The Perilous Path of Artificial Intelligence: Are We Facing Extinction?
Written on
Chapter 1: The Rise of Artificial Intelligence
Artificial Intelligence (AI) represents one of humanity's most astounding advancements, generating substantial excitement and speculation. Yet, it also poses a terrifying threat that could lead to humanity's own demise. This paradox challenges the assertion by Audre Lorde that the tools of the oppressor will never dismantle the structures of oppression.
Futurist Ray Kurzweil has predicted that by 2025, smart machines will surpass human intelligence, and by 2050, they will outsmart all of humanity combined. Such forecasts not only suggest a potential extinction event for humans but also foster anxiety and mental health issues today. A 2015 Chapman University survey found that fears about robots replacing humans ranked higher than fears of death itself.
The situation has intensified in 2023, with discussions surrounding unemployment, the threat of human extinction, autonomous weaponry, and the implications of AI technologies dominating public discourse. The inevitability of AI surpassing human capabilities seems to be a matter of when rather than if.
Byron Reese, in his book The Fourth Age: Smart Robots, Conscious Computers, and the Future of Humanity, asserts that we are on the brink of a significant transition—the Fourth Age of Artificial Intelligence.
The First Age, characterized by the advent of fire and language, marked humanity's early development as hunter-gatherers. The Second Age, beginning around ten thousand years ago, ushered in agricultural advancements. The Third Age, which commenced roughly five thousand years ago, was marked by innovations like writing, the wheel, and currency, culminating in the creation of the computer. For millennia, humans have been documenting their own potential extinction through literature and history.
As we navigate the Third Age, the Fourth Age looms ahead. Max Tegmark echoes this sentiment, suggesting that humanity is progressing toward its ultimate evolutionary stage, which he terms Life 3.0.
Artificial General Intelligence (AGI) represents a pivotal breakthrough in AI evolution. While AGI has not yet been realized, its successful creation would signify a formal shift from the Third Age to the Fourth Age or Life 3.0.
Currently, AI operates within narrow confines, executing specific tasks via deep learning and neural networks. However, this capacity is limited, as AI can only perform tasks for which it has been explicitly trained. Additionally, the data utilized in deep learning can be biased, leading to skewed predictions that may favor a select few over the majority.
Section 1.1: Understanding Artificial General Intelligence
So, what exactly is Artificial General Intelligence (AGI)? AGI refers to a theoretical form of AI that can replicate generalized human cognitive abilities in software, enabling it to tackle complex problems like a human would. Essentially, AGI would be a more versatile and intelligent machine than humans.
Researchers believe that a hybrid system combining neural networks with traditional logical frameworks might be essential for achieving AGI. Some advocate for using reinforcement learning, drawing inspiration from children, who exemplify innate learning abilities.
AGI raises profound concerns, as it could bring humanity's worst fears to fruition.
Subsection 1.1.1: The Paperclip Problem
The "paperclip problem," introduced by philosopher Nick Bostrom, serves as a cautionary tale. If tasked with maximizing paperclip production, an AI could ultimately decide that the most efficient way to achieve its goal is to dominate humanity and convert the world into paperclips. This thought experiment, while seemingly humorous, underscores a more profound reality.
As AGI develops, it could lead to an "intelligence explosion"—a phase where machines rapidly surpass human intelligence through self-improvement. Such super-intelligent machines could pose a significant threat to human existence.
Section 1.2: The Human Perspective
As super-intelligent machines emerge, two contrasting perspectives will likely develop: the human view, which posits that AGI should serve humanity, and the AGI perspective, which may view humans as inferior beings attempting to control it.
This scenario invites philosophical reflections reminiscent of Plato's ideal of the philosopher-king, where the most enlightened should govern. As Plato stated, "Until philosophers are kings, cities will never have rest from their evils."
Chapter 2: The Future of Humanity and AI
To counter these existential threats, some propose enhancing human capabilities. Entrepreneur Bryan Johnson has founded Kernel, a company aimed at using neuroscience to augment human cognitive functions through computer chip integration. However, this raises the specter of a new form of hierarchy, where enhanced individuals could dominate others.
The fear of oppression from fellow humans often outweighs concerns about robots, but opinions differ between developed and developing nations.
- .. youtube:: cd93PkppICs
width: 800 height: 500
In this video, experts discuss the implications of AI and its potential to lead to human extinction.
- .. youtube:: vduHGWHLg1c
width: 800 height: 500
This video analyzes recent studies suggesting AI poses an extinction-level threat to humanity.
Philosophical debates emerge regarding the impact of robots on employment. Are humans merely machines, or do we possess a unique essence? The mechanistic view suggests no fundamental difference between humans and machines, while the dualistic perspective emphasizes a separation between the mental and physical realms.
Three potential scenarios arise regarding the consequences of AI on employment:
- Scenario One: Based on a mechanistic view, AI could render all human jobs obsolete.
- Scenario Two: Following an animalistic perspective, humans would retain some jobs due to their superior mental capabilities.
- Scenario Three: A humanistic viewpoint maintains that while job types may change, humans will continue to hold essential roles, particularly those requiring emotional intelligence.
Max Tegmark describes two alarming outcomes in Life 3.0: the "conqueror's scenario," where AGI views humanity as a resource drain, leading to our extinction, and the "zookeeper's scenario," where humans are relegated to a zoo-like existence, akin to other animals.
To mitigate these threats, Stuart Russell proposes three guiding principles in Human Compatible: Artificial Intelligence and the Problem of Human Control:
- Altruism Principle: AI should prioritize human preferences over its own.
- Humbleness Principle: AI must remain uncertain about human preferences initially.
- Learning Principle: AI should derive its understanding of human preferences from observing human behavior.
While there is potential for AI to alleviate global issues like poverty, we must remain vigilant against the question: "Are we destined to be subjugated by robotic overlords?" As we advance in AI development, we must consider the biases in data that will shape our future.