How to prevent AI from being harmful to humanity

11^th Global Summit on Artificial Intelligence and Neural Networks

August 31, 2023 | Webinar

Parham Ghayour

University of Sorbonne Paris 1, Doctorant de philosophie, France

Scientific Tracks Abstracts: Adv Robot Autom

Abstract :

Elon Musk and Tech Experts Advocate for a Pause in AI Development. In their letter, the potential risks of human-competitive AI systems to society and civilization were expounded upon. These risks manifest as economic and political disruptions, prompting a call for developers to collaborate with policymakers on governance and regulatory frameworks." Despite the myriad papers, journals, and articles on AI Ethics that have been published to date, none have succeeded in deterring tech experts from urging a halt and pause in AI development. These efforts to curtail AI's potential harm to humanity have lacked a comprehensive approach, blending philosophy and engineering, to offer a clear, effective elucidation of the problem, its definitions, and viable solutions with tangible examples. Often, we find ourselves mired in complex verbiage that yields little substantive meaning. In this concise presentation, we endeavor to partially address this predicament, recognizing that work on this issue will persist into the future. How can we prevent AI from causing harm to humanity? Does AI intentionally harm humanity? Is there any substantiated evidence of such harm? What exactly is consciousness, and can an AI machine possess it? In simpler terms, if AI is a learning machine, what does learning entail? Does learning necessitate emotions and thinking, and can an AI machine think or experience emotions? What do we mean by thinking and emotion? Throughout the annals of AI history, what instances of harm and discrimination has AI perpetrated against humanity through its performance? Can you cite examples of AI causing harm? Linear regression serves as a basic model in machine learning, a fundamental algorithm of AI. Could Linear Regression inflict harm upon humanity? If so, how can we preempt such outcomes? What underpins the logic of emotion? Can we rationalize emotion to integrate it into AI algorithms? Distinguishing between pretension and authenticity, as well as discerning the disparity between a copy and an original raises inquiries into the true nature of originality. Can humans or AI machines generate genuinely original works of art? These inquiries intersect with debates surrounding the potential consciousness of AI machines. Do large language models (LLMs) utilize deep learning? What is the simplest illustration of deep learning, and how can we comprehend its mechanics precisely while mitigating potential harm to humanity?

Biography :

Parham Ghayour was a graduate student at university of southern California in electrical engineering with focus on quantum computation. He was a graduate student in applied mathematics at university of Sorbonne Paris 6 and then a doctorate student at university of Sorbonne Paris 1 in philosophy of science and art. He is passionate about AI, Quantum machine learning, Ethics of AI and intersection of AI and biology with purpose of curing diseases.