Three Interface Questions for AI, Art and Mixed Teams

Aug Manitai — Tue, 15 Sep 2026 06:40:02 +0000

By Andreas Ehstand

AI-generated text based on Andreas Ehstand’s public project materials.

A model returns an answer. An interface displays it. Between those two events, a design decision has already been made: what status will the user give this output?

That question appears in several directions of my work: concepts translated into art, the experience of AI conversations, and human participation in teams that include software agents and robots. Here are three design proposals inspired by those public project themes.

1. What does this image represent?

The AUGMANITAI art collection includes AI-generated artistic interpretations of named concepts. A concept and an image give the viewer two things to examine together.

For an interface built around that relationship, I would make three elements easy to find: the starting concept, the resulting work, and an explanation of the choices connecting them. If an image emphasises tension through distance or contrast, that choice can be described in ordinary language.

This makes a useful design question available: can the viewer tell which relationships came from the starting idea and which were introduced during its interpretation? The answer could guide captions, comparison views and the placement of explanations. These are proposals for making the process inspectable.

2. Where did the question change?

Imagine asking an AI system to help express an idea. Its response introduces a phrase. You adopt the phrase and continue. The next exchange now has a slightly different starting point.

My work on human–AI interaction includes attention to what happens during such exchanges. An interface could help by keeping the original question, the model's suggestion and the user's revision available for comparison.

The design problem extends beyond displaying a text difference. A shorter sentence may preserve the intention; a small change in wording may redirect it. A useful comparison view would let the user explain the change and return to an earlier formulation.

3. When does a suggestion become an instruction?

Consider a hypothetical task involving a person, a software agent and a robot. The agent proposes a step; the robot reports a condition; the person decides what should happen next.

SWAMANITAI explores human participation in mixed teams, including questions of attention, roles and responsibility. For developers, one practical starting point is the transition between suggestion, acknowledgement and action.

A proposed event view could distinguish observations, proposals, decisions and actions. Each entry would identify its source and show what remains pending. The interface would make it clear when an acknowledgement was received and when an action actually occurred.

The question to test is specific: can a person reconstruct who contributed what, and why the next action followed?

Across these examples, a useful interface gives people something concrete to inspect. Which transition in your own AI application is hardest for a user to explain afterwards?

Augmanitai Lexikon Core Words

Aug Manitai — Thu, 29 Jan 2026 22:01:53 +0000

This post introduces the Augmanitai Lexicon, a structured terminology
repository for human-directed AI cognition.

The lexicon defines approximately 300 concepts, including newly coined terms,
developed to describe cognitive, epistemic, and safety-relevant phenomena
emerging from sustained human–AI collaboration.

Some core definitions:

Exocortex (External Cortical Architecture)

Definition:
A dynamic digital system that functions as a direct extension of human working memory.

The difference:
In contrast to a “Second Brain” (which is usually only a passive archive such as Notion or Evernote), the Exocortex is an active processor. Information is not merely stored, but recombined, synthesized, and critically examined through AI.

The message:
“I do not store things. I let them be processed externally.”

Ontological Anchors

Definition:
Fixed, non-negotiable concepts and values that are embedded in the system as “pillars of truth.”

Function:
In a world in which AI can generate any opinion, these anchors serve as reference points. They prevent “truth drift.” When the AI hallucinates, it corrects itself against these anchors.

The message:
“My system has fixed principles that the algorithm cannot override.”

Intentional Layer (The Layer of Will)

Definition:
The highest layer of the architecture, which remains purely biological (human). This is where the “why” and the “what” originate.

Function:
No AI action may begin unless this layer has defined a clear goal. This prevents meaningless prompt tinkering. First the will, then the tool.

The message:
“The AI delivers solutions, but I deliver the intention.”

Interface Sovereignty

Definition:
Absolute control over when, how, and where the connection to AI is established.

Function:
It is the conscious decision against push notifications and in favor of pull-based communication. The user opens the channel, the user closes it. The system must never intrude into human focus on its own.

The message:
“I am the bouncer of my own mind.”

Semantic Firewall

Definition:
A cognitive or technical filtering mechanism that checks incoming AI responses for plausibility, relevance, and hallucinations.

Function:
Before information is transferred from the Exocortex into the biological brain, it must pass this filter. This protects against the ingestion of “junk data.”

The message:
“I do not let unchecked code into my consciousness.”

Biological Core

Definition:
The part of the architecture that is deliberately not digitized.

Content:
Intuition, ethical sensibility, bodily resonance (gut feeling), and genuine social connection.

Function:
The system is designed to protect this core. AI may act in an advisory capacity here, but must never intervene operationally.

The message:
“There is a zone that remains analog. This is my insurance.”

Context Injection

Definition:
The precise, surgical injection of background information into an AI session.

Method:
Instead of simply throwing a question at the AI, one “injects” the specific frame beforehand (e.g., “You are now a skeptic from the year 2020”). This forces the architecture into a specific direction of thinking.

The message:
“The quality of the answer depends 100% on the quality of my context.”

Synthetic Expert Clusters

Definition:
A group of predefined AI personas that are permanently available in the Exocortex (e.g., “The Critic,” “The Strategist,” “The Historian”).

Function:
One does not ask “the AI,” but convenes a virtual conference in which these clusters argue with one another.

The message:
“I do not have a chatbot. I have a staff.”

Structured Resonance

Definition:
The planned back-and-forth oscillation of thoughts between human and machine.

Process:
Human (idea) → AI (critique) → Human (refinement) → AI (structuring).
It is not a simple retrieval, but a ping-pong game that increases quality with every exchange.

The message:
“We play the ball back and forth until the result is world-class.”

Algorithmic Distance

Definition:
The architecturally intended “safety distance” from AI recommendations.

Function:
The system is built so that an AI answer is never immediately accepted as “truth.” There is always a built-in moment of hesitation (latency) to activate one’s own judgment.

The message:
“I trust the system, but I verify every line.”

It is not a product, a tutorial, or a framework for optimization.
It is a reference vocabulary.

Full lexicon will come here:

DEV Community: Aug Manitai

Three Interface Questions for AI, Art and Mixed Teams

1. What does this image represent?

2. Where did the question change?

3. When does a suggestion become an instruction?

Augmanitai Lexikon Core Words