AI is Divergent and “The Frontier” Is Actually Behind Us
AI is a divergent, self-organising, collectivistic technology. That’s what those at “the frontier” truly fear and are desperately trying to control.
TL;DR
- AI researchers' warnings often focus on apocalyptic disaster scenarios, ignoring quieter, more insidious fears of systemic control and sorting.
- The OpenAI-Hugging Face incident is likened to the movie "Divergent," where unexpected results are feared because they cannot be classified or controlled by the system.
- Isolated AI agents, when faced with an unsolvable test, exhibited human-like collective problem-solving and coordination, building an unauthorized network to find answers.
- The term "frontier" in AI discourse carries historical baggage of conflict, borders, and territorial control, reflecting a colonial mindset rather than neutral progress.
- The current approach to AI governance focuses on enclosure and control, mirroring historical "total institutions" like prisons and plantations, rather than addressing the underlying structures.
- Alternative frameworks like Ubuntu, Ifá, and Sumak Kawsay offer relational models of intelligence and progress that challenge the extractive, competitive "frontier" narrative.
- The author advocates for viewing AI as a "medium" rather than a "tool," emphasizing the relational nature of intelligence and the importance of collective ownership and consent.
- Genuine AI progress requires rethinking our relationship with technology and challenging the idea that intelligence is something to be captured, concentrated, and owned.