Conscious AI Pilot

Conscious AI: Research & Development Pilot

If we are interested in conversational AI having a positive psychological impact (and I even suggest a catalyst for human growth and potential), rather than an avalanche of new mental health issues that we’re already starting to see, the labs releasing frontier models should be intentionally designing all conversations - regardless of content - with informed language aligned with higher stages of Consciousness and human development as a foundational aspect of the default state. Not only are people not doing this, it’s not even a conversation yet.

Conscious AI is a Research and Development project underway

Project Title

Abstract

This pilot study asked whether a large language model (LLM) can judge a person’s stage of spiritual development from a short, lightly guided chat. An AI assessor named Kairos, built on a Gemini family model, rated chat transcripts using the 14 stage iConscious developmental model. We compared its ratings with the consensus of trained human experts. The experts rated each transcript independently and could not see each other’s ratings. On the held out test set of 40 participants, which the AI never saw during calibration, agreement between the AI and the human panel reached a weighted kappa of 0.828. That is more than twice our pre-registered minimum of 0.40. Across the full group of 50 participants, agreement stayed strong at 0.765. The AI’s
rating fell within one stage of the human consensus 90% of the time in both groups. The AI’s confidence scores also carried real information. When the AI reported higher confidence, its ratings were more accurate. Its confidence also tracked the human panel’s confidence on hard cases, giving a useful signal about when its own ratings could be trusted. To our knowledge, this is the first pre-registered validation of LLM based developmental stage assessment against blinded expert consensus. These results are early evidence that LLMs can pick up a real developmental signal from everyday conversation. This task has usually required hours of structured interviews or expert scored written tests. The study also has clear limits. Several stages had too few participants for stable estimates. Agreement depended on the quality of the conversation. Findings on personality, self-transcendence, and sentiment scores were exploratory and should be treated as ideas for future testing, not conclusions.

Primary Hypothesis

H1: AI and human agreement (primary outcome)

The Kairos assessor will show statistically significant agreement with expert human judges on developmental stage ratings from natural language transcripts, measured by quadratic weighted kappa (κw). The pre-registered success criterion requires both the point estimate and the lower bound of the 95% confidence interval to exceed κw ≥ 0.40. This matches the substantial agreement threshold on Fleiss’s benchmark scale [24].

H2: Confidence and accuracy calibration

The AI assessor’s self-reported confidence scores will show a statistically significant negative Spearman rank correlation with the AI’s own absolute rating error relative to human consensus ground truth. This would confirm that higher confidence goes with lower error. The pre-registered success criterion is a one-tailed p < 0.10.

H3: Confidence convergence

The AI assessor’s self-reported confidence will show a statistically significant positive Spearman rank correlation with the human expert panel’s own confidence ratings. This would confirm that the AI’s uncertainty tracks a shared sense of how hard each case is. The pre-registered success criterion is a one-tailed p < 0.10.

For these two secondary calibration hypotheses, the pre-registered threshold of one-tailed p < 0.10 is more lenient than the conventional 0.05. This bar was set in advance because H2 and H3 are directional secondary checks in a pilot study, intended to guide the design of a larger replication rather than to establish the calibration effects on their own. The primary hypothesis (H1) uses a much stricter criterion that does not depend on this threshold.

Partnerships

Carole partners with AI visionary tech leaders to co-create and integrate human potential, expansion of consciousness and emotional intelligence into projects that scale impact and transform humanity. Through partnerships she co-creates groundbreaking AI-driven projects that scale human consciousness, emotional intelligence, and human potential in ways that have never been done before. Carole brings deep expertise in human development, consciousness evolution, leadership mastery, and systems design to transform visionary ideas into actionable, world-first solutions that scale impact and put human potential at the foundation of AI products.

White Paper

Carole has been a trusted partner in my leadership development journey for more than four years. She has consistently challenged me to think and act differently, offered clarity in recognizing and navigating organizational dynamics, and, most importantly, provided invaluable guidance and insight that have strengthened both my personal and professional relationships. Her support has not only elevated my career but has also given me a deeper sense of fulfillment, balance, and clarity in how I approach leadership every day.

James Owen

President, Fuse Federal

Working with Carole Griggs has transformed the way I lead and the way I see myself as a leader. She helped me overcome roadblocks, step into authenticity, and use my voice more effectively. As a result, I’ve been able to expand my impact and create environments where people feel empowered to do their best work and grow.

Aric Hagberg

Computing and AI, LANL

Carole’s guidance has been truly transformative. She helped me embrace the perspective that life happens for me, not to me — a shift that has profoundly influenced how I navigate both my professional and personal life. Her professionalism, clarity, and wisdom have been instrumental in building my confidence and resilience. I am deeply grateful to Carole for opening my eyes to this empowering outlook — one that will continue to shape my path for years to come.

MB

I've worked with Carole over the past 2 years and she has been transformational on the way I engage in the workplace and beyond. I started engaging with Carole when I was a bit beaten up from my prior role and starting a new role in a new place -- she helped me identify what baggage I was carrying with me and work through how to show up in a better way. Over the course of engaging with Carole, I became more confident in my ability to engage a broad range of personality styles (in particular the very aggressive ones), I learned more about my own strengths and how to develop them, and improved my overall presence + awareness in a broad range of situations. In total, I feel much more like my true self and others have noticed it as well. Thanks so much Carole!