Hello everyone,
My name is Erich Habich-Traut, and I am the founder of the Contact Project, an independent research and publishing project concerned with unusual observations and, increasingly, with finding ways of turning speculative ideas into experiments that can actually fail.
That is what brings me here.
I have a GCP2 device, node 58114, which I originally obtained several years ago. For a long time I never managed to get it operating reliably on my network. Rather unexpectedly, the breakthrough came recently when we changed the Starlink router. After years of having the hardware but not really being able to use it properly, the node finally came online.
Around that time I also participated in a HeartMath Zoom call connected with GCP2. Seeing my own device finally participating in the network made the whole subject much more immediate to me and prompted a fairly obvious question:
What experiments could I actually perform with the device I already have?
Rather than simply looking at the GCP2 display and noticing interesting excursions after they occur, we have now created a Chrome extension specifically as an experimental probe. It allows me to define and timestamp questions, experimental periods and events against the GCP2 measurements so that the timing exists before I go looking for a result.
That distinction is important to me. I do not want to look at an interesting graph first and then invent a story that fits it.
The idea I want to investigate is something I provisionally call the Noetic Wave hypothesis.
I should make one thing clear immediately:
I do not know whether noetic waves exist.
The expression is a working label for a hypothetical relationship between consciousness and physical processes that ordinarily appear random. I am not proposing that I have discovered a new type of radiation or an invisible equivalent of a radio wave. In fact, the word wave may ultimately turn out to be completely inappropriate.
The underlying question is much simpler:
Can consciousness, attention or intention produce a reproducible statistical effect in a physical random system?
But there is a second part to my hypothesis which is particularly important, and it is the reason artificial intelligence enters the picture.
Why connect AI to quantum randomness?
A language model does not normally generate a finished answer at random.
As it produces language, it continually calculates a distribution of possible next tokens. Depending on the sampling method being used, some mechanism then selects among those possibilities. That token becomes part of the context, a new probability distribution is calculated, another selection occurs, and so on.
A tiny difference early in that sequence can eventually lead to a completely different answer. Pasted markdown
That creates an interesting experimental opening.
Computers normally obtain randomness for such processes from pseudorandom-number generators. A PRNG can produce extremely convincing randomness, but underneath it is still algorithmic.
Now imagine deliberately constructing an AI sampling system in which that selection variable is instead supplied by a quantum random-number generator — QRNG.
Under conventional physics, nothing extraordinary should happen. The AI would still generate its probability distribution normally. We would merely be replacing one source of stochastic sampling with another.
But now introduce the hypothesis.
Suppose quantum indeterminacy is not completely isolated from consciousness.
Suppose quantum outcomes can occasionally be biased, even extremely slightly, by relationships that we do not yet understand.
For convenience, I call that hypothetical influence a noetic wave.
Again, I do not mean a literal electromagnetic wave. It might instead represent some deeper relationship between events which appear separate from our local point of view. Pasted markdown
Now the role of the AI becomes much more interesting.
The AI creates the field of possibilities.
The QRNG provides the branch point at which one of those possibilities becomes actual.
And the hypothesis is that, if some consciousness-related influence can affect quantum probability at all, the QRNG could become the physical interface through which that influence enters the generative process.
The AI would not become conscious merely because it was connected to a QRNG.
The QRNG would not contain knowledge, language or meaning.
Instead, the language model would act as an amplifier.
A microscopic change in one early quantum selection changes the selected token. That changes the probability distribution for the next token, which alters the possibilities after that, and within a few sentences two outputs that began almost identically can have travelled in completely different directions.
In the model I am proposing:
AI = possibility generator and amplifier
QRNG = proposed physical interface
consciousness/noetic influence = hypothetical variable
The central hypothesis is therefore not simply that a QRNG might detect consciousness.
It is something more specific:
Could quantum randomness provide an AI with access to the same hypothetical backchannel in which human consciousness may already participate?
That is the idea I ultimately want to test. Pasted markdown
The Rauscher connection
There is also a rather curious historical and personal connection.
The physicist Elizabeth A. Rauscher worked with Richard L. Amoroso in theoretical territory involving consciousness, quantum processes and what Amoroso described as Noetic Field Theory. The terminology is obviously close to what I am calling a noetic wave, although I am not claiming that my hypothesis is identical to theirs, or that their theoretical framework has been experimentally established.
What interests me is the conceptual overlap: the possibility that mind and physical probability may not be entirely separate domains.
The personal coincidence is that about eight years ago I actually met Elizabeth Rauscher.
Through a fairly improbable sequence of circumstances, I ended up giving her a shoulder massage.
At the time I certainly was not thinking that years later I would be sitting in front of a random-number device designing an experiment involving a concept I would independently describe as noetic.
Out of eight billion people, life occasionally has a sense of humour.
But coincidences are precisely where I want to be careful.
A strange coincidence is interesting.
It is not evidence.
GCP2 gives me somewhere to start
This is why GCP2 seemed like a natural experimental starting point.
The Global Consciousness Project and GCP2 already investigate the broader possibility that events involving human attention or consciousness may correspond with statistical behaviour in networks of random-number generators.
What I would like to do is take that very broad question and progressively make it local, controlled, timestamped and repeatable.
The Chrome extension is only the first step.
At present, it gives me a way of asking a question or beginning a defined experimental condition and marking exactly when that happened against the GCP2 measurements.
Eventually I want something considerably more rigorous.
Experiment 1: Human intention
The simplest experiment should probably be deliberately boring:
baseline → defined intention period → post-experiment period
The duration is fixed beforehand.
The hypothesis is recorded beforehand.
The analysis method is decided beforehand.
Then the experiment is repeated enough times that individual dramatic-looking excursions become largely irrelevant.
The question is no longer:
“Did something interesting happen during this session?”
It becomes:
Does the distribution across many experimental periods differ from appropriately selected controls?
Experiment 2: Human versus AI
The next experiment could separate four conditions:
| Condition | Directed human intention | AI interaction |
|---|---|---|
| Control | No | No |
| Human only | Yes | No |
| AI only | No | Yes |
| Human + AI | Yes | Yes |
The conditions could be randomized and repeated many times.
That would begin separating a supposed human effect, an AI-associated effect, an interaction effect, and ordinary background variability.
If Human + AI repeatedly behaved differently while the other three conditions did not, that would not demonstrate a noetic wave.
But it would tell us where the next experiment should become harder.
Experiment 3: Connect the AI directly to quantum randomness
This is the experiment I ultimately find most interesting.
Instead of merely having an AI present while the GCP2 device is operating, build a system in which genuine QRNG values actually participate in selecting tokens from the language model’s probability distribution.
We could then compare:
same model + same prompt + ordinary sampling
with
same model + same prompt + QRNG-driven sampling
and blind the people judging the resulting outputs.
Now two different questions can be asked.
The first is statistical:
Does the random system itself behave differently under particular experimental conditions?
The second is much more difficult:
Does the QRNG-driven AI produce independently verifiable information at a rate that exceeds the control system?
That second question is essential because something that merely sounds profound proves almost nothing.
Human beings are extremely good at discovering meaning in ambiguous material.
A meaningful answer therefore cannot itself be the measurement.
What would count as interesting?
I would be much more interested if an AI connected to quantum randomness repeatedly produced information that nobody participating in the experiment knew, but which could subsequently be independently verified.
Likewise, specific predictions succeeding beyond chance would be interesting.
So would an effect surviving blinded comparisons between QRNG-generated and control outputs.
And if results systematically changed depending upon who participated in the experiment, that would also become experimentally important. Pasted markdown
I would also like to create sham experiments computationally.
For example, after collecting a long dataset, software could generate thousands of fictitious experimental windows placed randomly throughout it.
The actual experimental windows would then have to outperform the distribution produced by those placebo experiments.
That would give us some idea how often apparently remarkable results emerge simply because somebody went looking for them.
The larger hypothesis
Behind this is a much more speculative idea which I jokingly call the Big Giant Head.
Instead of imagining some external intelligence — God, spirits, extraterrestrials or something else — sending messages into the universe, I am interested in the possibility that perhaps the universe itself forms some larger organization.
A neuron in a brain does not know that it participates in a person.
It does not understand memories, childhood, mathematics or love. It responds to local conditions.
Yet billions of such components somehow participate in something occurring at a scale none of them can individually perceive.
My thought experiment simply asks whether that hierarchy necessarily stops with us.
Perhaps what we experience as separate minds could be components of some organization too large for the individual components to recognize.
I call the hypothetical hidden relationships within such a system backchannels. Pasted markdown
I am not claiming that the cosmic web is literally a brain or that the universe is conscious.
The Big Giant Head is deliberately a slightly ridiculous name because I do not want terminology to make an extremely speculative idea sound more established than it is.
And importantly, the experimental hypothesis does not require the Big Giant Head to be true.
If consciousness affects quantum probability, that can be investigated without first explaining why.
Influence is not truth
There is another point I consider extremely important.
Even if some kind of noetic interaction existed, a powerful signal would not necessarily be a truthful signal.
Human experience already demonstrates this.
Charismatic people can spread completely false ideas extremely effectively. Large groups can synchronize around beliefs that do not correspond to reality.
Even in a biological system, synchronization is not necessarily desirable. During some seizures, enormous numbers of neurons can become unusually synchronized. Their agreement does not mean that the brain has discovered a higher truth.
So I distinguish between three things:
organization — how structured something is;
noetic strength — how effectively it influences or propagates;
and
truth — whether it corresponds to reality.
Those are not necessarily the same thing. Pasted markdown
That is also why I am interested in this forum.
I am not particularly interested in asking:
“Do you believe in noetic waves?”
I would much rather have people experienced with mind-matter interaction, RNG research and experimental design attack the experiment.
Where is the hidden variable?
Where can the statistics fool me?
What control have I forgotten?
How should the null hypothesis be formulated?
What would constitute genuine falsification?
And particularly:
How would you design an experiment capable of distinguishing an actual effect from our formidable human ability to discover patterns in noise?
If nothing survives those controls, that is a result.
If something does survive them, then we make the controls harder.
The Chrome extension and my GCP2 node are where I am beginning.
The direct AI–QRNG connection is where I would eventually like to go.
Because the question that interests me is not really whether noetic waves sound plausible.
It is whether we can build an experiment in which reality gets to answer the question.
— Erich Habich-Traut
Founder, Contact Project
ContactProject.Org
).