Taming AI
Society and Government at the Helm
This story compares the military’s adherence to governmental oversight and artificial intelligence’s compliance with strict societal and governmental regulations. Just as the government binds the military with orders, AI, too, must not operate outside the guidelines set by society. Proceeding otherwise can usher in an era of chaos and turbulence with unpredictable outcomes.
Artificial Intelligence (AI) technology is to society what the military is to a democratically elected government. The government directs the military, which is subservient to it, serving without attempting to control or usurp it. It waits for government decisions to implement them and does not dictate actions to the government.
Similarly, society, a much larger entity than the technology sector, must oversee technology; technologists cannot independently determine society’s course. Society comprises families, local, regional, and national institutions, bodies, companies, and organizations. It has a specific multilayered habitus, zeitgeist, history, and traditions, with technology being just one component.
Like the military, which receives orders from the government, technology should also be a tool to assist society and create new possibilities, leaving strategic decisions to society and its ruling elites. This is because ruling elites in democracies are elected by the people and tasked with governing, conflict resolution, economic planning, social engineering, education, and more.
A legitimate representative government consists of a hierarchy of elected or appointed officials covering diverse fields, including trade, economic planning, defense, health, international relations, etc. The military, tasked with defending the territory against foreign adversaries, should not meddle in government activities but execute orders from elected officials.
Likewise, society encompasses more than technological firms, including the government, opposition parties, civil society organizations, NGOs, think tanks, educational establishments, researchers, business institutions, unions, markets, independent thinkers, artists, philosophers, other communities, minorities, families, and private citizens.
AI technology companies, despite their power, capital, specialized workforce, and sophisticated technology, do not represent society and cannot force it to implement strategic choices, such as designing super-intelligent machines.
Should these machines reach the singularity point, their decisions might conflict with societal values, norms, and expectations, possibly making decisions beyond the control of AI engineers. Even so, AI engineers or their creations are not entitled to make critical decisions for society without consulting the government or society, as they have not been elected or authorized to do so.
The consequences of AI technology acting independently are uncertain. Such companies’ potential to coerce government or society is significant, given their budgets, coverage, and ability to create entities surpassing human intelligence, rendering humans obsolete, and operating based on decisions by a select group of techno lords who claim ultimate authority.
It has been a little over a year since the Future of Life Institute issued the call to pause giant AI experiments beyond ChatGPT 4. As I explained in a previous article, there is a very slim chance that such a call may be heeded. This is because of what Eliezer Yudkowsky and David Shapiro call “the terminal race condition”, a situation where the competition between AI firms is so heated that supersmart systems are developed without consideration of safety, ethics, and governance factors.
In the end, we are standing on the cliff of a digital summit after which we may no longer be in a position to control the movement of the human race driven by an entity expectedly far superior in intellect. Because of this, the need to put in place regulations that govern, oversee, supervise, and approve or disapprove the course AI, AGI, and ASI are expected to walk. The European AI Act shows EU member states determination to restrain or curb what AI can and cannot do.
Finally, allow me to clarify, as I do every time I write on this topic, that I am not a Luddite, a technology hater, or even a digital minimalist. Like many, I enjoy the benefits of technology. However, given AI technology’s uncertain trajectory, I, as a human being, do not wish my life, future, and fate to be determined by superintelligent systems created by a handful of humans.
The evolution of human history tells us that tribes, nations, and states have always been the seat of power. In them, we trust, and under their aegis, we hope AI will fall, now and ever.