Before We Build AI Gods, We Need AI Rules
THE FXBG ADVANCE THURSDAY 8/6/26 MIDDAY READ
By Phil Huber, ADVANCE COLUMNIST
If you can envision it, you can build it. Given enough money and time, nearly every powerful invention can also be weaponized. Artificial intelligence is no exception. It is already moving through our networks, our skies, outer space, and even into the human nervous system. It can watch, predict, persuade, and act at speeds no human can match.
The ancient Greeks imagined gods who could see everything, hurl lightning, and bend human fate. For thousands of years, that remained mythology. Today, AI is bringing parts of that myth disturbingly close to reality. The real question is no longer whether these systems can be built. It is who will control them, who will be protected from them, and whether democracy can survive if too much of that power gathers in too few hands.
From Olympus To Orbital AI
On the military side, AI-enabled weapons can coordinate drone swarms, identify targets, optimize attacks, and run cyber operations at machine speed. The most dangerous versions could include autonomous weapons and AI-assisted chemical, biological, radiological, or nuclear systems. The risk is not only deliberate misuse. It is accidental escalation. Software can act on incomplete, biased, or misunderstood information before human judgment has time to intervene. A conflict that once unfolded over hours or days could accelerate in seconds.
AI is also transforming surveillance. Systems that combine location data, communications, financial transactions, purchases, and social media can reconstruct a person’s life in extraordinary detail. They can identify dissent, predict behavior, and tailor propaganda to specific fears and vulnerabilities. Deepfakes and microtargeted messages make it possible to manipulate not one audience, but millions of individuals at once. The Greeks imagined gods who could see and judge every human action. Modern AI offers governments and corporations something uncomfortably similar, but with far fewer moral limits.
Now add space. Satellites are no longer passive cameras or relay stations. They increasingly use onboard AI to decide what data to collect, where to point sensors, and when to transmit. In October 2025, an AI agent performed a complete attitude maneuver on a satellite in orbit. Governments and companies are also exploring orbital AI infrastructure: computing centers in space, powered by sunlight, cooled by the vacuum, and connected by laser. If we build an intelligent shell around the planet, controlled by a small number of actors, we will have created an Olympus of steel, code, and permanent observation.
Control Inside the Nervous System
It would be comforting to believe this power stops at the edge of the body. It does not. Neuralink has already implanted brain-computer interfaces in human patients. The medical promise is real. A person with paralysis may be able to move a cursor, operate a robotic arm, or communicate by thought. That could restore independence and dignity to people who have lost both.
But the architecture matters. A brain-computer interface does not end at the implant. It connects the nervous system to software, networks, and platforms. Musk has spoken of high-volume production and largely automated surgeries, with early trials potentially reaching hundreds or low thousands of patients in the late 2020s. Broader medical, and perhaps nonmedical, uses could emerge through the 2030s.
The danger is not the medical breakthrough. The danger is the system that may grow around it. If the same companies control the implants, the AI models, the cloud infrastructure, and the satellites, surveillance and influence are no longer merely over our shoulders. They begin to reach inside our heads. The technical question is whether the devices can work. The political and constitutional question is who will own them, regulate them, and stop them from becoming tools of coercion.
A Timeline We Can See
None of this is centuries away. Brain implants already exist. AI has already controlled spacecraft in orbit. Companies are openly planning space-based computing systems, with pilots and early deployments aimed at the late 2020s and early 2030s. Many AI researchers also believe there is a real chance that human-level or transformative AI could arrive between 2040 and 2060.
That means the pieces are appearing within the lifetime of people already voting, working, and raising families. Chips in brains, AI in orbit, autonomous weapons, and increasingly powerful models on the ground are not separate stories. They are parts of a larger structure. We are building systems that can see more, decide faster, and reach deeper into human life than any government, military, or corporation in history.
Why the Promise Makes the Danger Greater
AI can do enormous good. It can help doctors spot disease earlier, assist people with paralysis, tailor education to individual students, improve emergency response, manage traffic and energy use, and reduce repetitive work. Properly governed, it could expand access to knowledge, strengthen public services, and give people more time for care, creativity, and civic life.
That promise is precisely why the stakes are so high. Technologies that become indispensable are hard to challenge. Once hospitals, schools, militaries, banks, and governments depend on the same systems, the people who control those systems gain influence that may be difficult to reverse. AI does not have to become evil to become dangerous. It only has to become essential while remaining unaccountable.
Enemies Today, Partners Tomorrow
We are often told that the United States must move faster because China is moving faster. There is truth in that concern. Democracies should not hand advanced AI capabilities to authoritarian governments. But a race between great powers does not guarantee protection for ordinary people. It may instead produce a compact among the few who control the chips, models, satellites, cloud systems, and networks.
Rivals often cooperate when it serves their shared interests. Governments and corporations could eventually agree on standards for surveillance, borders, information control, and access to computing power. The benefits would move upward to those who own the infrastructure. The risks—job displacement, biased policing, misinformation, and lost privacy—would move downward to everyone else. The gods would stop fighting and begin managing their estate.
An FAA for AI
We already know how to govern complex, high-speed systems that can kill people when they fail. We do it in aviation. The Federal Aviation Administration classifies airspace, certifies aircraft and equipment, sets operating rules, tests new technologies, investigates failures, and insists on human oversight.
AI needs a similar model. An “AI FAA” would classify systems by risk. It would impose the strictest rules on weapons, policing, elections, healthcare, education, and critical infrastructure, while using lighter rules for low-risk applications. High-risk systems would require testing, documentation, certification, independent review, and clear lines of accountability before they could be deployed at scale.
The principle is simple: safety first, rights first, speed and profit second. No company should be allowed to write its own rules for systems that can shape elections, make life-and-death decisions, or monitor whole populations.
Keeping the Gods’ Chair Empty
The goal is not to create a benevolent AI ruler. It is to ensure that no person, corporation, agency, or nation can occupy that position at all. At home, that means binding rules for transparency, accountability, nondiscrimination, and human control. It means public oversight of high-risk systems and a meaningful role for states and local communities in regulating effects on labor, energy, the environment, and civil rights.
Internationally, we have taken the first step. The Council of Europe’s Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law is the world’s first binding AI treaty. It seeks to ensure that AI is developed and used in ways that respect human rights, democracy, and the rule of law—placing principles like transparency, accountability, and non‑discrimination at its core. The world will need arms-control-style agreements for autonomous weapons, surveillance AI, and cross-border platforms. If nuclear weapons required global rules, then systems capable of reshaping militaries, information, and human behavior require them too.
Here Is My Ask
Do not treat this as science fiction. Treat it as a political agenda. Ask every elected official and every candidate who wants your vote one question: What will you do to ensure AI remains accountable to democracy?
Ask your city council and school board. Ask your state legislators. Ask your members of Congress. Ask the President. Right now, policy is moving toward concentrated, unaccountable AI power, sold to us as innovation and national security.
The architecture is already rising around us—in our phones, our workplaces, our skies, in orbit, and now inside the human nervous system. We still have time to govern it, but we should not assume we always will.
The Greeks feared their gods because no one could challenge them. Democracy was founded on the opposite belief: no one should possess unlimited power. AI may become one of humanity’s greatest achievements. It may also become the greatest concentration of power we have ever built.
The question is no longer whether someone can build the gods’ chair. The question is whether we will stop anyone from sitting in it before the chair becomes a throne.
***
Phil Huber is a retired Army Reserve colonel, federal civil servant and consultant who writes on civic education. He lives in Fredericksburg.
Clear-headed assessment