By Leeann Sanon | Staff Writer
It seems as though every day there’s breaking news claiming Artificial Intelligence (AI) is bringing the world towards a dystopian reality. With constant breaking news about AI abuse, it is understandable why there is so much skepticism and anxiety surrounding AI and its adoption.
Anthropic and Amazon are one of these AI companies which practice destructive scanning. Destructive scanning is when a book is dismembered so that the pages can be quickly scanned. This form of scanning allowed Anthropic to avoid gaining author permission for the use of their books for AI training.
AI companies are avoiding copyright claims which lead them to obtain physical books, destroy them, and then regurgitate the information from said book. By avoiding copyright, AI cannot produce entire text from books rather summaries or excerpts. Avoiding copyright liability is nothing but a tactic for AI companies to avoid regulation.
AI avoiding accountability today is a precursor to what is to come; doom and despair. The discretion that AI companies have today is unacceptable. Allowing these companies to get away with this destruction is reaffirming their position.
It seems as though books are important for now, but not later. Clearly, these AI companies understand the value of original content. Without human made content, how would they be able to train their AI models?
The “replication” of physical books into digital feeder information for AI fundamentally replaces them as primary sources. Limiting access to the original material creates a dependency on digital and secondary sources, changing the distribution of knowledge.
The overeliance on digital forums is quite unsettling when you incorporate the destruction of books. Digital information can be altered and manipulated. The digital is never truly authentic.
AI as a primary source is unreliable, so why are they destroying books? AI models can learn without books being destroyed and information being distorted. This process primarily benefits AI as it allows AI to learn easier. Not only this, but AI is not verifiable. I would argue that’s what these conglomerates are trying to change, who can verify information.
The basis of AI is manipulation. Manipulating the information they have with the information you (the user) provide or suggest.
DISCLAIMER:
The opinions presented on this page do not represent the views of the PantherNOW Editorial Board. These views are separate from editorials and reflect the perspectives of contributing writers and/or university community members.
About Post Author
