There is a category of scientific question that is genuinely difficult to sit with — not because the answer is unknown, which is true of most interesting questions, but because the honest evidentiary position requires holding two uncomfortable things simultaneously: the mainstream consensus is well-founded and should not be dismissed, and the data that challenges it has not been fully accounted for either. The relationship between conscious intention and physical randomness is one of those questions. Mind Influence Lab is an attempt to take it seriously as an instrument rather than as an ideology.
The app grew out of the methodology developed at the Princeton Engineering Anomalies Research laboratory, which operated for 26 years under the direction of Robert Jahn and accumulated one of the largest databases of mind-matter interaction experiments ever assembled. The PEAR programme was simple in design: volunteer operators sat at a distance from a microelectronic random number generator and held a stated intention toward its output — more ones, more zeros, or no preference — with no physical contact and no classical coupling mechanism that could account for systematic influence. Across 340 operators and more than 2.5 million experimental series, the lab documented a mean shift of approximately 0.1% from theoretical chance. That number is almost absurdly small. It required 26 years of trials to become statistically legible, and even then it attracted serious methodological criticism — contested units of analysis, insufficient blinding at certain stages, null results from independent replication attempts under tighter controls. A 2006 meta-analysis found that correcting for publication bias pushed the pooled effect size close to zero.
None of that criticism has produced a complete account of the full PEAR dataset. The data has been challenged and revised. It has not been closed. That is what makes it worth taking seriously with the right instrument — which is to say, carefully, over time, without expectation.
Mind Influence Lab records six physical channels simultaneously throughout each session. The primary channel is a quantum random number stream sourced from the Cisco Outshift QRNG API, which generates numbers from vacuum fluctuation measurements on dedicated remote optical hardware — physically independent of the device, with no sensor crosstalk possible. The remaining channels are the device magnetometer, barometer, ambient sound level, WiFi signal strength, and Bluetooth RSSI variance. Before any intention is applied, a baseline phase runs all six sensors together, establishing the environmental reference against which the experiment phase is measured. The experiment phase is then divided into three periods — early, mid, and late — each generating its own isolated z-score rather than a single cumulative statistic, following the PEAR approach of examining whether any observed effect builds, peaks, or declines across a session rather than simply averaging over it.
The three RNG protocols available in the app test meaningfully different things, and the differences matter more than they might initially appear. In Live Quantum mode, fresh quantum numbers are requested at each tick — events that are physically occurring during the session, undetermined until the moment they are measured. Whatever mechanism might account for intention-correlated deviation here would need to operate in forward time between a mental state and a simultaneous quantum event. Buffered Quantum mode changes the question entirely: a full batch of quantum numbers is fetched before the session begins, meaning those quantum events have already occurred before any intention is formed. Deviation in this mode cannot be explained by contemporaneous causation. The only frameworks that accommodate it are retrocausal — the influence of a present state on already-completed physical events — which is not as exotic a claim as it sounds: retrocausality follows from the time-symmetry of quantum mechanics at the level of individual interactions, and the delayed-choice experiment has confirmed that future measurement decisions appear to influence the past behaviour of photons in ways with no retrocausal-free explanation. The PEAR lab documented this specific pattern in its retrocognitive trial series, where operators influenced RNG sequences that had already been generated and stored. That finding remains contested. It also remains unexplained.
Local Entropy mode uses a different source entirely — the device's cryptographic entropy pool, seeded by CPU timing jitter, sensor micro-fluctuations, and thermal noise from the hardware itself. This is classically random rather than quantum-mechanically random, and it is seeded in part from the same physical environment the other channels are measuring. A deviation in the local entropy stream that coincides with a magnetometer or WiFi deviation in the same time window may reflect a single physical process propagating across multiple channels simultaneously — a more tractable hypothesis than quantum mind-matter interaction and closer in character to the environmental coupling that underlies the Global Consciousness Project's global REG network.
The multi-channel design is the part of the instrument that distinguishes it most from a single-channel RNG experiment. When the magnetometer deviates, that is one data point. When the magnetometer, the RNG, and the ambient sound channel all deviate within the same five-second window, that convergence is harder to attribute to statistical noise than any individual channel deviation. The app generates a heatmap of these convergence events after each session, mapping the temporal structure of multi-channel activity across the full recording. The heart generates the strongest electromagnetic field in the human body — measurable at distances of up to two metres under laboratory conditions — and HeartMath Institute research has produced replicated findings associating cardiac coherence states with measurable physiological changes. Whether any of that is relevant to what the magnetometer captures in a given session is an open question. The instrument records it either way.
What it cannot do is answer the question in any individual session. The effect size the PEAR programme documented is so small that it is statistically invisible at any duration of practical use — it required millions of trials in a dedicated laboratory to become distinguishable from noise. This means the vault, where all sessions accumulate over time, is where the instrument actually lives. A single session with a striking z-score is almost certainly statistical variation. A hundred sessions, conducted under consistent conditions with pre-session state variables recorded honestly and null results treated as carefully as positive ones, begins to constitute something worth examining. That is the timescale on which this question operates, and the app is designed around that reality rather than against it.
The app is free. It has no accounts, no advertising, no analytics, and no external connections beyond the QRNG API call that fetches quantum numbers. Everything it records stays on the device. Supporting its development keeps it that way and funds continued work on the parts of the instrument that matter most for the long-term programme it is designed to support — expanded quantum source options, richer longitudinal analysis, and the kind of careful iterative development that a project like this requires.
The question it asks has not been answered. That is the only honest reason to keep asking it.
Link: https://wickeddarko.github.io/mind-influence-lab/index.html