User guide

How Noosphere works

Noosphere doesn't search the web: it answers from its own document base covering every psychoactive substance, and shows where every claim comes from.

What it is

A document brain on psychoactive substances — psychedelics, cannabis, opioids, stimulants, depressants, dissociatives, traditional plants and animal-derived preparations — covering their pharmacology, neuroscience, clinical use and harm reduction. Every document that comes in is split, indexed and turned into a network of connected concepts. When you ask, Noosphere retrieves the relevant fragments — whatever language they're in — and writes an answer citing its sources.

Why multidisciplinary

Almost everything written about these substances is written from a single discipline. A pharmacology paper says nothing about ritual context; an anthropology paper cites no receptors; a legal report ignores the neuroscience. Noosphere brings those views together and cross-references them, because real questions rarely fit inside just one.

Pharmacology and neuroscienceMechanisms, receptors, dosing, interactions, long-term effects.
Clinical and psychiatryTrials, indications, contraindications, outcomes and their limits.
PsychologyTherapeutic process, experience, personality, trauma, measurement.
Botany, mycology and ethnobotanyWhich species it is exactly, what it contains, how the plant or fungus builds its alkaloids, and what traditional knowledge surrounds its preparation.
Anthropology and historyTraditional use, ritual, colonial trade, counterculture.
Law and policyRegulation, prohibition, markets, criminalisation, reform.
Harm reductionReal risks, dangerous combinations, drug checking, overdose.

One example: asking about ayahuasca can return its pharmacology (the Banisteriopsis MAOIs and their dangerous interactions with antidepressants), its ceremonial use as documented by anthropology, its legal status across countries and the specific harm-reduction precautions — all at once. That answer exists in none of those literatures on its own.

What you can ask

It works better with specific questions than with single words. These are real, and they get good answers:

How to use it

  1. Ask naturally

    Write as you'd speak: "what does the evidence say about psilocybin in treatment-resistant depression?". It answers in the language you ask in.

  2. Read the sources

    Under each answer you'll find the documents used, numbered and linked to the original. The bracketed numbers in the text point to that list.

  3. Keep asking

    It remembers the conversation: you can say "and what are the risks?" without repeating the topic.

  4. Explore the graph

    On the right you'll see the concept network behind each answer. Click any of them to open its card: what it is, what it relates to and which documents back it up.

Who it is for

Research and academia

Find in minutes what would take an afternoon of cross-searching, with the reference in front of you. The graph also shows which lines of work connect and which stand alone — useful for spotting gaps.

Journalism and science writing

Check a claim before publishing it and reach the original source, not the headline that summarised it. Every fact comes with its document so you can cite it properly.

Harm reduction

Look up interactions, risky combinations and warning signs with the literature behind them. Built to answer without scaremongering and without advocacy, which is how you actually inform people.

Clinical and therapy

Review the state of the evidence for a specific indication, with its trials, sample sizes and contraindications. As a documented starting point, never as clinical judgement.

Students

Enter a new subject through the right door: a clear explanation, the sources to go deeper, and a map of the concepts you need to master.

Informed curiosity

If the subject interests you and you are tired of forums and headlines, here the answers come with their provenance and you can check them yourself.

The graph's three tools

Hubs

The most connected concepts in the corpus: the field's pillars.

Map

Thematic blocks the system detects on its own, such as kappa receptor pharmacology or PTSD therapy.

Path

The chain of relations linking two concepts — say, from ibogaine to opioid addiction.

Where its knowledge comes from

8,802indexed documents
49,369connected concepts
62,998relations between them
1957-2026years covered

The bulk is peer-reviewed scientific literature: over 1,000 different journals — among them Nature, Cell, PNAS, Neuron, Molecular Psychiatry, Biological Psychiatry, Journal of Psychopharmacology and Neuropsychopharmacology — with work spanning 1957 to today. They are found through PubMed and Europe PMC, which are the indexes covering all biomedical literature rather than publications themselves; open-access studies are taken in full, the rest as abstract plus link.

But these substances are not only pharmacology, so the corpus is multidisciplinary. It starts where they start: botany, mycology and ethnobotany — which species it is, what it contains, how the plant or fungus builds its alkaloids and what traditional knowledge surrounds its preparation — together with the phytochemistry, pharmacognosy and analytical chemistry that let you check what is actually in a sample. Added to that: psychology (psychotherapy, cognition, personality, transpersonal, psychometrics), anthropology and history, philosophy of mind, religious studies, sociology and drug policy, law, criminology, toxicology and ethics.

Added to that: documents contributed by the team and the project's collaborators, plus new material the system tracks on its own — clinical trial registries, regulatory agencies, public health bodies and preprint repositories.

Worth knowing

Frequently asked questions

How is this different from asking ChatGPT?

Noosphere does not answer from memory. It searches a closed, cited document base and only says what is in it. A general-purpose assistant can give you a convincing answer with invented references — a well-documented problem; here every claim carries the document it came from and you can open it. It also covers literature general models do not prioritise: anthropology, history, legal reports and harm-reduction material.

Does it search the internet?

No. It answers only from its own document base, built and reviewed by the project. That base does grow continuously — it crawls new publications on its own and the team adds material — but nothing enters without being indexed and citable.

Why does it sometimes say it does not know?

Because it is designed to. If its documents do not contain the answer, it says so instead of filling the gap. Less impressive, far more reliable: an acknowledged gap is worth more than a plausible invented paragraph.

What languages does it work in?

Ask in yours and it answers in yours. Internally it searches in all of them: if the best evidence for your question is in an English paper, it uses it anyway and explains it in your language. That is half a library more than most people ever reach.

Which substances does it cover?

All psychoactive ones: psychedelics, cannabis, opioids, stimulants, depressants, dissociatives, deliriants, traditional-use plants, animal-derived preparations and novel substances. Also alcohol, tobacco and caffeine, which tend to be left out of these conversations for no scientific reason.

Can it tell me what to take, or what dose?

No, and that is a decision rather than a technical limit. Noosphere informs — what the literature says, which risks are documented, which interactions to avoid — but it does not prescribe or recommend use. For any clinical decision, see a healthcare professional.

Can I trust what it answers?

You can check it, which is better than trusting it. Every answer brings its numbered, linked documents; open them. And bear in mind this field moves fast and often contradicts itself: one study does not settle a debate, and Noosphere shows you the debate rather than hiding it.

Does it store my questions?

The site carries no trackers, no third-party analytics and no external web fonts: nobody outside this server knows you were here. Queries are logged anonymously only to measure usage, and are deleted automatically after sixty days.

Can I contribute documents?

Yes, and they are welcome. If you have relevant literature — theses, reports, proceedings, harm-reduction material — write to us and we review it before adding it. That is mostly how the corpus grows.

What does it cost?

There is a free plan with a limited number of daily queries, enough to try it out and for occasional use. Paid plans raise that limit and sustain the project: the corpus, the server and the work of maintaining them are real costs that advertising does not cover, because there is no advertising.

Try it with a hard question

The best way to know whether it works is to ask something you already know well, and check whether it gets it right and where it got it from.