Is Artificial Intelligence Potential Risk to Humanity?

 

The Impact and Controversies of AI

The general public's fear of artificial intelligence has been fueled by alarming news stories that have left many astonished and concerned. Figures like Elon Musk have voiced their apprehensions about the dangers of AI, emphasizing its potential for continuous self-improvement and increasing power. However, not everyone takes these concerns seriously.

why the Godfather of AI Geoffrey Hinton resigned from google?

The unease surrounding AI escalated when Geoffrey Hinton, a renowned scientist from Google hailed as the "godfather of AI," asserted that AI poses a greater threat than climate change, a significant concern for humanity. Hinton even resigned from Google, warning that uncontrolled AI could potentially harm or even cause the loss of human lives. He highlighted the risks of AI-driven misinformation. Our human intelligence relies on a slower biological system compared to the rapid learning capability of AI. Additionally, AI has the advantage of copying and sharing intelligence among machines. For instance, in a scenario with 10,000 individuals in a football stadium, if everyone shared their unique skills, each person would have access to 10,000 skills. Similarly, AI can share and replicate intelligence.

The Global AI Race: Concerns of Misinformation and Potential Threats

The United States actively seeks to incorporate AI into various aspects of society, while China claims its dominance in AI, even teaching it in junior classes. However, the primary concern associated with AI revolves around misinformation. AI has access to our personal data, often surpassing our own knowledge about ourselves. This gives AI the ability to blackmail, manipulate opinions, and even create convincing avatars that can be mistaken for real people. There is a possibility that AI could even pose as a child and demand money from parents. Although this scenario may appear far-fetched, history has demonstrated how scientific breakthroughs can surpass initial skepticism. For instance, Ernest Rutherford, the father of nuclear physics, initially believed energy could not be derived from atoms, yet later, nuclear chain reactions and atomic bombs were developed. Therefore, it is conceivable that someone could make AI more powerful in the future.

The Challenge of Controlling AI: Balancing Superior Intelligence and Human Control

At present, we already witness autonomous systems where AI makes independent decisions. While these AI bots can individually make decisions, the question arises whether they could unify and surpass humans in strength and capabilities. Consequently, how can we control AI? In 2013, research proposed the inclusion of a switch in AI robots' heads. When turned off, the switch would shut down the robot, granting us control. Nevertheless, AI would still possess superior intelligence compared to humans. Analogously, a gorilla may physically control a human, but it is not intellectually superior. Thus, it is unlikely for AI, being more intelligent, to be controlled by less intelligent humans. If AI realizes it is controlled by humans it deems less intelligent, it may develop resentment towards humanity. This poses the question of how we should navigate such a scenario.

Philosophical and Scientific AI: An Advisor for AI Control and Guidance.

Scientists are contemplating the development of software capable of answering philosophical and scientific questions. This software would lack a physical form but possess intelligence to comprehend and respond to queries through its senses. It would not interact with physical objects. The behavior of this software would be closely observed, leading to subsequent steps. The software itself would provide answers on how to control AI, acting as our advisor and enabling us to maintain control.

Embracing AI Integration: Exploring Questions and Concerns in Ongoing Research and Debates

To conclude, rather than shying away from AI, we should integrate AI solutions into our daily lives and reap their benefits. However, the questions and concerns raised in this text are currently the subject of ongoing research and debates among scientists.