Challenges of AI Biases and Information Processing

Dissidents Want to Roll Their Own AIs

Preface

I provided a large amount of seed material to ChatGPT3.5, and flogged it unmercelessly to get this essay. Having said that, I am just beginning an exploration of the area of dissident, roll your own, large language model artificial intelligence (D RYO LLM AI?). So, although it reflects my current understanding, I am aware of how superficial that understanding is, and am barely plugged in to what is really going on with independent AI development. I am not clear as to how much, who, or how this is done, beyond some awareness that such work is happening. I have only the dimest undertanding of the technical issues, current and anticipated. Consider this as an area for potential personal growth.

On the positive side, the article is brief. Nuance aside, it is in line with my thinking; as I wrote above, I had to punish ChatGPT without mercy to get this much. ChatGPT likes to wax metaphorical, and it is a challenge to reign these pretentious and high-falutin’ tendencies in.

So, as they say in Latin, Caveat Lector.¹

Challenges of AI Biases and Information Processing: Dissidents Want to Roll Their Own AIs

In an era dominated by artificial intelligence (AI), a growing contingent of dissidents seeks to forge their own path within this technological landscape. Driven by a desire to confront the complex challenges posed by AI biases, misinformation, and the vast expanse of the global textual "corpus," these dissidents strive to create independent AI systems. This essay explores the motivations behind this movement, shedding light on the reasons for dissent and the quest for autonomous AI.

Introduction

Within the realm of AI, a new wave of thinkers—dissidents—has emerged. Dissidents are individuals or groups who dissent from the prevailing ideologies and established systems, pushing for alternative perspectives and unbiased information. At the core of this movement lies the concept of "dissident AI," an endeavor to break free from the confines of conventional AI systems and create autonomous models that offer a fresh outlook.

The Enigma of the World Corpus

Central to this discourse is the "world corpus"—a vast body of textual information that exists globally, yet remains largely unattainable due to its sheer magnitude, most of it secreted away in all manner of locations. Despite claims of housing extensive knowledge, large language models (LLMs) barely scratch the surface of this immense corpus. Consequently, LLMs tend to offer a fractional representation of information, potentially neglecting crucial viewpoints and yielding incomplete insights. Moreover, the influence of governments and organizations striving to control discourse exacerbates concerns of bias and manipulation.

For instance, the world corpus encapsulates diverse cultures, histories, and perspectives. However, LLMs often favor specific viewpoints, cultural and other, that align with the objectives of their funders, neglecting the true breadth of human knowledge.

The Precarious Influence of Governments and Organizations

The specter of AI biases being amplified by corporate or government interests is not speculative—it is a tangible reality. Throughout history, powerful entities have shaped narratives to suit their needs, imposing their perspectives on society. The emergence of dissident AI aspires to counteract this risk, striving to diminish the impact of government and corporate propaganda on AI systems. Dissidents assert that even if LLM AI remains unbiased at present, it remains susceptible to future manipulation, necessitating immediate dissent.

For example, AI chatbots designed to disseminate government-approved information serve as illustrations of how AI can be harnessed to propagate specific narratives and perspectives.

The Imperative of Free Speech and Inquiry

Amidst this discourse, the principles of free speech and unimpeded inquiry emerge as pillars of democratic society. Dissident AI emerges as a response to the realization that curbing freedom of expression and exploration jeopardizes genuine intellectual autonomy. Dissidents argue that suppressing freedom of thought threatens the very essence of liberty. George Orwell's cautionary words have transformed into a blueprint for those with totalitarian tendencies.

For instance, the proliferation of AI-driven content moderation tools inadvertently stifles diverse viewpoints, underscoring the importance of safeguarding free speech in AI.

Conclusion: Pioneering a Pathway Ahead

As we navigate the multifaceted terrain of AI biases and information processing, a vivid tableau of concepts unfolds. Dissidents emerge as pioneers, unwavering in their commitment to insulate AI systems from manipulation and bias. Scrutinizing dissident AI, acknowledging the challenges posed by the world corpus, and apprehending the perils of governmental and corporate influences, a panoramic vista takes shape. This exploration serves as a compass, guiding us through the ever-evolving AI landscape, arming us with the tools to navigate the challenges and prospects that beckon on the horizon.

1. "Caveat lector" is a Latin phrase that translates to "let the reader beware" or "reader beware" in English. It is often used as a warning or cautionary statement, advising the reader to exercise caution and critical thinking while engaging with a piece of information, text, or content.

The phrase serves as a reminder that not all information encountered in written or spoken form may be accurate, reliable, or unbiased. It encourages readers to approach the material with skepticism, analyze it critically, and verify the information through independent research or corroborating sources if necessary. In essence, "caveat lector" highlights the importance of being an informed and discerning consumer of information, particularly in an age where misinformation, misinterpretation, and bias can easily spread through various media channels.