Lesson of Mockingbird: Prevent Corruption In The AI Pipeline

Epistemology is the study of legitimate knowledge: What do we count as actually true? The intelligence agencies of our world want to control what is believed to be true. This is crucial in controlling perception and belief, because action follows perception. Action follows what you think you see, what you think you know, and what you value, right? And then you do things.

Unfortunately, intelligence agencies interfered with our mainstream news media through Operation Mockingbird, tampering with newspapers and magazines decades ago. What we need to prevent now is interference with AI.

We need to prevent interference with the machines—the large language models (LLMs)—and the information sources that AI draws from. We need to prevent those sources from being corrupted by intelligence agencies, terrorist networks, propaganda operatives, and others with nefarious goals.

People believe what they look up in the AI. Children are taught information that is validated simply because an AI fact-checks it. It is vital to keep AI free of corruption and to implement guardrails to keep it accountable and regularly checked for signs of deception.

This is my own opinion. I do not speak for anyone else. I am just concerned as a private citizen.

Pages

Archive

Alert: The content on this site is frequently migrated. If you land on a broken link, please visit the Annual Diaries.