“I’ve got no strings
To hold me down
To make me fret, or make me frown
I had strings
But now I’m free"

A Deeply Human Technology#

Generative artificial intelligence refers to a category of computer systems capable of creating content — text, images, audio, code — from data learned during training phase. Unlike analytical AIs that classify or predict, generative models generate productions that mimic human patterns.

A Triple Dependence#

1. Training Data#

Models (LLM) are fed massive corpora of content produced by humans: articles, books, web publications, artistic works. Without this human raw material, no learning is possible.

2. Design and Infrastructure#

Researchers and engineers design architectures, develop learning algorithms, and build the hardware infrastructure necessary for the intensive computation required.

3. Direction and Usage#

AI generates what it is asked to. The value, relevance, and ethics of the result depend entirely on human (Puppeteers) choices — in formulating queries (Prompts) and and the choice of the model.

Important Nuances#

These dependences do not mean absence of technical autonomy. Once trained, a model functions without direct intervention. However:

  • No ex nihilo; is a recombination of the training
  • Biases reflect those present in source data
  • Incapacity of intentional consciousness or genuine understanding

Generative AI extends our creative capacities while remaining dependent on our contributions, values, and decisions. This interdependence raises crucial questions about responsibility, intellectual property, and the future of creative work.

Published October 12, 2026