I’M PEACHY extends the investigation of political repetition into the domain of machine cognition. Using a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM), the work is trained on tweets produced during the impeachment period—a moment marked by obsessive return, denial, and rhetorical looping. The model generates new language by continuously feeding its prior output back into itself, creating a synthetic echo of thought.
Here, recurrence is no longer metaphorical; it is operational. The machine does not critique repetition—it embodies it. Patterns reinforce themselves, contradictions are absorbed, and language cycles without progression. The generated text reflects a behavioral logic already present in the source material: a mind locked into feedback loops where memory does not produce learning.
Where Pliant Desire fractures and rearranges rhetoric to reveal its pliancy, I’M PEACHY allows repetition to run unchecked—demonstrating how both political speech and algorithmic systems can become self-sealing. Together, the works trace a shift from exposure to simulation, from dismantling language to letting it endlessly repeat itself.
Pliant Desire and I’M PEACHY examine political language as a system defined by recurrence. The first breaks rhetoric apart to reveal its pliability under power; the second reconstructs that recurrence within a machine that cannot stop repeating. Across both works, language ceases to communicate and instead circulates—trapped in loops that sustain authority through repetition rather than meaning.
Note that due to limitations in processing power, it takes a while for a thought to be generated – please be patient as it loads below. It may also be long and rambling and not make much sense.
The work traces its roots to THE HARPY AND THE HARE , a (political) bedtime story for children and grown-ups, when Canada had its own fascist-leaning government and prime minister.
CREATED BY: Stephen Chen | COMPLETED: 15/02/2020
Generate your own RECURRENT thought: https://pyhook.dissonanceonline.com/impeach
