After the Privatization of the Internet, Silicon Valley Begins Privatizing Human Civilization
"The Privatization of Human Civilization"
The article critiques how AI companies like Anthropic are systematically acquiring and digitizing millions of books—sometimes by destroying physical copies—to build proprietary training datasets for models like Claude. While a lawsuit resulted in a settlement, the author argues the deeper issue transcends copyright: it is about the privatization and centralized control of human knowledge and civilization.
This process coincides with a powerful Silicon Valley ideology, exemplified by Marc Andreessen's "Techno-Optimist Manifesto" and movements like e/acc (Effective Accelerationism). This worldview frames technological growth and speed as inherently moral, portraying caution, regulation, and public dissent as obstacles to progress. It often envisions intelligence itself, rather than human well-being, as the ultimate goal, potentially sidelining present human concerns.
Figures like Peter Thiel and Curtis Yarvin express skepticism towards democratic processes as too slow, suggesting more centralized, founder-led governance is efficient. This logic extends to AI, where a small team within a company defines the model's "constitution"—its rules, values, and definitions of truth and safety—effectively governing how millions understand the world.
Thus, the scanned books symbolize a new form of control. Knowledge isn't erased but is ingested into private, opaque systems. The original, decentralized, and contestable nature of books and public knowledge is replaced by a curated, company-controlled output. The public's access to their own cultural heritage becomes mediated by corporate AI, which remembers civilization only in the form its creators dictate. This is not book-burning but a subtler, potentially more complete privatization of human memory and understanding.
marsbitHá 1h
