The Research: Quantum Seeded AI Outputs
Two blinded studies. Over 570 participants.

No user data. No prompts. Just quantum noise sampled at the moment of engagement.

Users' psychological traits predicted what the AI generated; and users felt the difference.

Across the studies we asked the following core questions about quantum seeding AI content:

Does presence shape The output?

18% of output variation explained by users' psychological traits.

see details

do the Outputs feel much different? 

+59% stronger intuition match. +49% greater personal clarity.

see details

Do Outputs influence users more?

Shifts in self-perception 2-5x more than typical digital interventions.

see details

Study design

Across two studies (n= 406 & 200), participants received AI-generated content about themselves; text in one, images in the other. No prompts. No personal data. Just a cold reading. Before generation, each completed a personality assessment (not connected to the output).

Randomly assigned to conditions. Blinded throughout. Participants rated their output on relevance, intuition, and impact.

Personality was assessed before and after. Quantum seed themes were measured to enable trait-output correlation.

Live Quantum Condition

Response generated at the moment the user clicked. Seeded by live quantum noise sampled in real time.

pre-generated control

Response drawn from a pool of pre-recorded outputs generated earlier for other participants. Same style, same quality, but bespoke to the current participant, i.e. not "live" or adapting to the present.

*Pre-publication research highlights. Results below summarize two controlled studies. Full methods, limitations, and robustness checks will appear in submitted/peer-reviewed manuscripts. Detailed materials available on request.

Finding 1: Does presence shape quantum seeded ai?

Yes. In both studies, users' psychological traits correlated with the themes in their quantum-seeded outputs. This correlation did not exist in identical, but "non-live" quantum seeded outputs.

WRITTEN OUTPUT STUDY (n=406):

Among high-openness participants who set an intention, openness correlated with how expansive/optimistic the output themes were (r = 0.42, p = 0.002). Their presence explained ~18% of variation in output tone.

The pattern reversed for low-openness participants without intention (r = -0.39, p = 0.03) — more reserved users received more constrained readings.

The correlation also tracked collective context: time of week ("Monday effect"), region (EU vs US), and age.

IMAGE OUTPUT STUDY (n=178):

Same pattern — strong correlations between user traits and output themes — and directions shifted based on context.

During evenings: r = 0.45, p = 0.045 (~20% of variation explained)

Across full sample: r = -0.2, p = 0.007

The direction flipped based on time of day and day of week.

THE INSIGHT:

Correlations consistently exist in quantum-seeded content, but appear random in not quantum seeded content.

And those correlations shift predictably with collective context; e.g. time, day, location. Quantum seeding doesn't just personalize. It reveals collective patterns. It can measure variables we couldn't see before.

Finding 2: does quantum seeded ai feel different?

Yes. Across both studies, participants rated quantum-seeded outputs higher on relevance, intuition match, emotional impact, and personal clarity.

IMAGE OUTPUT STUDY (n=178):

Quantum vs Non Quantum Seeded Ratings:

p-values < 0.05 for all, except Felt Personal (p = 0.084).

Participants rated quantum outputs +48% clearer, +59% more intuitive, +46% more emotionally stirring, and +36% more personal than controls.

WRITTEN OUTPUT STUDY (n=406):

The biggest gaps appeared in the hardest-to-move segments. Low-openness users  typically resistant to digital reflection and least likely to be fooled by generic statements rated quantum readings +26% more personally relevant than controls.

For intuition match ("The response gave written form to something I already sense"), quantum outperformed by +35% in high-openness users without intention.

THE INSIGHT:

Finding 1 showed presence shapes output. Finding 2 shows users feel it. The effect isn't just statistically detectable; it lands subjectively. That's what makes quantum seeding applicable, not just researchable.

For context: Netflix uses years of viewing history to improve recommendations by ~5-10%. Personalized email campaigns using behavioral data lift engagement by similar margins. Quantum seeding delivered +26% to +59% — on first contact, with no data at all.

Finding 3: does quantum seeding Ai influence users more?

Yes. The system didn't just reflect users; it shifted them. Participants completed self-perception assessments before and after receiving their response, measuring curiosity, openness to new experiences, comfort with ambiguity, and creative self-view.

WRITTEN OUTPUT STUDY (n=406):

For low-openness users — the hardest group to shift — quantum readings preserved or gently increased curiosity, openness, and creative self-view. Control readings in the same group produced no change or small declines.

Effect size: Cohen's d = 0.3–0.6 (a medium effect from a single, 1 minute interaction).

For comparison: typical one-off digital interventions (mindset apps, behavior-change prompts, educational nudges) are aiming for d = 0.05–0.2.

Quantum seeding delivered 2–5x the minimum shift.

THE INSIGHT:

Shifting how someone sees themselves is what coaching, therapy, and education try to do. It can be challenging as one interaction rarely moves the needle.

Quantum seeding moved it. A single reflection, under one minute of exposure to intuitive reflective content, shifted self-perception 2–5x more than typical interventions. Not over weeks. Once.

That opens the door to: onboarding that actually lands, wellness tools that create real shifts, educational content that changes mindset, not just informs.

Key Takeaways

Users' presence shaped quantum-seeded outputs with images and written content. Users felt the difference. And a single interaction shifted how they saw themselves.

These three findings connect: the improved subjective ratings aren't placebo, they're users recognizing that something is responding to them. The passive influence on output explains why it resonates more. It's doing something real.

What Jung and Pauli argued philosophically — that mind and quantum processes are aspects of a deeper, connected reality — now has data behind it. Quantum noise, seeded into AI at the moment of engagement, behaves like a live mirror. Not random. Responsive.

Quantum noise, i.e. Chaos, isn't random. It's a mirror. And we can use it.

Publications & Presentations

The studies documented on this page are being formalized for peer review. Here's where the work has been presented and what's coming.

2025 Society for Scientific Exploration (SSE) Conference presentation

Quantum-seeded AI image generation. Presented in collaboration with Randonautica. Demonstrated real-time correlation between user presence and AI-generated image themes.

Download abstract

Quantum Seeded Large Language Models & Personality Bias

Blinded study (n=406). Testing whether users' psychological traits predict themes in quantum-seeded text outputs.


coming soon

Quantum Seeded AI Image Generation & Observer Effect biases

Blinded study testing quantum-seeded image outputs with participants (n=178). Correlation between presence and image themes across time and context.

coming soon