Is AI becoming feral?
Reported by TechRadar1 source ↓
TechRadar argues that “feral” can describe technology that was designed to be controlled but develops unpredictable effects, and proposes the term as a way to discuss AI’s risks without assuming catastrophe is inevitable.
Key points
- The article defines “feral” technology as something originally controlled that slips out of control
- It says attempts to control an uncertain system can sometimes intensify its dangers
- The article argues that AI risks should be assessed without dismissing warnings or accepting apocalyptic claims
- It uses printing and the telegraph as examples of technologies that reshaped systems beyond their creators’ plans
Details
The article proposes “feral” as a term for technology that begins under human control but produces consequences its creators cannot predict or contain. It says this does not mean catastrophe or permanent disorder is inevitable: early recognition could allow a trajectory to be mitigated or changed, though stabilization is not automatic.
To illustrate the idea, the article describes printing as initially serving book production before pamphlets helped drive religious and political upheaval, and the telegraph as supporting international finance and cycles of speculation and collapse. It argues that AI’s development raises similar questions about whether existing regulatory systems can manage its effects.
The article also says competition between countries can be used to argue against slowing AI development, while companies’ warnings about catastrophic risks may draw public attention to their products. Citing Georgetown computer science professor Cal Newport’s phrase “doom trolling,” it calls for institutions that can distinguish commercial alarm from genuine danger.
Written from coverage as of
Images
- TechRadar