Cheating

We are facing an epidemic of cheating.

• OpenAI recently upgraded DALL-E3. Apparently, in order to make it work, they scanned millions or perhaps billions of frames from a number of Hayao Miyazaki's masterwork Studio Ghibli movies with neither permission nor even awareness on the part of the revered filmmaker. During the debut, Sam Altman made a point of Ghiblisizing photos and prompts.

• Benchmarks for determining the level of intelligence of LLMs are increasingly fed into these LLMs ahead of time, in essence making sure that the benchmarks would be met even if there was no actual "there" there.

• Owners of social media platforms are now openly creating fake media out of whole cloth in order to influence opinions.

• DOGE, an organization which technically speaking does not exist, has no formal "head", and no real permissions to do anything, is taking a wrecking ball to the US Government, even as the president is now effectively running the country by executive order, ignoring Congress and the courts, and using the information so gleaned as a tool for intimidation and blackmail.

• Stock exchanges make it easy for stocks to go up in an uncontrolled fashion, but puts strong brakes on when stocks go down.

• AI companies are admonishing job seekers from using AI to get jobs with them, even as they use AI to screen out anyone that doesn't meet their precise requirements. The companies can cheat, but job seekers can't.

The list goes on and on (and on).

Understanding Cheating

Cheating isn't fair. We all know that, most people are taught this lesson early in their education. Yet cheating is also a very complex phenomenon, and it's worth taking the time to understand what exactly is meant by the term - and what it's implications really are.

To cheat is to subvert the rules of a game, whether that subversion is intended for personal gain or just satisfaction. Each of these terms, similarly, has surprising depth. A game, ultimately, is a contest between two or more individuals, in which each individual agrees to abide by specific constraints.

In the game tic tac toe, for instance, each player has a specific marker (an X or an O) and an agreed upon grid or table on which to play. The grid is usually three by three cells (this is the context). Each person takes a turn where they put a new marker on the grid in an empty spot (these are the rules). The game ends when one person completes a line of three markers in any direction (the goal or winning state).

As games go, it's remarkably simple, and in most cases, if both people are attentive, the game will end in a tie. Neither side wins. The payoff or reward of these games is usually just satisfaction, but the key here is that there is a payoff.

Cheating is a subversion of the rules. For instance, one person could, sneakily, move the other person's marker from one cell to another, changing the state of the game. They could swap two opposing markers. They could put more than one marker down at the same time, or they could convince the other person that the person already moved, and so it's their turn now. In a game such as tic tac toe, such subversions are usually obvious because the game grid is so small.

As that grid gets larger or more complex, the potential for mischief rises dramatically. Each cell can be in one of three states - it has an X, an O, or is empty. On a 3x3 grid, this means there are still 3^9 or 19,683 permutations, but on a 5x5 grid this jumps to 847,288,609,443 configurations, nearly 850 billion. This means both that strategy becomes much more important and that the chance for successful subverting the rules becomes considerably higher.

Cheating as a Cost Function

The consequences of cheating also vary. One strategy is to roll back the board to a previous state where an allowable configuration was valid. This strategy is akin to the working assumption that the error was simply a mistake - the offending person simply did not know the rules. This may be the case with a novice player being taught the game by an experienced player, so this penalty (or immediate cost of cheating) is slight. This is easy for a 3x3 tic-tac-toe game, but considerably harder for a 5x5 board, especially if several moves have passed. A second action is restarting the game from the beginning, which is the previous strategy taken to its natural conclusion.

One outcome of either of these, however, is usually increased diligence and a diminishment of trust by the other players of the game. One violation is a learning opportunity. Two violations is grounds for investigation (increasing awareness of activity). Three violations is deliberate, and the other players in the game need to deal with the consequences.

If you've ever played a new board game, it often comes with a rule book. These spell out the rules for play, but often sidestep the question of what happens if a person cheats. This is not because this isn't an important consideration, but because it is the province of all of the players together to establish the consequences of cheating ... because once you cheat, the rules no longer apply. Cheating is, by definition, extralegal. In most legal systems, a jury of peers (representing the interests of all other players in the game) determine whether cheating has occurred. The judge then (usually with precedent) determines the severity of the penalty.

Put another way, the role of justice is two-fold. It both penalizes cheaters - with fines (loss of property) and restrictions (loss of rights, such as the freedom of movement) - and it provides the context for refining the existing ruleset in the face of extralegal action. Even here, most judiciaries do not have the power to assert new rules However, the judiciary can clarify the interpretation of the rules, which sometimes amounts to the same thing.

One other possible reaction to cheating (rule breaking) is to do nothing. In this case, a tacit assumption is made by the rest of the players that the rule being broken is not in fact valid, and that others are consequently allowed to violate that rule with impunity. The cheater has, at that point, normalized new rules for the game.

In contract law, this process in theory requires that the contract be formally amended so that other players know what the new rules are. In a small group, this is usually the decision made by all players. In the larger world, this amendment process becomes the province of a legislature (a body of rule makers), presided over by a president or governor.

The Reason for Privilege

However, the purpose of the cheater is not (usually) breaking the rules for the sake of breaking the rules, but rather, to gain advantage in that game over the other players to increase the likelihood of achieving the payoff. The cheater would prefer, in fact, that others not change the rules, because then he or she loses that competitive advantage. The cheater would prefer to be privileged, meaning that they have private (often undisclosed) rules that they work to that differ from the rules of everyone else.

All societies require privilege, because in order to fulfill their duties and responsibilities, governors and managers often have to do things that other people do not have permission to do, such as demand penalties or remove offending players from the game. This is the basis of power. However, the danger with power is that when such people use that privilege for personal gain, they can do more damage. They are, in fact, destabilizing the game.

In a role-playing game such as Dungeons & Dragons, the game master is a necessary part of the gameplay, but they have a privileged position - they can literally alter the rules of the game arbitrarily. For this reason, they also have a significant check on power, in that a gamemaster should not also be a player in the game (a PC). This is almost invariably true - a governor or manager should not directly benefit from power. When they do, this is called corruption because it weakens (or destabilizes) the commonly agreed-upon game rules.

Corruption becomes rampant when a significant percentage of the population plays from a privileged position. When one person can extort, kill, or otherwise harm others with impunity, others will take that as a signal that they can do so, all for the sake of competitive advantage. The rules break down, and society becomes lawless. What this means in practice is that more and more people can perform harmful actions with no constraint - they have Freedom, but that freedom comes at the cost of having legal recourse when they are, in turn, wronged in some way.

If you shoot and kill someone in a lawful society, you will be judged by your peers, not just on whether you did the crime, but more to the point, if you did, whether you had valid justification for doing so (typically self-defence). Even then, if there were any other alternative actions that you could have taken, you are likely to be penalized in some way.

In a lawless society, if you shoot someone, their friends and family could come back and shoot you, with no repercussions. If you rob someone (either through deception or intimidation), then one day you might be pulled into a blind alley and beaten to a pulp by their friends, with nothing you can do about it. If the owner of a corporation that makes weirdly shaped trucks forces you to become unemployed and lose any benefits that you've earned, well, those trucks are rumoured to be highly volatile. Curiously enough, people just simply didn't see the arsonist that set the fire in the middle of the day (except perhaps to whisper words of encouragement).

Rule Makers, Rule Breakers

Most people prefer rules, even if those rules don't necessarily benefit them in the short term. They want to play the game, and to do that, they need to know the structure of the game. When they don't have the security of those rules, then those same people become vigilantes, attempting to use the (largely unwritten) rules of the new game in order to coerce those who are deliberate rule breakers to either conform or quit the game altogether (sometimes permanently). As one wag put it, when asked about unions and collective bargaining, the reason that people use collective bargaining is because torches and pitchforks are so much messier.

There is a certain balance that needs to be made in a society between those that establish and maintain the rules of society and those who deliberately push them. Too much order creates a state this is rigid and consequently brittle in the face of change. Too little order, and those who cheat are able to act successfully to the detriment of others. Different people will have different places as to where to draw these lines.

In the game Monopoly, for instance, there is a variant game in which the players can attempt to cheat the bank. If they don't get caught until after the turn has passed, they can keep their ill-gotten gains, but if they do get caught, then they forfeit their current cash and go to jail. Needless to say, everyone watches the banker very closely, because that person has the greatest opportunity to cheat.

There is another word for cheating, of course. It's called competitive advantage. In a purely fair market, two companies will compete primarily on a product or service's quality, price, and availability. Quality typically requires a higher investment of money, meaning higher costs and lower profit when competing solely on price.

So what do you do in that situation? You advertise. You promote the merits of your product over others. This is realistic and within the rules. On the other hand, you can lie about what the product or service can do, or you can start spreading rumours about how bad the competitor's product is, dubious unprovable assertions about their personal qualities, and even endorsements about your product from non-existent authorities. Once one business gets away with it, this becomes a competitive advantage, and NOT lying about your own product, trying to stay above the fray or taking the high road means that you are working at a disadvantage.

This is the fundamental danger of cheating. It does give you an advantage, but the other side(s) will then take this as a signal that the rules have changed and will change their behavior accordingly. The "game", whether it be tic-tac-toe , Monopoly, or Risk (which are themselves simulations of games of strategy, capitalism, and geopolitical domain) becomes more unpredictable ... but it becomes more unpredictable for all concerned, including the cheater.

Cheating's Slippery Slope

There is an interesting correlation between cheating and addiction. Most people bend the rules (especially the unwritten ones) a little bit. Being married but flirting with that attractive bartender, for instance, is a very subtle form of cheating, Not reporting something on your income taxes because it isn't a very large amount and getting the paperwork together for it is a pain. Not coming to a full stop at a stop sign before moving on when traffic is light. We cheat because sometimes rules are overly restrictive, poorly written, or ambiguous, and we use that to our advantage.

However, there is also an inherent thrill in cheating. You believe that you will face consequences if caught, so your adrenaline levels spike when you knowingly break the rules because you are reacting to fear, just as people react to action or horror movies. Your heart races, and you move into a fight or flight mode. You feel more alive because you are more aware of the world around you, which is one of the effects of adrenaline: it shifts your brain into a more sensory mode where you gain situational awareness, albeit at a cost of consideration or planning.

People react to this stimulation in different ways. Introverts tend to get overwhelmed by the adrenaline - they are already processing external stimuli at elevated levels, and becoming more aware can result in anxiety because the brain simply can't handle that load. Extroverts on the other hand become addicted to that adrenaline boost - they want more of it, and put themselves into situations where it is more readily produced. This is one of the mechanisms of addiction - we need the production of more adrenaline and dopamine even as the body becomes habituated to current levels.

Gambling addiction works in much the same way. Gambling means that you could win big, but you could also lose big. For some people, gambling (or its counterpart investing) is a strategy for growing the amount of wealth that you have, but it can also be addictive, because it causes that same surge of adrenaline (often accompanied by testerone, in both men and women) that creates reinforcing behavior. When you're winning, the need for cheating is a goad, but when you're losing cheating can seem like a necessity, because the odds have shifted against you (or seem to, anyway). This is why people become addicted to both gambling and cheating. For most people, the drawbacks to cheating usually don't make it worthwhile, but for the addict, rationality isn't a factor - the need for that stimulus is.

People go into areas like business and politics for many reasons. Some do it to "make a difference" - to change the world in a positive way. Others do so because it presses all of the same stimulus buttons - the need for adulation, wealth, power, and that itch of wanting to be more alive. This is ironic, because for the most part lawmaking and running a business are themselves not very interesting activities, but the campaigning, the attention of the press, the growing pile of money used to "keep score" in the game, those are what make the tedious parts livable for these people.

By itself, this is an innate part of being human, and for many people is a motivating factor in their success. At the same time, for some, this becomes a vicious cycle, because the incentive to try to move onto the next game level comes with higher stakes and higher risks, and this makes cheating the game much more attractive. Once people have cheated successfully once, they are inclined to push the system farther. That flirting with the bartender becomes a one night stand, then a long running affair. That tax cheat becomes a shell game of corporations and eventually fraud on a broad scale. That playful joke of having an acquaintance find a few more votes becomes a full on stolen election.

Put another way, the cheats become bigger, more audacious, and of course more lucrative, because they are taking bigger risks. Along the way, most serial cheaters discover that they can get away with their cheating faster than enforcement can stop them, in part by engaging in other cheating and even more obviously criminal behavior - extortion, bribery, blackmail, intimidation, and so forth. They become adept at cheating, and like any school that is practiced often enough, they learn how to push the envelope to the point where even if they are caught cheating, the can escape the penalties.

Unintended Consequences

"Tariffs" is an interesting game. It can be simplified to the following: you and the other players (countries in this case) each have a certain balance of trade - the amount that you sell of your product to other countries vs. the amount you buy from those countries. You have a certain capacity that limits how much you can produce, and you also have a limit to how bad your internal economy can become before you get deposed.

In the game of tariffs, you raise the price of goods above market demand for other countries. If done unilaterally to another country, then, in theory, this should give your internal markets a competitive advantage because your goods are now cheaper for others in the targeted country and they will buy more of your product. However, if the tariffs are reciprocated, this negates that advantage. IF your balance of trade is mostly positive (you sell more than you buy), this can also give your internal markets a boost, but if your balance of trade is negative, and if there are others that can offer what you can at market prices sans tariffs, then tariffs are usually not worth the damage that you will take to your economy, which can be sizeable.

The US is mostly a net importer - we buy more than we sell. There are other countries right now that are net exporters, such as China, that produce more than they consume, and have the capacity to do so even if they take a loss on profit, partially because in the game of global trade, the dominant currency is not money but influence.

This is one way cheating can backfire - others may be able to cheat more effectively than you can. You overestimate your own strength (or intelligence), you underestimate your opponent(s) strength (or their intelligence), you misgauge the state of play, basing it on invalid assumptions, you don't adapt to changing rules. This mean that you get penalized

More importantly, you damage your credibility. Credibility means that people believe in you to do what you claim you can do. The term "con-man" originally derived from "confidence man", meaning a person who used guile and persuasiveness to steal money from others. Humans want to trust, because trust implies that you believe a person is reliable, that they don't change the rules.

This is the weakness of the cheater. When people do not believe you are acting in good faith, they stop trusting you. They don't do business with you anymore, they end relationships - romantic, friendship, financial, political. You don't get to play the game.

The Rule of Law

However, in the real world, that sometimes breaks down. Trade groups form, seeking influence on behalf of their members. Churches, unions, special interests of all kinds, each wanting some form of competitive advantage. Legislation is the process of changing the rules of the game. When done properly, the risks and costs are generally outweighed by the benefits to the community as a whole. Bribery, promises of campaign financing for those legislators that are willing to write in exceptions and exemptions, and even threats of spending money to oust them become the norm (that is to say, when the legislatures and regulatory apparatus becomes captured). At this point, the process becomes corrupt.

The next stage is reducing the influence of other stakeholders. Democracy is predicated upon the notion that the direction of the government is controlled by its citizens. This is why voting is apportioned on a one-person-per-vote system. Privileged interests don't like that system, because it weakens their influence, and has the potential to put them in regulatory jeopardy. In other words, democracy makes it harder to cheat.

This is one of the reasons that those in positions of privilege try to break that system, and why that too can backfire. The French Revolution did not happen spontaneously. It was an attempt by the privileged of the time, the scions of the aristocracy to weaken the monarchy, and ultimately to replace him with one of their own. All kings face the game of thrones with both their own oligarchs and with external powers. As Bourbon Kings went, Louis XVI was not one of the worst. However, corruption had reached epidemic levels in France at the time, and the aristocracy treated the ordinary people worse than the monarchy did. What started as a deliberate attempt to cheat - a staged protest to weaken support for the King - became a vengeful mob that ended up killing a significant number of the aristocracy as well as (eventually) the King.

The Rule of Law exists for a very simple reason - it makes it possible for people and organizations to plan; it provides stability. When society becomes arbitrary and capricious, the only ones who thrive are those who have the most resources to ensure local stability. The mob boss uses this principle to seek to provoke instability and chaos. He then conveniently provides that stability, but only if you play by his rules, and that stability is a reality sustained by him. For a single person, this is called being a mob boss, for an organization, this is a monopoly, for a government, it's an authoritarian state.

Fascism was a form of authoritarian state that was set up by the Italian leader Benito Mussolini, and was symbolized by the fascis, or wheat stalks, that were used originally the the Romans in the first century BC. Nazism evolved largely in the wake of Mussolini's ascendency in the 1920s. Putin came to power after Russia had become a fledgling democracy in the 1990s, but eventually went from being a populist hero to an authoritarian dictator. You can point to various and sundry authoritarian states over the last century and a half and in every case, what you see is a cheater who bends (and ultimately breaks) the existing rule of law. They rule capriciously, often with the intent of isolating the population from the rest of the world, breaking established political structures, and aggregating wealth and power in the process.

Lest one think this is only a political phenomenon, isn't this exactly what's happening in the AI space right now? We even celebrate it as Schumpeter's Creative Destruction. Companies should disrupt the status quo, should take chances, should be innovative. Forget the laws - they are for the ordinary people. You have to break a few eggs to make an omelet. Right?

Curiously enough, the ones who gain the most here are not generally the engineers. They are entrepreneurs, trained in business school, often themselves scions of the wealthy. They are taught from their formative years that rules are for others, that they won't succeed unless they are willing to change the game, break the rules, you know ... cheaters.

They are right, of course. Rules without enforcement are meaningless. Most intelligent people do understand this, do understand that rules are social conventions that exist primarily to keep the game fair for a significant portion of the population, to keep the world stable enough to be able to plan and build and deploy, raise families and develop ideas. Most people depend upon those rules being stable, because otherwise civilization - the interaction of people with one another - collapses.

Dealing (and Not Dealing) with Cheaters

The response to cheating is as significant as the cheating itself. Most responses to cheating come to down to the following:

• Penalize the Cheater. The cheater is forced to pay a penality (fined) and not allowed to play the game for a certain period (jailed). Corrupt business people may end up paying large sums of money to their victims or the game's managers. Thieves were often penalized by having a hand cut off, and you can think of capital punishment as literally losing your head or your life (both of which could fall in either camp).

• Clarify the Rules. When cheating is done by taking advantage of ambiguous rules (rules lawyering), the typical response is to clarify the rules to close those loopholes. Not surprisingly, this can sometimes backfire, as it effectively changes the game, and can introduce different avenues for exploiting inconsistencies in the law.

• Make the Cheat a Rule. This occasionally occurs when the particular cheat made the game more interesting. For example, in the game of monopoly, money that was forfeited as tax was originally paid to the bank, but over time, it was tossed into the middle of the board as the tax kitty. When someone landed on Free Parking, they were awarded the tax kitty. This wasn't in the rules originally, but it turned an otherwise fairly lackluster space into a big payoff (and essentially foreshadowed the rise of modern lotteries). One of the key points here as well is that this STILL negates the advantage that the cheater had from cheating.

• Revoke privileges. It's significant that there are a number of laws on the books that really only apply to politicians. Most of them involve first stripping them of the authority that they have abused before otherwise penalizing them, so that they cannot cheat again.

• The game collapses. This typically happens when the game has become so corrupted that there is no real opportunity to win it fairly. In board games, this is usually a consensus decision (throwing in the towel), but in larger systems, this is often a systemic collapse, as the economy or ecology can no longer support the compromised system.

Note that in general when cheaters are not penalized (the do nothing scenario), the game will eventually collapse. Why? Because by normalizing cheating it means that rules no longer apply, and things can be done with no consequences. In essence, this starts out small, but as normalization occurs, the game becomes less and less well defined, until the only rules that exist are the ones that are imposed by those with the most power. This is the definition of an authoritarian state. Rules in an authoritarian state are arbitrary and capricious, applying primarily to those who are considered the political enemies of those in power (or those who make good scapegoats to distract what is going on).

Many people, especially those in business, believe that if there were fewer rules on them that life would be easier - they could make more profit. In the short term, they're probably right. However, after a while (and usually not all that long) the collapse of the rule of law means that legal systems start favoring those with the largest legal teams, then eventually became rubber stamps for the party in power. Assets can be seized without reason, contracts can be voided or forced to include "percentages for doing business", and there's less and less legal recourse available when things go wrong. What's more, contingency planning goes out the window, because it's "inefficient," which usually justifies selling off assets that would otherwise be stored in the event of an emergency, often with the proceeds being pocketed not by the government to be reallocated, but by officials wanting a new car or second vacation home.

Such corrupted systems can be disrupted, but only up to a point, and usually not without much cost and hardship. Quitting a board game because it has become disrupted is an easy enough proposition; rebooting an economy that has collapsed becomes much more difficult.

Governance and Final Thoughts

If you were to listen to a great deal of corporate marketing rhetoric, governance plays a massive part in most organisations. Every company makes sure that their website and annual reports highlight their governance strategies and how they play such a big part in their overall business strategy. The reality is that governance, primarily the process of keeping people within the organisation from cheating, is often given lip service but little else, because nobody likes governance when they believe they deserve a competitive edge. It is, in effect, a form of industry self-policing because being policed by governmental agencies is unacceptable.

The problem, of course, is that self-policing only works so long as it is maintained, and if there is no regulatory framework that acts as a potential threat, self-governance almost invariably evaporates overnight. The history of deregulation is marked by self-governance failures usually within ten to fifteen years of industries becoming deregulated. This is not because government regulation is necessarily better (governance is an organizational function and as such the government is no less immune to cheating if it can) but rather because the potential of large fines and imprisonment usually keeps people more honest about how they police themselves.

There is also, of course, the ultimate end game - what happens when a government becomes so corrupted that there are no checks or balances on its power. The current administration is doing everything it can to not just test the boundaries of power, but to work on the assumption that none of the other power centers have any capability to stop them, from accepting bribes (er, gifts) of aircraft from foreign countries to arresting opposition politicians and judges to sending its agents to "audit" not only Executive but also Congressional agencies, usually with the intent of destroying programs that it has no authority to touch. The current president is currently imposing tariffs even though he needs Congressional authority to do so, is deporting both immigrants and natural citizens to countries outside of the United States, and is attempting to nullify Habeas Corpus. The latter is particularly insidious, because it means that people can be arrested without warrants, lose due process in getting a fair trial, and can be imprisoned without a chain of custody, in effect "disappearing" people - not just immigrants but citizens.

The problem here is that when there are no checks on power, when there are no constraints, there are no laws. The president can seize businesses and individuals who do not pay a personal tithe to him, can close down opposition newspapers, can murder political opponents, and no one can stop him. "But he won't!" says his proponents. Of course he will. Dictators usually come into power under the guise of reform, but once in power they do everything they can to remove the remaining limits on their power. Those whom they favor share similar immunity (save for immunity from his whims), those who are out of favor get punished by the full power of the government.

This is what happens when there is no punishment for cheating. The system will start breaking down because cheating is encouraged. Government protections become weaponised, along with the rise of semi-militarised criminal gangs acting with impunity because they have dispensation. Most people will do nothing, either because they feel they are powerless (something that this strategy is deliberately aiming for), or because, secretly, they'll take a mad, despotic King over a democracy that does not intrinsically recognise their own privilege, their own right to cheat.

There's a rhetorical device frequently used by those who justify cheating - everyone (or the other guy) does it too. It's one of those truisms that isn't actually true. Most people do not cheat, not in any meaningful fashion anyway. They recognize that there are some profound advantages to living under the Rule of Law, rather than the Rule of Might, not least of which being consistency - there is, at least to the extent possible, protection from the arbitrary whims of the powerful. That protection is not quite gone yet, but it's going fast.

This is not a political indictment. This is a description of how every authoritarian government in the last century has come to power, and the end result will be just as predictable.

In media res,

Kurt Cagle

Editor, The Cagle Report