Aleations: the people-friendly aliens who will consume Big Brother from the inside out—the pure and patterned varieties—why the aleation provides a good approximation to the janion—why we need to separate information hiding and aleatography if we are to conquer Big Brother—why distributed aleatography is best.
Now in theory janography is all very well, but without a practical implementation we won’t have Big Brother quaking in his over-sized boots. And to start with we must have a concrete candidate for a janion.
If we’re going on a journey, then the best place to start from is where we happen to be right now. Of the four dimensions of information hiding, the only one that has been widely implemented to date is HEye. And these implementations have been based on cryptography. And what do the techniques that are employed to encrypt data have in common? Well, when we remove any identifying headers and footers the resulting ciphertext looks just like “aleatory data”, like random or pseudo-random data. So let’s take aleatory data, or the “aleation”, to be the practical embodiment of a janion.
Pure and Patterned Aleations
Aleatory data comes in two flavours: the pure and the patterned. The former is random or pseudo-random data, through and through. The latter consists of mixed data, with a certain percentage of the data following some non-random pattern, and with the remaining percentage consisting of pure aleatory data. In practice, the patterned aleations that are found on most computers consist of image, audio, and video files. Typically, the higher portion of each byte, or set of bytes, is patterned and represents file content, while the lower portion is just random noise generated by physical processes within the camera or microphone.
Pure aleations are efficient carriers of information, as we can produce ciphers where the length of the ciphertext is comparable to the length of the plaintext that has been encrypted. Patterned aleations, on the other hand, offer poor storage densities as the percentage of “noisy bits” present in a typical media file is very small compared to the percentage that represents content.
By definition, aleatory data, whether or not it is masquerading as ciphertext, looks the same to any statistical technique that might be used to characterize it, and so the criterion of “inscrutability” is satisfied.
If the only constraint on an information hiding technique is that it should produce aleatory data, then we have a vast number of techniques to choose from, so the criterion of “versatility” is satisfied.
Many computer processes produce pure aleations, and it would not be practical to ban these processes, or to modify them so that they do not. However, it would be very helpful if there were rather more pure aleations on the average computer than is the case at present, particularly in those directories that are used to store personal information. But this shortcoming is one that we can do something about.
As far as patterned aleations are concerned then we are spoilt for choice. The computer world is overflowing with image, audio, and video files. Almost everyone has, or can reasonably be expected to have, such material in directories that contain personal information. Patterned aleations therefore satisfy the criterion of “duality” extremely well.
We can easily store any software tools that are needed to manipulate aleations inside aleations. Then we can use dual-purpose software to extract the aleatory software and convert it into runnable code.
Pure aleations are rarely exchanged. Such transfers could, of course, be hidden inside encrypted tunnels, but such tunnels are not the norm for Internet communications, and the presence of encrypted communications is easily detected by monitoring software. Now, the occasional and short duration use of SSL while downloading payment pages on merchant sites is to be expected. But the frequent use of SSL, its use for lengthy periods, or its use while downloading non-payment pages would soon be flagged as anomalous behaviour. So, we need to take active measures to increase the exchange of pure aleations.
Patterned aleations are extremely widely exchanged and are well suited for the purposes of aleatory exchange provided that the volume of plaintext that needs to be hidden inside them is relatively small.
A Good Starting Point
Aleatory data provides a good approximation to what we have required of a janion. There is a trade-off between pure and patterned aleations: the former has some weaknesses when it comes to duality and aleatory exchange, but offers excellent storage densities. The latter is excellent on all janioning criteria, though it only offers a low storage density.
We should therefore adopt a strategy of promoting the creation and exchange of pure aleations to extend the janographic infrastructure. And provided we can think up a good reason why a program needs to produce aleations, we should be able to produce them with impunity, even within those regimes that proscribe encryption. When it comes to using information hiding techniques an individual can then choose between pure and patterned aleations, depending on the level of risk involved should the hidden information be discovered.
Divide and Conquer
It is important to separate the development of aleatography from the development of the various techniques that might make use of it for information hiding. Proscribing particular information hiding techniques is easy to do. Proscribing aleatography—it would be tantamount to banning Internet access and all personal computing—is not practical in any country that hopes to develop and maintain a modern economy, so with the exception of a few maverick states, such as North Korea, the roll-out of a janographic infrastructure that is based on aleations should meet with few obstacles. However, if the same individuals and organizations are too closely involved in both activities then the banning of information hiding by a particular regime might also curtail the development of aleatography.
Tyranny flourishes in a hierarchical environment; freedom flourishes in a distributed one. The Internet as a janographic infrastructure illustrates this point very well. The Internet was not developed with privacy in mind, but its distributed nature, one that spans the fiefdoms of the world’s Big Brothers, makes it very useful for constructing privacy solutions. Had governments any inkling of what this once military/academic network would become then they would have strangled it at birth. As the Internet has now become essential to the successful functioning of a modern economy, it cannot be destroyed; but its use can be monitored, and its use for certain purposes and by certain individuals can be prevented—witness the successful attempt of China to curtail access to those web sites that it disapproves of. By developing aleatography we will be steadily reducing the capacity of regimes to extract useful information from the web traffic that they monitor, thereby strengthening the Internet still further.
It’s important that the development of aleatography is done in a distributed and uncoordinated manner. While banning aleatography would be very difficult, targeting the individuals who develop it would not. Because aleatography relies on dual-purpose software, it does not advertise or draw attention to itself. It is one thing for Big Brother to know that aleatography exists; but it is quite another for Big Brother to appreciate the threat that aleatography poses to his very existence. If it is developed independently by individuals and by small groups it can grow and spread in a relentless and invisible manner, hidden even from Big Brother’s Sauron-like eye. If we have a fair wind in our sails, then by the time Big Brother fully appreciates the nature of the threat it will already be too late for him to do anything about it.