< Back to Consciousness & Deep State | Main Deep State Project

Nick Cammarata

AI interpretability researcher and experienced meditator whose exploration of jhana states, psychedelics, and the nature of consciousness bridges cutting-edge artificial intelligence research with ancient contemplative traditions — offering a rigorous, empirically grounded perspective on altered states and the mechanics of subjective experience.

Field	Details
Full Name	Nick Cammarata
Role	AI Researcher / Consciousness Explorer / Meditation Practitioner
Platform	X/Twitter (@nickcammarata), OpenAI research, Goodfire (core contributor)
Status	ACTIVE
Current Location	San Francisco, California
Current Affiliation	Goodfire (AI interpretability lab); formerly OpenAI Interpretability Team
Notable Works	"Zoom In: An Introduction to Circuits" (Distill.pub, 2020); "Language Models Can Explain Neurons in Language Models"; "Multimodal Neurons in Artificial Neural Networks"; featured in Scott Alexander's "Nick Cammarata on Jhana" (Astral Codex Ten, 2022); featured in "Manufacturing Bliss" (Asterisk Magazine)
Category	AI Researcher / Consciousness Explorer

Background

Nick Cammarata is a research scientist who co-founded the interpretability team at OpenAI alongside Chris Olah (who later co-founded Anthropic). His professional work focuses on mechanistic interpretability — understanding the internal representations and circuits within neural networks. He co-authored foundational papers in the field, including "Zoom In: An Introduction to Circuits" published on Distill.pub in 2020, which demonstrated that individual neurons in vision models correspond to recognizable concepts.

Cammarata describes himself as having "an unusually high baseline mood" — a trait so pronounced that as a child he was bullied for his positivity until he learned to adopt an outwardly depressed affect. He is a former startup founder and angel investor who has cultivated a significant following on X/Twitter for his discussions of consciousness, meditation, and the intersection of AI research with subjective experience.

In 2020, Cammarata's interpretability work led him to describe neural networks in terms that echo consciousness research: "It turns out their inner world is beautiful, and I helped try to share a glimpse of it with Zoom In." This framing — treating artificial neural networks as having an "inner world" worth exploring — reflects his broader philosophical orientation: that understanding how information processing systems generate subjective-seeming representations is one of the most important questions in both AI and consciousness research.

As of 2026, Cammarata is a core contributor to Goodfire, an AI interpretability research lab that raised $150 million at a $1.25 billion valuation in February 2026. He has stated: "There is a critical gap right now between frontier research and practical usage of interpretability methods. The Goodfire team is the best team to bridge that gap."

Current Situation

Cammarata occupies an unusual position in the consciousness discourse. He is not a psychedelic researcher, a religious scholar, or a traditional meditation teacher. He is an AI researcher who has independently explored extreme states of consciousness through meditation and psychedelics, and who brings a computational and empirical lens to questions about the nature of mind, subjective experience, and altered states.

His influence operates primarily through X/Twitter, where his threads about jhana, emptiness, and consciousness mechanics have sparked significant interest in the Bay Area rationalist and tech communities. His tweets about jhana meditation catalyzed the creation of Jhourney, a meditation retreat startup, and a Discord community of over 400 members focused on achieving jhanic states. TIME Magazine covered the resulting jhana movement in 2024.

Cammarata does not position himself as either a true believer or a dismissive skeptic regarding altered states of consciousness. He treats consciousness exploration as an empirical project — something to be investigated with the same rigor he applies to neural network interpretability. This makes him a distinctive voice in the DMT and consciousness discourse: someone who takes the phenomenology of altered states seriously while maintaining the analytical frameworks of a computational researcher.

His Perspective on Consciousness

Consciousness as Mechanism

Cammarata's core perspective treats consciousness not as a mystical phenomenon but as a mechanical one — a system with identifiable sub-second operations that can be observed, mapped, and understood. He has proposed that traditional Buddhist terminology obscures rather than reveals these mechanics, offering alternative descriptions:

"My preferred translations of tanha and dukkha are fast-grabby-thing and evil-vibrating-blob. Makes it obvious what kind of thing to look for to see them." — Nick Cammarata, X/Twitter

"Tanha is a grabby sensation so fast that you don't notice it without lots of practice." — Nick Cammarata, X/Twitter

This approach — translating ancient contemplative observations into concrete, mechanistic language — mirrors his professional work in AI interpretability, where the goal is to translate opaque neural network computations into human-understandable concepts.

The Emptiness Framework

Cammarata has described a progression in his meditation practice toward what Buddhist traditions call anatta (no-self) or sunyata (emptiness):

"What I previously identified with is gone. The only reasonable thing to identify with is the emptiness that all sensations arise from and return to." — Nick Cammarata, X/Twitter

This perspective positions consciousness not as a thing to be found but as a space in which experiences arise — a framework that has implications for how he interprets DMT experiences and entity encounters. If the self is already a construction, then entities encountered in altered states may be additional constructions arising in the same empty space, rather than autonomous beings from another dimension.

Jhana and the Neuroscience of Bliss

Cammarata's most publicly documented consciousness exploration involves the jhanas — eight progressive meditative absorption states described in Buddhist texts. After approximately 1,000 hours of solo meditation practice, he achieved first jhana in the summer of 2021. His descriptions challenged conventional neuroscience models:

"If I could have my max ideal everything is 100% perfect casual sex fantasy situation... or sit in quiet in jhana I'd definitely do the latter." — Nick Cammarata, Astral Codex Ten, 2022

"Jhana made me not crave pleasure so much anymore. Cured that 'addiction' via surplus." — Nick Cammarata, Astral Codex Ten, 2022

"Happiness buttons are real and you can press them whenever with ~only positive side effects... they're so non-addictive that you forget about them." — Nick Cammarata, X/Twitter

He described the jhanas as "clean joy," a "sensory overload of pleasure and serenity," and the "freest of free lunches." His account revealed a paradox that challenges standard reward models in neuroscience: maximum hedonic pleasure with minimal behavioral reinforcement. This distinction between "liking" (hedonic experience) and "wanting" (motivational drive) has implications for understanding how consciousness processes reward — and by extension, how DMT and other psychedelics might generate intensely "real-feeling" experiences without those experiences necessarily corresponding to external reality.

Micro-Psychology: The Sub-Second Mechanics of Mind

Cammarata advocates for what he calls "micro-psychology" — the study of rapid, sub-second mental mechanics that conventional psychology largely ignores:

"The secret to happiness turns out to be not doing a specific fast clenching mechanic called tanha that we've known about for 1000y+, is easily reproducible by tons of people, and we don't teach people about ever." — Nick Cammarata, X/Twitter

"Modern psychology is largely about what macro things lead to a happy life, but actually high valence comes from the stopping of subtle nearly-invisible mental movements." — Nick Cammarata, X/Twitter

This framework is directly relevant to DMT entity encounters. If consciousness operates through sub-second mechanical processes that most people never observe, then the "beings" encountered under DMT could be understood as those same mechanical processes rendered visible — the brain's pattern-generation machinery, normally invisible, suddenly manifesting as perceived entities when the usual perceptual filters are disrupted.

On Psychedelics and 5-MeO-DMT

Cammarata has publicly discussed 5-MeO-DMT, the more potent relative of N,N-DMT, noting its therapeutic potential and its relationship to meditative states:

"5-MeO-DMT is the perfect treatment if you just figure out how to reduce the nightmare probability." — Nick Cammarata, X/Twitter

He has described 5-MeO-DMT as "the easiest way to one shot wellness" for a significant percentage of people, and has listed it alongside general relativity, black holes, and consciousness evolution as things that "uncontroversially exist" — suggesting he takes psychedelic phenomenology seriously as data about the nature of consciousness, even if his interpretation of that data skews mechanistic rather than metaphysical.

This positions Cammarata in an interesting middle ground in the DMT entity debate. Unlike Terence McKenna, who treated DMT entities as genuinely autonomous beings from another dimension, and unlike pure materialist skeptics who dismiss entity encounters as meaningless hallucinations, Cammarata appears to treat the phenomenology as real and important data — while remaining agnostic or skeptical about whether the entities have autonomous existence outside the experiencing consciousness.

Bridging AI Interpretability and Consciousness Research

Cammarata's dual expertise creates a unique analytical framework. In AI interpretability, he studies how neural networks generate internal representations — "multimodal neurons" that respond to concepts across different sensory modalities. This work demonstrates that information processing systems can generate rich, concept-laden internal representations without those representations necessarily corresponding to external entities.

The parallel to DMT entity encounters is suggestive. If an artificial neural network can generate a "neuron" that activates for the concept of a face across images, text, and drawings — without there being any actual face — then a biological neural network under DMT might generate vivid, coherent entity representations without those entities existing outside the brain.

Conversely, Cammarata's meditation experiences — particularly the jhanas and emptiness practices — demonstrate that biological consciousness is capable of generating experiences far more intense and "real-feeling" than ordinary waking life. This cuts both ways in the entity debate: if consciousness can generate experiences that feel "more real than real" through meditation alone, then DMT entities might be products of the same generative capacity. But it also means consciousness has capacities far beyond what materialist neuroscience typically acknowledges.

"Neural network interpretability is so fun, it's always nice when things that are important also turn out to be a blast to do... neural network interpretability is like all the beauty of going to the zoo mixed with all the beauty of mathematics." — Nick Cammarata, X/Twitter

"A lot of discussion around artificial intelligence implicitly conflates intelligence with consciousness." — Nick Cammarata, X/Twitter

This last observation is crucial. Cammarata explicitly distinguishes intelligence (information processing, pattern recognition, behavioral competence) from consciousness (subjective experience, qualia, awareness). This distinction means he does not assume that AI systems are conscious simply because they are intelligent — and symmetrically, he does not assume that consciousness experiences (including DMT entities) are "intelligent" or "autonomous" simply because they feel conscious.

Key Arguments

Consciousness has sub-second mechanical operations that can be directly observed through meditation practice — these mechanics are the actual substrate of happiness, suffering, and altered states
Jhana demonstrates that consciousness can generate extreme hedonic states without external input, proving the brain's capacity to create vivid experiential realities from internal processes alone
AI interpretability reveals how information processing systems generate rich internal representations — a parallel that suggests DMT entity encounters may be the brain's pattern-generation machinery made visible
Intelligence and consciousness are distinct phenomena — a critical distinction for evaluating whether DMT entities are autonomous conscious beings or complex information patterns generated by the brain
5-MeO-DMT and deep meditation access overlapping territory — both produce states that challenge ordinary assumptions about the nature of consciousness and subjective experience
The emptiness framework suggests that both the experiencing self and perceived entities arise from the same "empty" ground — neither is more fundamentally real than the other
Modern psychology neglects micro-mechanics — the sub-second operations of consciousness that meditation makes observable and that psychedelics dramatically disrupt

Where He Has Said It

X/Twitter (@nickcammarata) — Primary platform for consciousness discussions; prolific threads on jhana, emptiness, psychedelics, and the nature of mind (2019–present)
"Nick Cammarata on Jhana" — Scott Alexander, Astral Codex Ten, October 2022 — detailed profile of Cammarata's jhana experiences and their implications for neuroscience
"Manufacturing Bliss" — Asterisk Magazine, Issue 6 — feature article on the jhana movement Cammarata helped catalyze
TIME Magazine — "My Week at the Buzzy Meditation Retreat That Promises Bliss on Demand" (2024) — coverage of Jhourney, the retreat startup influenced by Cammarata's public jhana advocacy
"Zoom In: An Introduction to Circuits" — Distill.pub, 2020 — foundational interpretability paper that implicitly frames neural networks as having an explorable "inner world"
Jhourney blog — "A response to ACX's 'Nick Cammarata and Jhana' post" — Jhourney's response detailing Cammarata's influence on their founding
nickcammarata.com — Personal website with writing on meditation and consciousness experiments

The Counterargument

Against the mechanistic interpretation of DMT entities:

Researchers like Rick Strassman and Terence McKenna argue that the consistency of entity encounters across independent users — including specific entity types, behaviors, and "geographies" — is difficult to explain as mere neural pattern generation. If entities were simply the brain's machinery made visible, one would expect far more variation between individuals.
The DMT Consciousness Travel thesis argues that the coherence and interactivity of DMT entities suggests genuine contact with autonomous beings rather than passive hallucination.
Tom Campbell's "My Big TOE" framework proposes that consciousness is fundamental rather than generated by brains — in which case, DMT entities could be genuinely real entities in a larger consciousness system, not products of neural computation.

Against the jhana-as-proof-of-internal-generation argument:

That the brain can generate intense experiences internally does not prove that all intense experiences are internally generated. A radio can produce static on its own, but that does not mean all radio signals are self-generated.
The jhana states themselves may involve contact with real aspects of consciousness that exist beyond the brain — a possibility Cammarata's own emptiness practice does not rule out.

Against conflating AI interpretability with consciousness:

Critics argue that artificial neural networks and biological brains are fundamentally different substrates. Finding that AI generates internal representations without external referents does not prove that biological consciousness does the same under DMT.
The "hard problem of consciousness" — why information processing gives rise to subjective experience at all — remains unsolved. Cammarata's interpretability work addresses the "easy problems" of information processing but does not touch the hard problem.

On Cammarata's credibility:

Cammarata is not a neuroscientist, psychedelic researcher, or credentialed consciousness researcher. His perspectives, while informed and analytically rigorous, come from personal experience and professional analogy rather than formal study of consciousness.
His unusually high baseline mood and exceptional capacity for jhana may make his experiences non-representative of typical consciousness — what works for his neurology may not generalize.

Terence McKenna — Coined "machine elves" to describe DMT entities; treated them as genuinely autonomous beings from another dimension. Cammarata's mechanistic framework offers a contrasting interpretation of the same phenomenology.
Rick Strassman — Conducted the first FDA-approved DMT research in the 1990s; documented consistent entity encounters across subjects. His clinical data is the evidential foundation that Cammarata's mechanistic interpretation must account for.
Tom Campbell — Former NASA physicist whose "My Big TOE" (Theory of Everything) treats consciousness as fundamental and physical reality as a simulation within a larger consciousness system. Cammarata's emptiness framework and Campbell's simulation framework share structural similarities but differ on whether consciousness is generated by brains or is fundamental.
DMT Consciousness Travel — The thesis that DMT enables consciousness to access other dimensions. Cammarata's perspective provides a skeptical but engaged counterpoint to this thesis.

DMT Entity Encounters — Documentation of the consistent entity types reported across DMT experiences, which Cammarata's mechanistic framework interprets as neural pattern generation rather than contact with autonomous beings
RAS Consciousness Filter — The reticular activating system as a consciousness filtering mechanism; Cammarata's "micro-psychology" framework addresses similar territory regarding how consciousness filters and constructs experience at sub-second timescales
Simulation Theory Consciousness — The hypothesis that consciousness operates within a simulated reality; Cammarata's AI interpretability work — demonstrating that neural networks generate rich internal "worlds" — provides a computational analogy for how a consciousness system might generate simulated experiences

Sources

This information was built by Grok and Claude AI research.

Background​

Current Situation​

His Perspective on Consciousness​

Consciousness as Mechanism​

The Emptiness Framework​

Jhana and the Neuroscience of Bliss​

Micro-Psychology: The Sub-Second Mechanics of Mind​

On Psychedelics and 5-MeO-DMT​

Bridging AI Interpretability and Consciousness Research​

Key Arguments​

Where He Has Said It​

The Counterargument​

Related Perspectives​

Related Concepts​

Other Coverage Worth Reading​

Sources​