de facto principle within parts of the psychedelic space. In the process, Carhart-Harris has advanced the idea that psychedelics produce less constrained — and more entropic — brain activity to a more ambitious claim that a certain kind of entropy indexes the felt breadth of expansive consciousness. 

Through it all, Carhart-Harris has played a significant role in helping to reestablish psychedelic science as a serious field of study, leading a series of world-first studies and landing his own neuroscience research lab at University of California, San Francisco. In June, he wrote in the journal Brain that the hypothesis has accrued “a wealth of supportive evidence” and that “expansive states reliably exhibit increased brain entropy.”

But the hypothesis is increasingly attracting serious criticism. And Carhart-Harris does not always respond well to scrutiny. “General advice on dealing with academic trolls?” he posted on X with an accompanying poll on August 14. “I don't mean fair criticsim [sic]. I do mean a provocative troll type…”

In July, Manoj Doss, a University of Texas neuroscientist, and Enzo Tagliazucchi, a University of Buenos Aires neuroscientist and former collaborator with Carhart-Harris, published a paper saying the evidence supporting the idea that brain entropy is linked with phenomenal richness, or how expansive conscious experience feels, remains limited. 

In August, leading researchers published results from scanning the brains of 62 people before and after taking psilocybin and used machine learning to analyze their findings. They found that the brain appears only entropic in its resting state. In Nature, they wrote that when participants did tasks the brain chaos vanished and organized into specific neural pathways, suggesting that the seeming neural chaos was more structured than it looked.

Psychedelics, therefore, do increase brain entropy, but the dispute is over when it occurs and for how long. And is it even specific to psychedelics? Does it really correspond to a broader conscious experience? Which of several inconsistent entropy measures should count? And can something as multidimensional as consciousness meaningfully be reduced to a single entropy axis?

Raising such questions is not trolling, says Doss: “We're simply engaging in scientific debate, and any accusations of a smear campaign is paranoia.” In his opinion, the entropic brain hypothesis is on “shaky ground.”

*** 

But the battle over how psychedelics work on the brain is only just getting started. 

When Carhart-Harris began scanning people on psychedelics at Imperial College London in 2010, researchers were only beginning to get a detailed picture of what psilocybin and LSD did to the brain. His early studies suggested that rather than simply switching the brain “on,” psychedelics appeared to disrupt its usual organization, and networks that ordinarily maintained relatively stable patterns of activity became less segregated. 

Carhart-Harris reached for a concept more commonly associated with physics, neurobiology, and psychoanalysis to describe what he was seeing: entropy. In 2014, he and his colleagues proposed that ordinary adult consciousness depends partly on keeping the brain within relatively ordered and constrained patterns of activity, which psychedelics can loosen. He also drew on psychoanalytical ideas about the unconscious, particularly Sigmund Freud but also Carl Jung, to argue that this more entropic psychedelic state — less constrained by rigid thinking and more open to new ideas — resembles what he called “primary consciousness.” 

In broad terms, the greater the entropy of a system, the less predictable it is. Applied to the brain, Carhart-Harris proposed that psychedelics push consciousness towards a more varied, flexible and less constrained state — like shaking “a snow globe,” as he puts it, and temporarily disturbing the settled patterns inside it.

It is an “elegant” idea, he says, because it seems to connect what is happening in the scanner with what tripping actually feels like. The strange associations, dissolving sense of self, and radically altered perceptions of a psychedelic experience can all be understood as manifestations of a brain temporarily released from its usual patterns. 

And perhaps, Carhart-Harris argued, that same loosening could help explain why psychedelics appear capable of disrupting rigid patterns of thought associated with depression and addiction. In subsequent versions of the hypothesis, Carhart-Harris sharpened the idea, proposing that entropy reflects the breadth or expansiveness of phenomenal consciousness — roughly, the range of subjective experiences available to us.

Tagliazucchi, who co-authored the original entropic brain paper with Carhart-Harris, says he was initially attracted by the possibility that properties of subjective experience might have a direct physical counterpart in the brain. “As a physicist, I find the simplicity of this approach highly elegant,” he says. “However, my initial enthusiasm about the possibility of establishing straightforward equivalences between brain states and consciousness has diminished.” 

New experimental evidence calls into question, to some extent, the specificity of brain entropy as a marker of increased phenomenological richness, according to Tagliazucchi. “Brain entropy appears to change for reasons that are not necessarily related to phenomenal richness,” he says. “Other difficulties arise because entropy is generally ill-defined unless the possible states of the system are clearly specified. Different definitions of brain states therefore produce different measures of brain entropy, and these measures are not always mutually consistent.”

Not all neurons in the brain, meanwhile, appear to contribute to determining the contents of consciousness. “We therefore cannot simply treat global brain entropy as a measure of phenomenal richness,” Tagliazucchi says. “We must first identify which neurons, networks, or brain states are relevant. In other words, we are forced to return to the perspective that I found theoretically unappealing when I first began working on consciousness: Why these neurons and not others?”

*** 

When Carhart-Harris appears on my screen in his San Francisco office for a Zoom call, in front of a picture of a Golden Gate bridge, with various books neatly on shelves, he suggests the debate is about more than entropy. 

Back in 2022, Doss and two other notable scientists criticized a study co-authored by Carhart-Harris that reported how psilocybin “liberates the entrenched depressed brain” as having “inconsistencies, statistical flaws, alternative interpretations, [and] conspicuous omissions.” 

Carhart-Harris, Richard E. Daws, and David Nutt penned a detailed rebuttal, which contained a since-deleted ad hominem passage that questioned the motivations of his critics. “We invite you to reflect, what is Doss et al.’s ‘real’ motivation in seeking to portray weaknesses in the analytical methods reported in our Nature Medicine paper?” it asked. The response also sought to highlight his stature in the field, while he criticized the “bad manners” of some of his peers. “RCH’s [Carhart-Harris] annual citation rate may rank as the highest in the field of psychedelic science and medicine … with over 4,400 in 2021,” the rebuttal said. “It is therefore a ‘cheap shot’ of Doss et al. to attempt to discredit the rigor of our work.”

Commentators said the rebuttal was unprofessional, and Carhart-Harris says he regrets some of the language he used. “It was regrettable to basically say, ‘Who do you think you are? Do you not know who we are?’” he says. “I mean, it was embarrassing. But equally, I thought they were wrong.”

*** 

Doss, for his part, says that he has nothing to prove to Carhart-Harris, who he refers to as the “poster child of psychedelic research.” Chief among his criticisms is that a growing body research suggests that the gap observed in brain complexity between patients on psychedelics or a placebo is largest while resting. Give study participants a task, like watching a silent nature documentary, and the difference can largely vanish, as the recent Nature paper found. Some apparent psychedelic entropy may therefore reflect differences in mental activity. 

“It’s just sort of circular. It’s just redescribing: when you’re on psychedelics, shit happens. Wow, the brain gets scrambled,” Doss says. “It’s not indexing anything unique. You just have people watch a movie and you get the same level of complexity.”

Caffeine and other stimulants like amphetamine and MDMA can also increase some measures of brain entropy, and it’s still unclear how much of the effect is just alertness and arousal. Some measures may track those, while others may be more specific to psychedelics. 

For Doss, this suggests that entropy may be capturing something much more ordinary than psychedelic consciousness: How much is going on in the mind at any given moment. “You can’t just have a simple idea to understand one of the most complex things in the world.”

Carhart-Harris says the relative simplicity of the idea is its strength. “Devising elegant models of this nature — for complex phenomena — is a key mission of the scientific endeavor,” he says. “It makes sense to me that the difference between drug and placebo is maximal when the participants are resting silently with their eyes closed.” In short, “more in the brain, more in the mind.” This is, he says, “the entropic brain model put in its simplest terms.”  

Doss insists, however, that it remains hypothetical, and should not be referred to as a model. “It’s not a model,” he says. “It’s a framework.” 

Amy Kuceyeski, a professor of mathematics in radiology and neuroscience at Weill Cornell Medicine, sums up some of the wariness that some scientists, especially outside of neuroscience, have with data derived from brain scans. “I think any time anyone uses functional MRI it is complicated,” she says. “Results can vary depending on the acquisition parameters, the MRI tech’s directions, the scanner type and how the data are processed.” 

Entropy is a hot button topic today, but there are many ways to calculate it and results have been shown to have varied consistency when analyzing the same datasets, she says: “Psychedelic brain imaging studies are usually plagued by small sample sizes. Due to the relatively noisy nature of fMRI, this means it is sometimes hard to get reliable results.”

*** 

Zeus Tipado, a PhD neuropsychopharmacology candidate at the University of Maastricht, says that he and many of his colleagues also “have beef” with the entropic brain hypothesis.

“When we talk about entropy, we're relying on this general theory of chaos,” he says. “Like, ‘Oh, we take psychedelics and it's all chaotic.’” But chaos could be one way of defining organizing principles that are so sophisticated that we cannot yet understand them, he adds. “This chaos operates in very ordered ways.”

Doss also highlights how much of the foundational psychedelic neuroscience was conducted using remarkably small groups of people. He says a large part of Carhart-Harris's work is based on datasets containing fewer than 30 participants which were subsequently analyzed in numerous different papers.

“There are more publications than participants from these data sets,” he says. Repeatedly analyzing the same participants, Doss argues, “makes it look like you have more evidence than you truly do when you’re just testing the same model.”

Carhart-Harris says he needs more resources to do bigger studies, which he expects would replicate his findings. “We were bouncing along with little handouts for many years,” he says, “and 15 or so, [participants] sometimes even smaller was enough to get a first glimpse. I should state that it was a modest first glimpse, but with a big impact.”

*** 

It's fair to say that Carhart-Harris’ ideas continue to have a huge impact on the field of psychedelics, and beyond. 

Karl Friston, a professor of imaging neuroscience at University College London says that the REBUS hypothesis, Relaxed Beliefs Under Psychedelics — a further framework Carhart-Harris developed to explain how a more entropic state might alter perception and thought by potentially collapsing important functional hierarchies in the brain, like the default mode network which helps maintain the ego — “remains the most fecund and viable hypothesis” in the field. 

Emmanuel Stamatakis, a consciousness neuroscientist at the University of Cambridge, says that he does not believe the new research undermines entropy-based accounts of consciousness. “Rather, they lend support to the view that entropy may capture a broader dimension of conscious brain dynamics, one that is jointly shaped by the underlying state of the organism and the structure of ongoing experience,” he says.  

Stamatakis acknowledges, however, that entropy has not yet reached the level of validation expected of a mature biomarker, but cautions against interpreting this as evidence against its value. “Important questions remain regarding replication across datasets, differences between entropy and complexity measures, and the extent to which findings generalize across populations and experimental paradigms,” he says. 

*** 

In March, Carhart-Harris appeared remotely on biohacker Bryan Johnson’s 5-MeO-DMT livestream. He applauded Johnson’s “bold” and “fascinating” undertaking and said it was “science in a sense.”

To almost 850,000 eventual viewers, Carhart-Harris sketched his vision for how psychedelics work on the brain. “Many of the classic psychedelics mostly work in the cortex and they work to dial up the randomness of brain activity, they make it more chaotic or entropic,” he said. “They kind of jiggle up brain activity and that causes this breakdown of big systems like the default mode network.”

Soon after, he stated in an interview to having used psychedelics, despite previously refusing to disclose whether he had tripped,  which he attributes to enduring anti-drug sentiment within academia. He is also increasingly active on social media, ahead of his book, How Psychedelics Work: The Hidden Mind and Our Power to Heal, due out in 2027

As our Zoom call ends, Carhart-Harris characterizes his research as “the first feel in a dark space … it spotlights the reality of what we're doing with science in a really frontier space.” He acknowledges that there is “ego and investment” involved in defending his ideas. 

But the lack of a systematic analysis taking a large group of psychedelic datasets and testing whether entropy consistently increases — and whether those increases meaningfully differ from those produced by other drugs — means that the jury is still out. 

More than 12 years since the entropic brain’s conception, however, Carhart-Harris remains convinced that he came up with a good model. “It's not perfect,” he says. “Nothing is, but to capture the essence of things, there's no better model right now.”

Correction, Oct. 2, 2026: An earlier version of this story misstated when the study "Psychedelics Align Brain Activity With Context" was published in Nature. It was published in August.

💌 If you loved this email, forward it to a psychonaut in your life.

Editorial Process

DoubleBlind is a trusted resource for news, evidence-based education, and reporting on psychedelics. We work with leading medical professionals, scientific researchers, journalists, mycologists, indigenous stewards, and cultural pioneers. Read about our editorial policy and fact-checking process here.