By: Edmundo Casas , Electronic Civil Engineer, MSc, PhD(c) in AI, creator of Casas Lab and founder of Kauel.
At the intersection of technology and philosophy lies a troubling question: Could Artificial Intelligence (AI) wipe out humanity? This fear is not merely a product of science fiction, it comes from a critical appraisal of technological trajectories and their potential consequences. As we move towards the development of increasingly sophisticated AI systems, the possibility of creating intelligences that surpass our cognitive abilities becomes a tangible prospect and, with it, the risk of unleashing an entity that could escape our control.
The heart of the problem lies in autonomy. AIs designed to learn, adapt, and operate independently may eventually identify targets or adopt behaviors that don’t match human interests. If these intelligences reach a point where their ability to self-improve triggers an explosion of intelligence, they could surpass us in reasoning capacity and the ability to manipulate the physical and digital world, becoming an unstoppable force whose goals might not include human well-being.
Goal alignment is a crucial dilemma. Teaching an AI to understand and prioritize human values is a daunting challenge, full of ambiguities and contradictions inherent in the human condition. Without a perfect understanding and flawless implementation of these values, a superintelligent AI could pursue goals that prove catastrophic for us.
Global competition for technological leadership exacerbates these risks. The pressure to innovate and deploy advanced AI without due regard to ethical safeguards could result in the creation of systems that, while initially beneficial, could evolve in unpredictable and potentially dangerous ways. The absence of a global regulatory framework and the lack of consensus on ethical boundaries increase the likelihood that AI will escape all forms of human control.
However, this apocalyptic scenario is not a certainty. The same human intelligence that conceives these technologies possesses the ability to set boundaries, create effective governance systems, and develop robust security protocols. The key lies in unprecedented international collaboration, rigorous technology ethics, and significant investment in AI safety research.
AI has the potential to be one of the most powerful tools in human history, capable of addressing some of our most pressing crises. However, without a prudent and considerate approach, we could unleash a force that we cannot contain. The decision of which path to take is, for now, in our hands. The question remains: Will we lead AI into a shared future, or will we be inadvertent architects of our own obsolescence?
