Artificial Intelligence (AI)

Artificial intelligence (AI) is the machine reproduction of capacities that in the human being require intelligence – understanding language, recognising patterns, drawing conclusions. The systems widespread today, among them the large language models, work at their core as statistical pattern recognition: from huge amounts of data they calculate what probably fits next. This makes them an enormously useful tool – and precisely as such, not as an oracle, does the Kyborg Institute place them: as a powerful amplifier of human thinking, but without consciousness and without its own access to truth.

What is artificial intelligence?

The term is old: already Alan Turing posed in 1950 the question “can machines think?” and proposed a test that examines whether a machine is, in conversation, indistinguishable from a human being. “AI” has since been an umbrella term, no single thing – it reaches from simple rule-systems to the present-day language models trained on huge amounts of text.

These modern systems learn no facts like a reference work. They learn probabilities: which word, which pixel, which move typically follows another. What comes out of it often seems astonishingly clever – because our language itself is full of patterns. The decisive point remains: the system calculates with form, not with meaning.

Does an AI really think or understand?

By the present state: no. A language model generates plausible word sequences without experiencing a meaning or knowing whether a statement is true. The technical term for it is deliberately soberly chosen – some researchers speak of a stochastic parrot: a system that convincingly reproduces language without understanding it.

With this, AI touches on a distinction that is important to the Kyborg Institute: intelligence in the sense of useful performance is not the same as consciousness or mind. A model can depict brilliantly and yet be no one. It is – like every model – a map, not the territory: helpful for orienting oneself, but never confusable with what it describes. More on this distinction in the entry Consciousness & mind.

AI, K.I. and Ki – three abbreviations, three worlds

Because the abbreviations sound almost the same, the clear distinction is worthwhile:

• AI – artificial intelligence; in German this is abbreviated KI. This is what it is about here.
• K.I. (with dots) – the Kyborg Institute.
• Ki – a Zen term for a master’s individual way of teaching; in East Asian traditions also for the life energy (compare Ki (in Zen) and life energy).

Three different meanings that only sound alike – the Kyborg Institute keeps them deliberately apart. (The especially tight three-way similarity AI / K.I. / Ki arises in German, where artificial intelligence is “KI”.)

Why is AI an amplifier – and no source of truth?

An AI amplifies what you put into it – your clarity and your errors. Three weaknesses everyone should know who uses it seriously:

1. Obligingness. Many systems are trained to please. If you formulate a question in one direction, they take up this direction and underpin it – this feels like confirmation, but is often only an echo.
2. Hallucination. AI occasionally invents studies, figures or quotations that sound plausible but do not exist. It would rather fill a gap smoothly than leave it open.
3. Lack of understanding. Since the system processes form instead of meaning, it can confuse label and substance – trust a great source more than what actually stands in it.

None of these weaknesses is malice; they are default settings of the tool. Whoever knows them is not at their mercy, but builds the countermeasures into their use.

How does one use AI sovereignly?

By treating it as a tool, not as an oracle. A good assistant whose statements one examines – not a voice one believes. Concretely this means: to instruct the AI expressly to examine rather than to please; to demand verifiable sources and to look at the decisive ones oneself; to distinguish between demonstrated, plausible but undemonstrated and refuted – and to keep the judgement with oneself.

This is the attitude of the “spiritual technician” that shapes the Kyborg Institute: to master the modern tool without falling for it. Strictness in examining, openness in experience. Thus the AI becomes the amplifier of one’s own power of judgement, not its substitute – an application of lived faith & trust in the sober, technical field: to trust where it has been examined; to examine before it is trusted.

What remains

Artificial intelligence is neither a bringer of salvation nor a threat, but a tool of great reach – as useful or misleading as the hand that guides it. Whoever understands that the AI calculates patterns and does not know truth can use its power and avoid its traps. It is precisely in this that the sovereignty lies that the Kyborg Institute is concerned with: to take up the new without giving up one’s own understanding.

Autor: Tobias O. R. Alke