Mindfulness ⚛️ · Studio Toriumi
No religion, no enlightenment, no spirituality. What is here is
a way to breathe and an accurate account of how strong the
evidence for it is.
The most important thing first: the famous finding that eight weeks of meditation
changes the structure of your brain did not replicate in the largest, best-controlled
attempt. It is still worth doing. But the reason is not the one you have been told.
The reading can wait. The fastest way to find out whether this does anything for you is your own body. No sign-up, no install.
Press Start
They do not need to be accurate. They exist only so you can see your own change later.
If it feels wrong, stop. This is not an endurance exercise. Breathlessness, rising anxiety, a sense of not quite being yourself — stopping is the correct response to those, not a failure. See it does not suit everyone.
"I think I concentrated well" is worth very little. The breath-counting task turns attention into an accuracy score, using a validated procedure.
It is simple. Breathe normally and count each exhale from 1 to 9. Press the left button for 1–8, and the middle button only on 9. Then start again at 1. If you lose track, press the right button to restart the count.
Here is the trick: when your mind wanders, you sail straight past 9. So the proportion of cycles where you actually landed on 9 is the proportion of time your attention held. You cannot inflate it, because it is not a self-report.
—
The limits of this measure, stated plainly. Breath counting was validated across four experiments with more than 400 participants, and independently replicated in 2018. However, a 2025 study found its validity is limited in highly ruminative adolescents. It is not a ruler that means the same thing for everybody.
Records are stored in this browser only. Nothing is sent to a server. There is no account to create.
Do not conclude from this that it worked. There is no control group. If you start on a bad day, anything you do will be followed by a better one (regression to the mean). If you expect improvement while rating yourself, you will find some (expectancy effects). What this actually shows is whether you kept going. That is still worth knowing.
There is one fact about mindfulness research that cannot be worked around: what you compare it to changes the answer almost completely.
Compared with people who did nothing
Effect size g = -1.03 / large
Compared with an equally demanding alternative activity
Effect size g = 0.03 / essentially none
Against a waitlist, meditators improve a great deal. Against people who spent the same time, with the same expectation, doing something else — exercise, a reading group, a relaxation course — the difference mostly disappears. No reliable advantage over other active alternatives has been demonstrated.
This does not mean meditation is worthless. The reading that stays faithful to the evidence is: doing something regularly, with the expectation that it helps, has an effect, and mindfulness is a perfectly good vehicle for that. It is free, needs no equipment, works anywhere, and has relatively few side effects. Those are real advantages.
276 adults randomised for eight weeks to MBSR or escitalopram, a standard first-line drug. Non-inferiority was demonstrated (Hoge et al., 2023, JAMA Psychiatry). Not better than the drug — not worse. Having one more option that works still matters a great deal.
Four conditions compared over one month. Cyclic sighing — emphasising a long exhale — produced greater mood improvement and a larger drop in respiratory rate than mindfulness meditation (Balban et al., 2023, Cell Reports Medicine). That result is why breathwork is the default on this page.
For recurrent depression, the form combined with cognitive therapy has reasonably well-established relapse-prevention effects. This is the closest mindfulness gets to clinical treatment — and it is treatment, which is a different question from self-improvement in healthy people.
Structure may not change, but reduced amygdala reactivity to emotional stimuli and increased amygdala–ventromedial prefrontal connectivity were observed (Kral et al., 2022). What legitimately changes is how the brain responds, not its shape.
Changes in the network active during mind-wandering are widely reported, but mostly in small samples. On attention tasks, you get better at the task you trained on; whether that transfers to everyday focus is a separate question and not settled.
Increased hippocampal grey matter, a smaller amygdala — a widely cited result. The largest and most rigorously controlled replication attempt found none of it (Kral et al., 2022, Science Advances). No group differences in grey matter volume, density, or cortical thickness, at the whole-brain level or in regions of interest. Under about 22 minutes a day for a few months is unlikely to change structure.
This site uses quantum random numbers, and there is no evidence whatsoever that this improves anything. Please do not expect it to. Why it is here anyway is explained in the last section.
More variety does not help you keep going. These are ordered by how clear the evidence is. Work down the list and keep the first one that sticks.
Inhale through the nose, add a second short inhale at the top, then a slow, long exhale through the mouth. Five minutes a day. The only method that beat meditation in a head-to-head trial, and the easiest of the four. Start here.
Four seconds in, hold four, four out, hold four. Effective in the same study, but behind cyclic sighing. If holding your breath feels bad, do not force this one.
Five seconds in, five out, tuned to the band where heart-rate variability peaks. An older research base than the others, but the effect on mood is less direct.
Move attention through the body in sequence. A core MBSR component, though few studies isolate its effect on its own. Easier to fall asleep during than the breathing practices, which makes it a reasonable one for bedtime.
Do not agonise over posture, place, or time of day. There is no good evidence that these determine the outcome. Whether you keep going matters far more, so pick whatever form survives contact with your actual life — a chair, a train, anywhere.
Most apps leave this out, so it goes here. Meditation is not universally safe.
A systematic review of adverse events found a pooled prevalence of 8.3%. The figure varies sharply with method: 3.7% in experimental studies, 33.2% in observational ones. What gets reported is anxiety, depression, depersonalisation (not feeling like yourself), and more rarely psychotic or manic symptoms (Farias et al., 2020).
So roughly one person in twelve has some unpleasant experience. Most are mild and pass. Not all of them.
Do not start unsupervised if any of these apply to you: you are in treatment for a psychiatric condition; you are living with the effects of trauma; you have been told you have dissociative tendencies. Talk to your doctor or a qualified professional first. This site is not diagnosis or treatment. If you are in distress right now, reaching medical care matters more than any practice here.
This site borrows from how Headspace and Calm are built. Having borrowed, it seems only fair to say where their evidence is thin.
There is a funding problem. A systematic review of randomised trials evaluating Headspace and Calm flagged conflicts of interest — a large share of the studies involve the app makers themselves (JMIR Mental Health, 2022). That does not mean the apps do nothing. It means the numbers in the marketing were often not produced by a neutral third party.
There is an attrition problem. Dropout across app trials ranges from 0% to 84.7%, median 23.4%. In 39% of studies, dropout was higher in the app group than in the control group. Look only at the people who stayed, and of course the results look good.
Genuine positive findings do exist — Headspace trials report improvements in stress and life satisfaction. But the figures you meet in an advert usually sit on top of the caveats above.
There is a ⚛️ in the name, so this section owes you a straight explanation.
The conclusion first: quantum randomness does not make meditation work better. There is no evidence for that at all. Plenty of people connect "quantum" to "consciousness." This site does not take that position.
So what is it for? Exactly one thing: it is a guarantee that this practice is not optimised to keep you here.
In an ordinary app, a server decides what you see next. Which order raises retention? A/B test it, ship the winner. That is not evil, but it is a choice made for time-on-site, not for you. On this page, the order the practices appear in and the route the body scan takes are decided by bits measured on IBM's quantum hardware. Nobody — including me — could have known them in advance.
A computer's "random number" is a sequence computed from a seed. Multiply, add,
take a remainder. Feed in the same seed and you get the same sequence, every time.
Math.random() is the same idea with more complication.
Cryptographic randomness — crypto.getRandomValues() — is better: it mixes in
noise the operating system has collected, including from the hardware random number
generators built into modern CPUs. But the shape is unchanged: gather a seed, run it
through a deterministic computation. Raw hardware output is not what comes out the other
end, which is a common misconception.
Quantum measurement is different in kind. The result is not determined by any variable that existed beforehand. There is no seed — not a hidden one, none at all. That is not "we haven't found it yet"; it is a mathematical result, Bell's theorem, confirmed experimentally. The experimental verification won the 2022 Nobel Prize in Physics.
The bits in use are shown under the guide when you start. When the hardware pool runs dry, the page falls back to your device's cryptographic randomness and says so. Continuing to claim "quantum" at that point would be a lie.
Other things running on the same hardware bits: Quantum Roulette, Quantum Draw, Quantum Randomness Beacon, Quantum Trip Solver.
Studies referred to above. The summaries are mine, and so are any errors in them.