Our world today, with its advancements in technology, has left many feeling disillusioned. While some promising technologies, such as nuclear energy and genetic engineering, have been held back due to Fear and hesitation, computers and artificial intelligence (AI) have flourished without much restraint. As an AI researcher from the 1980s and ’90s, I take pride in the progress made in the field.
There are concerns, however, that AI could pose a threat to humanity. Some call for strong regulation, fearing the misuse of AI by individuals or a loss of control over AI systems. A petition even demanded a moratorium on certain types of AI research. These fears, often referred to as “AI-doomers,” suggest that AI could surpass human intelligence and potentially harm or kill us all.
But are these fears founded? The argument put forth by AI-doomers is based on the possibility of a sudden breakthrough in AI combined with a malevolent inclination. Yet, historical trends and current capabilities of AI do not align with this doomsday scenario. Progress in automation and software has been made incrementally, allowing for predictable growth rates. AI is still decades away from replacing humans in most tasks, and the idea that AI would overpower all of humanity and other AI systems simultaneously is far-fetched.
Moreover, current AI systems are not inherently inclined to harm humans. Most AI operates as specific tools or agents designed to perform certain tasks. General AI agents, which possess their own desires, undergo rigorous testing and monitoring. The concept of AI suddenly turning against humanity contradicts the detailed understanding we have of these systems.
The AI-doomer argument ultimately comes down to the logical possibility rather than concrete evidence. It is a fear based on the unknown and the potential for unexpected outcomes. To address these concerns, some suggest that AI should only be allowed if it is completely “aligned” and controlled at all times. However, until we find a way to achieve this alignment, progress in AI should not be stifled.
As we discuss the future of AI, it is essential to recognize that future human-level AIs are not alien entities or competitors, but rather our descendants. Just as our instincts favor our own offspring, so should we favor and nurture the advancements made by AI. Evolution has shaped us to value and protect future generations, even when they differ greatly from our own generation.
In the grand scheme of things, AI represents the evolution of our culture and behavior. Cultural evolution now dominates over DNA evolution in humans, and AI is inheriting our tendencies, values, and characteristics. As we create new and better AI systems, these traits will be passed on to future generations of AI. In essence, AI will carry our genetic legacy as our descendants.
To sum it up, fears of AI bringing about our demise stem from a misinterpretation of AI as “other.” In reality, AI is an extension of ourselves, carrying forward our values, behaviors, and aspirations. It is crucial to embrace this future and not let fear hinder progress. Rather than viewing AI as a threat, we should see it as an opportunity to shape our legacy in the world of technology and innovation.
The post The Future of AI: Debunking Doom and Embracing Our Descendants appeared first on TS2 SPACE.
