Benevolent Dictator

Concept

In open-source culture, “Benevolent Dictator For Life” was the title given to figures like Linus Torvalds and Guido van Rossum—people who held final authority over their projects because they built them. This project transfers that authority to a large language model and puts it in direct competition with human judgment.

Anyone can pose a question about governance or civic policy. Anyone can answer. When an answer is submitted, an LLM answers too. Both responses go up for a vote, side by side, unlabeled—nothing indicates which is human and which is machine. After the vote, the source is revealed and a dashboard tallies the results. When the machine wins, a printer spits out the exchange.

In this blind vote, people don’t choose between “human” and “AI” as abstract categories but rather choose between two answers to a question about civic policy. It might cover housing, resource allocation, public safety, or any other societal topic, and only afterward do they learn which answer type they preferred. The tally doesn’t only measure what people say they value but measures what method of solution they’d actually select when labels come off.

All human respondents are anonymous and not necessarily experts, but just opinions entered into text fields. The AI is a commercial product built by a company, trained on scraped data and optimized for engagement. Neither has an obvious claim to authority, but the piece keeps an ongoing record. By the end of the exhibition, there’s a physical artifact that captures every time the algorithm was chosen over the crowd.

Benevolent Dictator Web Application 1/4

Benevolent Dictator Web Application 2/4

Benevolent Dictator Web Application 3/4

Benevolent Dictator Web Application 4/4

References

1. Raymond, Eric S. 1999. “Homesteading the Noosphere.”
2. Pasquale, Frank. 2015. The Black Box Society. Harvard University Press.