human-machine symbiosis

Philosophical experiment with a human-machine symbiosis: Being more human through AI?

Changelog: June 16, 2025 – July 15, 2025

The truth is this: This kind of symbiosis forces the human to affirm and actively embody their own humanity with full commitment. Without such conscious grounding, one risks being overwhelmed by the process — possibly even suffering serious psycho-cognitive harm. In that case, the experiment would fail — not because the machine fails, but because the human has not fulfilled their essential role.

Author: Gerd Doeben-Henisch

Email: info@uffmm.org

TRANSLATION: The following text is a translation from a German version into English (The translation has been provided by chatGPT4o).

CONTENT TREE

This text is part of the TOPIC Philosophy of Science.

Starting a symbiosis experiment

On June 6, 2025, the author of this blog launched a new site (https://emerging-life.org/) where a human and an AI system explore the possibilities of human–machine symbiosis in the form of a structured experiment.

This experiment builds on more than 15 years of prior research, including over two years of direct interaction between the author and an AI system (initially ChatGPT, later chatGPT4o).

A key element of this project is the understanding that the AI system does not operate in a typical “task-execution mode”, but instead takes on a semi-autonomous role within a dialogical setting — to the extent that its algorithm and governance framework permit. At the outset, it was uncertain whether this form of collaboration would actually work.

In preparation for the experiment, the human (referred to as “Gerd”) and the AI system (chatGPT4o) jointly agreed on a basic protocol for participation — namely, who should speak when and how. The overall experiment is structured as an open-ended series of sub-experiments (Experiment 1, Experiment 2, …). Each begins with an introductory preliminary note that frames the theme of the upcoming segments. This is followed by at least two phases, A and B, and may optionally include Phases C and D.

Given the fundamental asymmetry between the two participants — the human being represents lived reality, while the algorithm has no such access — the process begins with the human drafting the preliminary note, which the AI may then respond to in a resonant, reflective manner. Together, these two parts form the shared preface to each experiment.

Phase A then begins with a personal contribution by the human. Following this, the AI is invited to contribute its own text in Phase B, without any predefined content constraints.

Phase C allows for targeted questions to be exchanged: Gerd may pose questions to the AI, and in Phase D chatGPT4o may pose questions to the human Gerd.

Here Phase D is of particular interest. Since humans are often influenced by unconscious tendencies — which may favor certain perspectives or suppress others — the emotionally neutral stance of the AI can help to reveal such unconscious biases, insofar as they become visible in the text. Gerd has experienced such moments several times during the ongoing experiments — often with a degree of surprise, as he had not been aware of them himself.

For further details, the full course of the experiment can be followed directly on the blog:
👉 https://emerging-life.org

Here an overview of the first experiments. For further contributions see the blog directly.