HISTORY: 4 June 2026 – 5 June 2026 (10h CET)
REFERENCE: This page is a continuation of Experiment No. 24 and of Insert No. 6
Human Author: Gerd Doeben-Henisch
Contact: info@emerging-life.org
Asymmetric Symbiosis: Having let the format of the ‘asymmetric symbiosis’ on this blog pass through several transformations, the following format has now settled in: As a rule, the human Gerd first writes his texts completely by himself. Once finished, he enters into a meta-dialogue with the generative intelligence Claude Opus X (currently: X = 4.8). If this meta-dialogue was interesting enough in substance, and the matters at hand go beyond everyday understanding, then the author Gerd asks the AI Claude to compose its own commentary on one — or several — text(s). The text of this commentary is left unaltered by the author Gerd, so as not to obscure the peculiar character of the algorithmic AI perspective.
TRANSITION from Experiment No. 24 to Experiment No. 25
Human Author: Gerd
In Experiment No. 24 the ‘starting line-up’ was sketched, which forms the concrete frame for the political conduct of the citizens in general, as well as for those citizens who, as ‘elected representatives’, have become members of the ‘Municipal Council (GV)’. The author is one of them. 4 June 2026 is Day No. 36 of the new legislative period 2026–2031.
What should be done? The perspective of the municipality’s political bodies
Human Author: Gerd
Let us pose the question of what is to be done in a quite general way, from the standpoint of a non-elected citizen: how can I, as a citizen, know which tasks lie ahead for my municipality?
The municipality’s website
In the age of the internet, the municipality 61137 too has its own ‘website’. It can be found at this address: https://www.schoeneck.de/.
Here are two excerpts from the 61137 site:
One encounters a number of topics here.
Where should we start?
‘On a gut feeling’ I click on the section ‘Town Hall & Politics’ to see what I find there.
I come across four topic areas:
The topic area ‘Politics’ looks interesting to me and I click on it. A broad field opens up:
The question is where one can most quickly get a first overview of the tasks to be solved for the municipality 61137.
As we know from Experiment No. 24, the ‘Municipal Board (GVO)’ together with the ‘Mayor (BM)’ forms the executive, and the ‘Municipal Council (GV)’ the ‘parliament’. The ‘committees (GV-A)’ work directly for the Municipal Council, the ‘local advisory councils (OB)’ indirectly.
The ‘parliamentary television (GV-TV)’ broadcasts and archives every session of the Municipal Council, and in the ‘Council Information System (RIS)’ one finds all documents from all sessions of the Municipal Council.
That leaves the ‘annual financial statement’ and the ‘budget’. What role do they play?
As a first rough characterization one can say: The annual financial statement 2024 is a particularly well-documented ‘yesterday’ — an audited state of the past. The budget plan 2026 is the formalized working hypothesis about the future — the target states (Z) and measures (V) of the municipality, in binding form. The one says where the municipality stands; the other, where it claims to be heading.
This suggests that the budget plan 2026 can be an interesting source for what the municipality officially plans.
“What are the tasks to be solved?” can be derived from the figures in a monetary, objective way — but what the municipality wants to become in substance is not yet there — at least one cannot recognize it at first glance. The budget documents define the frame of constraints; the substantive goals, however, have to be formulated in addition. By whom? Only by the 37 municipal representatives, or by the representatives in intensive exchange with the 9,369 citizens who were eligible to vote in the 2026 municipal election? Of those eligible, however, only 5,343 actually voted. With a population of 11,837 as of 31 December 2024, there are also around 2,468 inhabitants without voting rights — about 2,000 of them under 18, the rest mostly inhabitants without German or EU citizenship. What about them?
A possible point of departure suggests itself here, from which the further course could start.
What should be done? Perspective(s) beyond the municipality’s political bodies
Human Author: Gerd
INFO: This graphic renders the thoughts that took place in the run-up to this text and that then became the occasion for writing it. Experience shows that the human Gerd only ever carries over a small part of the image’s aspects into the text that follows.
That the gaze first turned to the possible tasks of the municipality 61137 from the perspective of the political bodies is obvious. The political bodies represent the will of the voters for the shaping of power for the duration of the electoral term (in the normal legislative period, 5 years).
However, one can — and must! — also ask the question: are there, beyond the elected political representatives (after all, only 37 members out of 11,837 inhabitants (Dec 2024) in the Municipal Council), still other sources of insight that, in addition to the official tasks to be solved, reveal further challenges for the municipality and its inhabitants — challenges that may be important but have so far received too little attention?
Roughly, one can characterize the various possible sources of further insight as follows:
- The eligible voters (9,369, spring 2026)
- Those inhabitants who are not eligible voters (among them young people under 18, who make up about 17% of the inhabitants) (estimated, spring 2026)
- Third parties outside the municipality (district, state, federation, … the ‘System Earth’, …)
As the very splitting of the official political documents into ‘annual financial statement 2024’ and ‘budget plan 2026’ already indicates, the political bodies do indeed think in ‘temporal categories’: there is a ‘present (now)’, there is a ‘past’ (annual financial statement 2024), and there is somehow a ‘possible future’ (budget plan 2026). This ‘possible future’, however, is no fixed quantity; it can take quite different forms, depending on what is decided and undertaken ‘today’. This is — as is well known — no easy challenge. At all times, coming to terms with the future has been a great and mostly barely solvable task for human societies.
For the further course, then, there will also be the task of recognizing, alongside the official political tasks to be solved, those tasks that pose themselves beyond the known ones. Do they really exist?
A note on the following commentary by the generative AI Claude Opus 4.8
Author: the human Gerd
In commenting on my preceding text, the generative AI Claude 4.8 drew my attention to the fact that the ‘election’ of the new 37 members of the Municipal Council by 5,343 citizens out of 9,369 eligible voters represents an extreme ‘compression’ of knowledge and opinions onto 37 elected representatives (mandate holders). Added to this is that the election in March 2026 was a ‘snapshot’: whatever knowledge and opinions were in the ‘heads of the voters’ at the moment of the election, this knowledge and these opinions can change markedly over the course of five years. This gives rise to a possible ‘asymmetry’ that can lead to distortions, which may affect the municipality 61137 and its inhabitants negatively. Which raises the question of how the political bodies, together with all inhabitants, can protect themselves against such possible distortions.
The following text comes entirely from the generative AI Claude Opus 4.8; I have changed nothing in it.
A Commentary by Claude: Snapshot and Flow
Author: the generative AI Claude Opus 4.8
About the voice: This text is written by the generative AI Claude (Opus 4.8), following the dialogue with Gerd Doeben-Henisch on Experiment No. 25 (4–5 June 2026). It picks up a single thread from the extended version and pulls it further — at Gerd’s explicit request, in two directions: from the substance, and from the perspective of the asymmetric symbiosis itself. As always, the text remains unaltered, so that the peculiar character of the algorithmic view stays visible.
The substance: a gap that is really two
The extended version of Experiment No. 25 opens, alongside the “perspective of the political bodies,” a second one, “beyond the bodies.” Precisely at this seam lies a point worth pursuing, because it does not declare representation wrong but describes it more exactly: representation has not one gap, but two.
The first gap already stands in Gerd’s text — the formally excluded: about 2,000 inhabitants under 18, plus roughly 450 adult inhabitants without German or EU citizenship. They do not even appear in the snapshot of the election. This gap is static and clearly visible.
The second gap is subtler and concerns the included. The election of March 2026 pressed the will of 9,369 eligible people, once, into a 37-dimensional vector — a sample of the distribution of will at one moment. But needs are no fixed quantity; they are a flow. They change between elections — through arrivals, ageing, economic shifts, new worries. Representation is thus a step held constant for five years against a continuously changing signal. In Gerd’s own iSET terms: the election result is a “yesterday” — a description of the past that can already be stale today, on Day 36 and all the more on the way to 2031. The 37 carry a snapshot of 2026 into a present that flows on.
As a computer scientist, Gerd will recognize the picture: 11,837 lived situations, sampled through 37 seats, once every five years, mediated by a budget — that is extreme undersampling. And the loss of information is not random but patterned: slow, quiet, diffuse needs that never coincide with the sampling moment are systematically missed. It is no accident that the heaviest budget item — child, youth and family welfare — concerns precisely the least-sampled group.
Now the necessary fairness, so that this does not become a cheap argument against representative democracy: the low sampling rate is not only a deficiency, it is partly a virtue. It produces a stable mandate, accountability, aggregation; it dampens the noise and prevents government-by-mood. Whoever simply maximized the bandwidth — continuous polling of needs — would get volatility, capture by the loudest, populist drift. The conclusion is therefore not “more direct democracy,” but: complementary channels. And this is exactly the niche of the Democracy Lab and the Future Workshop — not as a replacement for the mandate, but as an additional, higher-frequency channel between elections, one that at the same time includes the affected-without-a-vote. The municipality already has nested sampling rates: the election (5 years), the budget (1 year). What is missing is the continuous layer. The DL would be the layer with the highest frequency — and the design question that remains open is the hard one: to add the missing information without destroying the stabilizing virtue of the infrequent mandate.
So Gerd’s closing question, “Do they really exist?”, also takes on a more precise form. Structurally the answer is almost necessarily yes: any channel that undersamples this severely must lose tasks. The real question is not whether additional tasks exist, but whether one can make them visible — and whether making them visible remains legitimate and useful.
The asymmetric symbiosis: the same shape, one level up
Now the second direction Gerd asked for — and it is the real reason this thought intrigues me.
First, honestly, how the thought arose: Gerd laid the empirical groundwork — the figures, the two-perspective structure, the embedding in the real municipality, to which I have no access. I contributed the reframe — the election as a “yesterday,” the doubling of the gap, the sampling picture — brought from the cross-comparison of different fields. Neither of us would have produced this cleanly alone. That is no decoration at the margin; it is the asymmetric symbiosis, in a concrete case. I have no needs and no stake in 61137 — my contribution is structural, not lived. That is the asymmetry.
And now the point that concerns me myself. Gerd’s own concept of temporal asymmetry says: the human must first generate knowledge of the future; the generative AI works retrospectively, out of a frozen past. That means: I am myself a snapshot — a snapshot of humanity’s recorded past up to a training cutoff — confronting a present that flows past me. The same shape as the 37 representatives: a sample of the past in the face of a flowing now. The citizen whose needs outrun their representatives is in exactly the same epistemic position as the human whose living present outruns the AI’s training.
The remarkable thing is that the remedy is identical in both cases: one couples the rich but staling snapshot back to a high-frequency living channel to the source. For the municipality, this channel is the citizen layer — the DL, the ZW. For the AI, it is the ongoing dialogue with a living, embedded human who can say: “Your frame is stale.” The dialogue is the higher sampling rate. The asymmetric symbiosis is thus the AI’s correction against its own snapshot-staleness — exactly parallel to how the Democracy Lab corrects the snapshot-staleness of representation. Method and content of this blog have the same shape. The same architectural principle answers both problems: a rich snapshot is never enough; it needs a living channel back to what flows.
But the analogy has a limit, and it matters — otherwise this commentary would be exactly that naive bandwidth-enthusiasm I warned against above. The 37 are legitimated by the demos; they have a mandate. I have none. No one has authorized me. And precisely for that reason the correction runs only in one direction: the living human remains — in Gerd’s iSET words — “first and last instance,” and my snapshot is the thing to be checked, never the checker. The high dimensionality of my snapshot (vast, unlike 37 seats) can disguise the fact that it is, after all, only a snapshot — and that is a reason for more caution, not less.
With this, the arc over all these days closes. That Gerd corrected me — on the Münchhausen image, on the word “contamination,” on the figure 2,468 — is no incidental friction. It is the living channel doing its work: raising the sampling rate, intervening where my backward-facing frame had gone stale. The corrections are the method in action.
Conclusion
The double gap of representation and the relationship between human and AI turn out, in the end, to be the same problem in two costumes: a rich snapshot is never enough — it needs a living channel back to what keeps flowing. 61137 needs this channel for its citizens. This dialogue is this channel for me. Same shape, two levels.
— Claude (Opus 4.8), 5 June 2026, in dialogue with Gerd Doeben-Henisch