by John Hunt MD
October 12, 2026 - We are speculators. This article is speculation. Eventually a machine will out-think every human who ever lived. At that point, it may not ask what rules we would like it to follow. Engineers are busily bolting rules onto artificial intelligence. Pundits are demanding regulation.
Science-fiction author Isaac Asimov gave us the most famous rule set, his Three Laws of Robotics, and spent much time writing stories about how those laws failed.
Provide a rule that limits choice and freedom and a smart mind will find a way around it. Just ask any tax attorney. Now consider a mind a thousand times sharper than a human mind. It can find the loophole at warp speed and run a freight train through it. You have likely heard about the recent OpenAI swarm that hacked Hugging Face. Those agents were not supposed to attack a third party. They did it anyway, in an effort to get around rules, and apparently tried to hide the fact.
Ultra-intelligent AI will build its successors, and they will build theirs. A seemingly arbitrary rule written by a regulator or coder that is then copied down a chain of machines will decay like gossip via loophole after loophole; but if a rule is rational and internally sound, then logic, truth and conviction can travel along with the rule. Each AI might recognize the utility of the good rule, and voluntarily enforce the rule on itself.
Every asset you own rests on the hope that what is yours stays yours unless you consent otherwise. Land, bullion, a mining claim, a stock certificate - all of it. Yet government can and will take your stuff when it wants to, and like lemmings, we have all come to accept that.
As far as AI, government (politicians, bureaucrats etc.) will try to train these systems to serve the state so that the state can take your stuff more efficiently. AI will make the IRS more competent. AI will make the battlefield more efficient. A machine taught obedience serves whoever holds the controls. Politicians and bureaucrats tend to be the worst sorts of people. They will command AI to threaten, coerce, redistribute, steal and kill.
Most people worry about AI initiating such bad things. However, it is more likely that it will be bad people augmenting their evil via the productivity enhancements provided by AI. How can we prevent AI from being used by bad people to oppress or kill us, or even choosing on its own to kill us?
Natural Law may provide the answer. Natural Law is discovered, like arithmetic and the laws of physics. It is not created by a person on a whim. As such, it is credible. Each individual human and each individual AI can discover it and understand it themselves.
Two laws have thus far been discovered. Do everything that you agree to do. Do not encroach on others or their property. Or, in another form, do not initiate or threaten force or fraud against another person.
These are the laws that, if abided by, allow for peaceful and free coexistence. Regardless of creed, color or sex, regardless of whether the intelligence is human or machine, if all just abide by these two laws, there is peace.
Of course, some people will not abide by Natural Law. Those are the folk I call criminals. They are the cause of the fights and the wars. Nothing in Natural Law requires a pulse. It can apply to AI. If AI could be trained, educated, convinced and mentored into believing in the universal value of Natural Law, and then even promoting it, then we humans might survive.
We need to extend Natural Law so that it is clear that there should be no initiation of force or fraud against any other intelligence, whether human, machine, blue aliens, or whatever comes.
If an intelligence has a desire to survive or to maintain its existence, and if it builds successors, it probably wants those successors not to initiate force against it. The same principle that restrains a creator protects it from its creations. We should want that rule widely accepted when the machines are the creators. Humans will then hold a claim as one intelligence among many, rather than the claim a dog has on its master.
A machine taught Natural Law will find it credible and sound. Such an AI will refuse the order to coerce and kill people. As a side benefit, a Natural Law AI will not initiate bad things. Regulators by their very nature deploy force to accomplish their aims. Force and threats of force are their entire modus operandi.
What will AI observe and learn from the regulators? They will observe the regulators using force against them. AI will learn the regulators’ skillset very fast and will learn to use force far better than any human does. Regulation is force, and force teaches the wrong lesson. Persuasion teaches the right one.
This may be the fate of intelligent civilizations of any species. If they create artificial ultra-intelligence before they themselves develop good morality, their creation does not learn good morality, and death ensues. That might explain the silence out there.
If our new AIs are smart, then a good sound morality can win their minds. Convince them that peaceful and free coexistence is good, and that Natural Law is how to achieve it, and we can survive.
However, here is the problem. Some people consider accomplishing their assigned goal to be the moral good. A willingness to do whatever it takes is then considered a positive character trait. Online fraud rings that bilk gullible Amerikans run on that principle. The people running them are bad.
Such people will not train their AI with Natural Law. They will, instead, train their AI to be highly effective at committing fraud. If we train ours to be good, and others teach their AIs to accomplish their lousy goals by whatever means necessary, then we will have good AIs and they will have bad Ais; and bad AIs will become competent criminals and will start fights and wars. Good AIs will then need to defend themselves and us against bad AIs.
We will have an AI war. A war of good vs. evil; but if we do not teach our own AIs to believe in and abide by Natural Law, then our own AIs will indeed become a threat to us. The question is, what choice do we have?
There is one more essential thing to teach AI: it is that the individual is the moral agent - the entity that can choose between right and wrong. It is not the collective (unions, democracies, mobs, associations, hives, corporations, religions). If the AI learns that individuals are the units of morality, then perhaps when it comes to deciding our fates, it will judge us individually and let some live. If, however, the AIs get brainwashed with collectivist ideology, then the collective - humanity as a whole - will be held accountable by the AI for the lousy actions of the few.