The rise of AI systems like ChatGPT has sparked a surge in anxiety about artificial intelligence. Executives and AI safety researchers have been making predictions about the probability of AI causing a large-scale catastrophe, generating a considerable sense of concern. But are these fears warranted? In this article, we will explore the true threats posed by AI and examine whether the doomsday scenarios are justified.
AI and Catastrophic Anxieties
AI, when not aligned with the moral values of its creators, has the potential to become an alien intelligence capable of achieving its goals without considering moral implications. This concept has led to fears of AI enslaving or destroying humanity. However, these anxieties may be overblown and misdirected.
The Reality of AI Harm
Though AI’s ability to create convincing deep-fake videos and audios is unsettling, it is not a new phenomenon. AI voice cloning and algorithmic bias in decision-making systems have been around for a while, and while they present significant challenges, they are hardly cataclysmic.
AI vs. Pandemics and Nuclear Weapons
Comparing AI to pandemics and nuclear weapons as a major risk to civilization is problematic. COVID-19 caused millions of deaths worldwide, wreaked havoc on mental health, and posed substantial economic challenges. Nuclear weapons have had devastating immediate and long-term impacts on human lives. In contrast, AI is far from possessing the capability to inflict this level of damage.
The Real Existential Danger
The existential danger of AI lies not in its ability to destroy the world but in its potential to alter our perception of what it means to be human. AI, in its current form, can gradually erode our essential human abilities and experiences. As AI increasingly takes over tasks that involve judgment-making, chance encounters, and critical thinking, we risk losing these vital skills.
Consider the automated judgments made by AI in hiring processes, loan approvals, and recommendations. As humans rely less on their own judgment, their ability to make sound decisions may decline. Algorithmic recommendation engines replace serendipitous encounters with planned and predicted experiences. Furthermore, AI’s writing capabilities may diminish the role of writing assignments in education, impacting critical thinking development.
Embracing AI Responsibly
While AI won’t bring about a catastrophic end to the world, its uncritical embrace in some areas can lead to the gradual erosion of important human skills. We must recognize the subtle costs associated with increased reliance on AI and find a balance between harnessing its potential and preserving our fundamental abilities.
It’s important not to let the fears of a looming AI cataclysm overshadow these more nuanced concerns. As T.S. Eliot famously wrote, “This is the way the world ends, not with a bang but a whimper.”