The New Runway of Rules: How AI Undermines the Rules-Based Order (and Why That’s Inevitable)
“But any rule is tyranny. The duty of the individual is to accept no rule, to be the initiator of his own acts, and to be responsible. Only if he does so will the society live, and change, and adapt, and survive.” — Ursula K. Le Guin, The Dispossessed
Dr. John Harvey Kellogg was a moral man. A man terribly concerned about the sin of sexual pollution, as masturbation was then called. Compelled by his concern, the good doctor made a quantum leap — from sin to … corn flakes.
[For]Dr. John Harvey Kellogg, the inventor of corn flakes, … cereal was not just a health food because it would improve Americans digestion. He believed a diet centered on bland foods like cereal would lead Americans away from sin. One very specific sin: masturbation.
Human history has no shortage of wacko stories like this. Some reformer-prophet type comes up with what was then a “novel” idea — corn flakes to stop masturbation — and sends it out into the world, believing it constitutes a new rule. Reformer-prophets were liable to show up anywhere: in church school, government, or business. We now know that Kellogg had a crackpot idea: corn flakes (nor any other food) does anything to halt masturbation. Even men of science (they were usually men) held similarly bizarre ideas — stumbling practices in search of a rule: Freud thought the seat of the libido was the nose; smoking was once thought to be a treatment for asthma, and heroin was the cure for a cough.
Kellogg and hundreds like him were rummaging for rules: a different order in a world they saw as disorderly. This is a noble, albeit misguided, goal. The problem was — and often still is — that their notion of “order” had no basis in fact. Factuality, factfulness, data flows and data stores were nonexistent. So, whether the rule entailed bloodletting or the segregation practice of labeling humans via racial characteristics, putting them into absurd categories such as quadroons and octaroons — humans kept thinking up ways to organize their world and rule or profit from that organization.
Then moral outrage over broken rules often triumphed over reason, observation, or experiment — which forced Socrates to drink hemlock, Galileo to admit to heresy and live under house arrest (it took more than 300 years for the Church to clear his name of heresy) and brought Oppenheimer before HUAC to name names. For most of human history, might triumphed over reason, outrage over sensibility — because we had no tool that was outside ourselves and accountable to reality. Lacking tools that provided accountability, humans invented rules. The idea of reality accountability was not known. Even today it is a new new thing and a huge step forward in human consciousness.
Rules, Rules, Rules: A Snapshot of Rules Prior to AI
There are so many rules, we forget or have lost count where they all come from. More than 30,000 statutes have been enacted by Congress since 1789. From 1995 to December 2016, there were 88,889 federal rules and regulations issued; over the decade 2005–2015, the average was 27 rules for every law.
Our games and streets have rules; swimming meets and baseball games have rules; the stock market and corporations must abide by rules. We rely on rules to guide us and to make our way clearer in a complicated world. But we never stop to ask ourselves: when did we start in earnest to make up rules and where does this notion of rules come from, anyway?
While many of these rules are, as Daniel Kahneman reminds us, heuristics — allowing us to think and act more quickly without having to think things through all the time; the origin of rules is not some human cellular propensity baked into our DNA. There was a time when the world had few or no rules; only the edicts of rulers as when Shadrach, Meshach and Abednego were thrown into a fiery furnace because King Nebuchadnezzar said anyone who didn’t bow down at his parade would be punished. The notion of a gaming setup with certain guidelines or rules came about — although it would not have described itself in this fashion — with the advent of alphabets. With the then-new technology of writing, humans were able to create commandments and eventually texts (circling the globe faster once the printing press was invented) that would, with the rise of literacy, be adopted widely and beget categorical rules among human populations. Writing a thing down gave a rule gravitas, power, supremacy.
Here’s where it gets tricky and may fly in the face of accepted notions. The very fact of writing a statement down in a book created the notion that what was written was holy; the fact that the U.S. Constitution says a certain thing induces many jurists and scholars to believe that they must be strict constructionists — meaning they must stick as closely as possible to what was written in the Constitution. Again, I’m saying that while we typically regard content, what was written, as the arbiter of our thinking — the act of writing is the rule-maker. Unlike these originalists, I do not believe that any set of words is more or less holy than another. But believers think differently; they unwittingly entrain with the tool logic of writing which aggrandizes writing itself and carries with it a host of supremacist prejudices and notions. The act of writing created the rules-based order because it gave humans a concrete marker — a book, a document, a constitution — to abide by. Then all hell broke loose as humans fought wars over rules; and they exalted what was written to make the written word unassailable, as many religious sects still believe today.
This is how rules began, the meta origin of the thing. Yes, there were edicts and bizarre rulings from potentates, but in a modern world rules are an outcome of what I have termed the alphabetic order. Writing a statement creates an order that didn’t exist prior to the writing. Once we have that order, rules follow because we take what was written down as a ruling.
How AI Rules Collide with the Rules-Based Order
What humans perceive to be ‘order’ is the accretion of written-down rules covered over by rhetoric and assertion. That assertion has often been based on specious, self-serving logic. American slave-holders (many of whom wrote the U.S. Constitution) believed in the legally enforced (and economically advantageous) order of racial division, just as most Indians believed in the order of the caste system and the lowest role of the untouchables. Order, then, is not a stand-alone ideal that is independent of nationality or political system; it is the expression of a society’s values, often complex and not infrequently contradictory.
The rules-based order among nations includes important elements such as treaties, tribunals, courts, injunctions, laws, and agreements among a host of legal instruments drafted often for a worthy cause such as peace, cessation of warfare, climate mitigation, division of land, or the right to rule, aka sovereignty. These have been enormously important throughout human history.
But as humans embrace AI and all that it entails, we are changing the wellspring of order itself; we are changing the dynamics of how order comes about, how we perceive order, how we implement it. Order is not absolute; it does not come fully formed from our political pundits, our parliaments, or legislatures. No longer can we abide made-up rules about corn flakes or any other thing. AI invites us to see: order is a tool-enabled phenomenon. When we change tools, we change our sense of order.
The key: tools create rules and newer tools can break them.
We are now moving from a rules-based order that derived from written (alphabetic) religious and territorial hegemonies, to neural network rules; AI rules that are software and machine-learning based. This is a change so profound it reaches into every area of human life, from religion to medicine to war and politics.
Here are a few basics:
- Humans operate subliminally with respect to their tools and their devices: we follow the tool’s logic without thinking about it
- Tools create (and break) rules: the origin of rules is not some cellular propensity in humans; it comes from our mindset after we entrained with writing things down, i.e., the creation of an alphabetic order
- Order is an inherent pattern recognition that follows tool logic
- Tool logic is the use dynamic built into any tool which humans must follow to use that tool. For example, using a cell phone you must pay attention to the phone’s screen, the phone’s sounds and signals, often at the expense of other things to which we must pay attention, like driving a car. When we use a tool, we unwittingly entrain with (follow) the tool logic built into the tool
- As we adopt different tools, we encounter different orders (and, of course, this is due to different tool logics)
- When tool logic changes, order changes: because we don’t acknowledge that we entrain with tool logic, we don’t see this
- Violence in response to order change is a result of ignorance, but it is predictable
- “Go fast and break stuff” speaks to a new dynamic, a new ethos possible because the tools of technology enable it: the “stuff” we break are the old rules as newer dynamics collide with fixed rules
- New technologies come with new rules; these are traditional-rule-indifferent; traditional-rule-agnostic
- The rules-based order that many are talking about (and wringing their hands over) is the alphabetic order that used injunctions and religions to make people adhere to the alphabetic order of texts, commandments, and pronouncements which became rules. Having never acknowledged this order, people do not see how form swallows function: how formal alphabets enabled the rules which created the order
- The central clash of our time (a soundless collision): rules vs. data interpretation
- Rules prescribe and proscribe; data interprets and infers: data rides facts as surfer rides a wave, to see where the facts lead
Here are the new rules that the arrival of neural networks and AI assert (and insert into our lives):
- Query and search rule AI; pronouncement and commandment play no role here
- AI is a data-driven technology: it exists because of facts and thrives on facts
- AI stacks networks of human thought, perception, and actions: AI built by humans uses human consciousness and then goes beyond that use to invent and ‘perceive’ according to new rules
- AI is an amalgam technology: it mashes together learning and knowledge
- This amalgamation collides with (and ultimately will break down) prior categorical walls like race, nationality, religion, tradition
- AI will often appear as an embedded (hidden) technology: AI will be infused into innumerable products and services
- AI is an adjunct to the human mind, built from innumerable human minds
- The human brain is now not enough; it needs the assistance of computers, algorithms, AI
- AI is ego-less, but humans using AI will bend their learning and instincts to AI prowess — which may cause some humans to question their self-worth
- AI is also built from neural network capacity which may lead to insights and conclusions the human mind has not seen or may not be capable of
- AI logic may be incomprehensible to humans and may not follow human logic
“AI already transcends human perception — in a sense through chronological compression or “time travel”: enabled by algorithms and computer power, it analyses and learns through processes that would take human minds decades or even centuries to complete.”
The New Runway of Rules: Implications
Because AI feeds on data — vast quantities of data — this single fact becomes an arbiter of the future and a harsh critic of the past. Previous civilizations had no data stores, no data mining mechanisms, no endless data flows that supported or refuted assertion, conjecture, invention. So one of the first implications of new rules is that we will begin to view ourselves differently, and especially view our past skeptically.
Data-driven realities change everything. They especially change the role of the guru, the so-called leader or visionary. What the data says is a profoundly different question than what the prophet says. Data access and analysis is a completely different dynamic than inherited, traditional rules and rule-based behavior; it ignores “Thou shalt” and “Thou shalt not” while favoring the restless movement of data, increasingly presented in colorful and well-designed visualizations.
Meanwhile junk data will become a thorny problem, as unscrupulous and self-serving media and social media platforms work to manipulate public opinion or foment discord for audience ratings and metrics. This will require new verification protocols and an ongoing awareness of the morphing nature of junk data.
Nonetheless, we will all be affected by the adjunct intelligence of everywhere-AI and the dramatic effect this has on our identity and individual perception. Human consciousness will experience a silent encroaching; AI will creep up on us in the night — and in the daytime. AI’s collective powers and uber-reasoning will challenge human abilities. This challenge will happen without much bother or awareness beyond cultural enthusiasm for AI because AI will be nestled neatly behind the tech curtain, contained and operating in almost everything we touch, invested in our objects and inventions.
Finally, a tech-paranoia backlash is coming. As jobs are lost to AI and the digital divide widens, the embedding of AI will be both a convenience and a point of contention. On the one hand we will enhance our lives with AI; on the other, as we entwine our lives with its hidden presence, AI encroachments on human consciousness will demand that humans become more meta-aware — realizing the ways we entrain with our tools that alter our thinking and behaviors. For some, this will be a diminishing of human centrality as the center of the universe — a psychological echo of the debate Galileo and the Church were embroiled in hundreds of years ago. As the authors of The Age of AI write:
“For humans accustomed to agency, centrality, and a monopoly on complex intelligence, AI will challenge self-perception.”
Coda: The Global Liberal Order
The global liberal order is a rules-based outcome.
In a bi-polar world moving to a multi-polar world that is witnessing the rise of populism, the global liberal order and the rules-based order itself are provoking criticism. Underneath these recent phenomena sits the existing edifice of the earlier alphabetic order, which was the original source of regional ethnic and cultural differences. The alphabetic order was valuable because it was some kind of order: writing down ideals gave humans a concrete thing which could be debated, amended, reconstituted — as opposed to the nihilism of endless wars which invited chaos and disorder.
Into that history now arrives a new order. It is not evenly distributed but it is already remarkably pervasive. AI and neural networks break apart the older order of prior rules because they provide an alt-logic to the alphabetic order — and new rules. When older structures break apart there is always uncertainty. Into that breakage come terrorists and populists who promote an autocratic order. But despite their scapegoating rhetoric, terrorists and populists foment chaos and will eventually seek to cure that chaos with force and repression.
So, to be clear: order is preferred over disorder, assuming that order enables a people to be free and encourages unfettered self-expression. Order, if it is enlightened, seeks cooperation among (former) polarities, not conflict. Is cooperation a realistic goal for self-serving autocrats? No. But it is a tenable goal for the new order of AI. The origin of rules is a necessary and useful understanding as we seek to craft and implement ground rules for a new order.
