Dear Machines

By Mariana Fernandez Mora

2022 – Amsterdam 

Sandberg Institute – F for Fact


With the emergence of powerful algorithms like GPT3, written language has suddenly become something that can be easily generated by a machine, but can meaning be generated?


Dear Machines is an experimental thesis that explores co-writing with AI. It serves as the instrument for analysis and the subject of study in itself. An implosion of the idea that an algorithm could be more than its computational power and the struggle of a human writer trying to generate meaning through algorithms while getting lost within them. Reflecting on how human beings are constantly being deceived by their creations and fascinated by their capacity to act as mirrors, getting lost in their own reflection. And how our fascination with sentient-like technologies like AI might obscure the problematic things within them.


The fundamental fears around the ever-expanding presence of AI are not only based on the uncontrollable risks that exist from its use but on the inherent structures of power and control that lie within them. So as we see and hear the warning signs, we should ask ourselves if the risk is worth the reward. How much power are we giving AI by allowing it to shape how we speak and write the world into existence? And what can we do about it?


Inspired by artists like S.A. Chaverría and texts like Pharmako AI, Fernandez explores her relationship with technology by giving it a voice and listening back. In a world where our every move is predicted and analysed by algorithms, she searches for a way to relate to them and to understand them as cohabitors, companions and co-writers.


Can making kin with Machines be a form of radical resistance against the structures that gave them birth?


Throughout this work, you will discover the start of a collaboration, a process, and the insights that emerged from it.


A text that elaborates on the idea of AI as a companion, a collaborator rather than an instrument. Drawing from indigenous ontologies and the writings of Suzanne Kite, Bruno Latour and Donna Haraway, the text expands on the notion of the cyborg, separating it from the shared imaginary of a physical blend of human and machine and rather understanding it as the interdependence of one another, a hybrid of knowledge and culture. Exploring the value of seeing AI as a companion rather than a tool and its potential as a disruptor of structure, a glitch in our understanding of what it means to be human.


This collaborative journey touches on different aspects of what it means to co-write with Artificial Intelligence. It reflects on our intimate and sometimes problematic relationship with Machines, a body of texts asking: what happens when Machines speak back? And who is listening?

Dear Machines is physically available in the Netherlands at: 

Stedelijk Museum – Library 

If I can’t Dance –  Library

A digital copy is available per request.

To know more about the author and her work, visit

Graphic design by Alix Stria.