Laconia 3 - Faults in the plan

1. After a long time. I hope this chapter remains true to the language, mood, and style of the previous two.
2. I found that describing the faults in the plan took so long that I was left with no space to describe the actual revolution itself. Much of what I planned for this chapter has now been postponed to the next. This also necessitated a renaming of this chapter.

(Development, nowhere in general)

The central assumption behind letting convicted felons free with Brain-Tap installed on them was that their thoughts would now be accessible to the police. The loss of every bit of one's privacy was believed to be a punishment as harsh as life imprisonment, and deterrent enough to those who thought of erring on the wrong side of the law. A nationwide thought network was established - contracts were signed with cell phone operators to accept signals beamed from Brain-Tap devices. And since the cell phone network covers most of the country, there was almost no restriction on the movement of subjects, unlike the case with a typical parole programme. The criminals participating in the Neuro-Electric Social Rehabilitation Programme (NESReP), were about as free in action as any other normal citizen of the country. In thought, they weren't, for there always was Big Brother, watching everything crossing their mind. There was no worry of them escaping if they couldn't think of doing so. To those who were quick on the uptake, this ruled out freedom of action as well, but it wasn't always mentioned that way. And if thought could be controlled, then so could one's actions.

Two years after NESReP came into effect, we see that the crime rate slowly doubled. But to put things in this way would be a gross oversimplification of what really happened. In short, although the crime rate initially increased, the rate of violent crimes actually came down drastically. The number of petty robberies started to dominate the crime statistics. The entire story of NESReP's two-year impact might interest a few.

The introduction of the clause had its desired effect - overcrowding in jails slowly came to an end. People started opting for Brain-Tap over spending the rest of their lives in jail. However, it appears that fear of punishment is what keeps many humans from committing inhuman crimes, and although the government had spent considerable efforts in showing the problems with opting for the Brain-Tap clause, it did not appear as harsh a punishment to many. The result was that the crime rate slowly grew from its value before the programme began.

The entire plan went off as expected. Whenever criminal ideas went through a subject's mind, the local police would be alerted, who would then swiftly proceed to prevent the crime from happening by various methods, such as providing the victim security, counselling the subject, and (rather rarely, it must be said) taking the subjects into preventive custody. Thus, although the crime rate doubled, there were no almost no repeat offenders, and first-time offenders are inexperienced and nervous about their actions, the crimes that did happen were small.

But this was the two-year impact. Things started looking differently in the long term. But again, it would be wrong to say that everything that went wrong happened because of poor implementation, or poor judgment. There were a few problems that were unfortunately overlooked - the scientists perhaps didn't stress them enough, and the politicians were perhaps not attentive enough.

In a country like India, the government controls most of the telecommunication network. If it wished to listen in on somebody's telephone conversation, it would not have to go out of its way to do so. The lengthy legal procedures guaranteeing a person's right to privacy are simply there to assuage people's fears, to tell them that the government won't really tap somebody's phone line if its not really necessary. If thought about critically, this promise is really nothing by way of a guarantee, but a man's right to talk to his wife without the government listening is saved by a very different reality of life. There are over 30 million wireline telephones connecting to the state-owned telecom network, with over 500 million telephones in all. Listening to every telephone conversation is simply impractical. It is rumoured that agencies like the NSA which do something like that for American telephone conversations have devices that react when they hear sensitive words - "bomb", "terrorism", ... It is quite obvious that such systems can easily be subverted, and they're quite ineffective against the very people whose conversations the authorities wish to hear, but this is sometimes how the authorities guarantee public safety.

Another problem was not entirely unavoidable - it had to do with the structure of the neuro-electric reception network. The telephone network again serves as a good example. Most modern telephones are of the DTMF type - if you pick up the receiver and dial a phone number, you hear a series of tones. What is interesting is that these are the actual tones that are transmitted over the telephone line, and if one were to somehow externally generate these noises, then the same number would be dialed. So what's the problem with this? The problem was dramatically demonstrated in the early days of computer hacking by "phreaks" (some of whom, such as Steve Wozniak and Steve Jobs, would go on to become important people in the computer age). The tiny loophole phreaking exploited was that telephone exchanges needed to communicate between themselves, and they chose to use some of these same tones to do this. So if someone figured out what messages were being sent over the telephone lines, he could do the same from outside, and do otherwise marvelous things such as not getting billed for long-distance calls. If a computer scientist were to examine what went wrong, he'd say it was because of the weakly-typed nature of the telephone network, in that it didn't actively distinguish between what are actually different types of signals - audio and dialing. It is interesting how often weakly-typed protocols have caused harm in human history. There is now a common method of hacking websites which is termed SQL injection. This is possible because of the weakly-typed nature of the SQL language used to query the underlying database. The appeal of weak-typing lies in how it allows an indiscipline of thought. And although numerous studies have shown that weak-typing doesn't actually boost productivity, weakly-typed programming languages, such as PHP, and weakly-typed styles, remain as popular as ever. And as engineers designed the Brain-Tap network, they didn't distinguish between packets of thought, and packets of control, which would carry control information back to the central servers.

Laconia 2 - The unsaid, the unseen, and the unheard

1. This chapter is meant to act as a big plot-dump.
2. Is CRW's role too cliche?

(Background, nowhere in general)

The lawyers had not woken up to the reality of Brain-Tap until it was too late. But they could not really be blamed for this - most research in neuro-electrics had been classified top-secret. Even the researchers working on the projects didn't know until then that there were some research topics that needed approval from some agency called the Confidential Research Wing, or the CRW. The Government didn't acknowledge their existence (nor was it clear how they obtained funding, without mention in the national budget), but it appeared that they were an independent organization reporting directly to the Home Ministry and the Information and Broadcasting Ministry. They could declare which journals were "permissible" to publish research results, and even say that some research was outright "unpublishable." It appeared that research in atomic and nuclear sciences was regulated by the CRW, but all this was mere speculation, since CRW agents never spoke about their other work. The scientists working on Brain-Tap couldn't care less - the CRW appropriated funds (which were in surprisingly large amounts for a government-funded project) and provided test subjects (who were surprisingly numerous for such "dangerous" human experimentation).

The first test subjects were not humans, but members of a phylum called Cnidaria, or jellyfish, as the public better knows them. With just a few hundred neurons, they have the simplest structures that can still be called nervous systems. For every 10 terabytes of neuro-imagery that a human produces, a jellyfish produces just a few hundred kilobytes. The devices for nervous sensing and algorithms for data compression were tried and perfected on these creatures before more ambitious projects were undertaken. Working upwards, later subjects had more complex nervous systems, and finally a device was made that could work on mammals like mice, cats and dogs. Progress ended here, and the device that was made so far could not really read an organism's thoughts - nobody knew what that cat sitting on the table was thinking (even when it occasionally bit the lab technician - was it angry, or just being playful?), and so nothing further could be done. Scientists didn't even know whether the cat was "conscious," as humans claim to be. It was only after Brain-Tap started working on humans did scientists see patterns that were possibly due to consciousness. Before any research could be initiated on this matter - both checking whether they truly represented consciousness dancing around in the brain, and whether they were present in other organisms - neuro-electrics research was outlawed. But we're getting ahead of ourselves here.

Basically, without human test subjects, it was difficult (impossible?) for neuro-electrics to develop. Scientists needed to know whether their data compression was sufficient for industrial applications, whether their devices were sensitive enough. More importantly, the pattern of human thought had still to be decoded. Without an understanding of how the neurons represent "thought", the 10 terabytes of data gathered each second simply represented a meaningless sequence of numbers. Most importantly, the ultimate goal of neuro-electrics was to have devices that worked on humans. One really can't say that there is a device that works on them, without first testing the device on them.

The first human subjects were convicted terrorists, on death row. They were supposedly housed in isolated confinement in the most secure of prisons, so nobody knew if they weren't there, but in a small laboratory that wasn't supposed to exist. Nobody cared. There were only a few days left for them to live, so there was no fear that they would cry foul. While working on them, "thought" was suddenly decoded. Connections were immediately made to AI - since we knew what thinking meant, we could now easily write computer programs that thought on their own. Even before that, we could make models of the nervous system (at least in principle) and observe a thinking brain, but without knowing what "thought" was, it was like knowing the answer to a question we didn't know. Man had solved the problem of producing artificial intelligence - this research was simplified (most of the innovations were omitted) and published in an obscure journal on the subject. Programs were written that went on to win the Loebner Prize, but since most researchers in AI think that the prize is a publicity stunt, it turned out nobody was paying any attention. The decoding of human thought was a most momentous discovery, but since the research wasn't supposed to have occurred, nobody was there to clap.

With the success of the first human trials, the CRW (must have) published a memo detailing the developments and requesting other similar top-secret departments for human subjects. This second wave of volunteers came from the Home Ministry, from terrorists who were caught, and who were going to prosecuted. The scientists were requested help in interrogation, immune as they were to other methods of coercion - torture, polygraph and narco-analysis. The courts were told that they were being hospitalized because of sudden illness, the subjects were told that they were being given an MRI scan to check for possible (future) disease - in the process, nobody knew, though some did care.

An interesting aspect of the polygraph test is that it only works because the subject believes that it will. When asked a sensitive question, the subject becomes nervous fearing that he will be caught. Involuntary body responses to this nervousness, such as increased heart rates, sweating, increased muscle tension, and fidgeting, are among the factors polygraph operators look for while detecting lies (The many graphs that are observed during such an examination are what earn the test the name polygraph). Subjects who are sure that the test will fail are so confident that they lie their way through such tests. Subjects who want the test to fail (secretly) cause themselves excruciating pain on all the questions, including the control questions (by hiding a sharp nail inside their shoes, maybe?) - in such cases, the test results are inconclusive.

A similar effect was initially observed when these second-round volunteers were being interrogated. Although it did work extraordinarily well on most subjects, a small but significant minority did something that would have been thought impossible. They concentrated so much on their confidence in beating the test, that researchers could not dig out information regarding their crimes. But neuro-electrics was by now one of the most well-funded fields of inquiry, and this seemed to be too simple a problem to leave unsolved. A solution was figured out, but therein lay the second great vice of Brain-Tap. The idea was to include a small receiver on the device, which interfered with the brain signals. The express purpose of the receiver was to artificially stimulate thoughts of the crime, and thus bring the subject to confessing on Brain-Tap. The problem with this approach is that it provides a trap-door leading to complete brain control.

The CRW now had, at its disposal, a large army of people who would follow its every order. The word army is appropriate in describing the subjects, not just because they were so numerous, but also because in unanimity of thought, they all behaved like disciplined soldiers marching to the front. Of course, in reading about Brain-Tap after it was made public, only a small passing mention was made of the receiver, and it was mentioned only as a stimulant to aid confession. The lawyers who were reading the scientists' descriptions of Brain-Tap were experts in law, not in neuro-science or electrical engineering. The trapdoor (which the scientists knew about, and had written about, but which was top-secret) was conveniently not mentioned.

An important question all this research left unanswered was the position of free will. If the weather all around the world, and at every altitude was known precisely at some instant, then with sufficient computational resources, we could calculate the weather everywhere at every moment of time afterwards. Since the laws governing the behavior of weather systems are now well-formalized in the Navier-Stokes equations, it is a simple matter to simulate the weather right upto some later instant. The problem with this simplified approach is that it ignores an important effect called turbulence, or what is well-known to the mathematicians as chaos. The computers simulating weather systems have only limited precision (even the data coming in is of limited precision), and the nature of the Navier-Stokes equations is such that the error does not stay bounded, but quickly spirals out of control. Even the best of computers working on the most accurate of data can predict the planet's weather for only so long before making no sense at all. A similar effect was thought to occur (although not proved) with the neuro-imagery obtained. Some of the data compression techniques were "lossy," so to speak, and the time at which neural turbulence (as it was called) set in was typically just a few seconds. Of course, the chaotic nature of the brain was not proven, and this was mere conjecture. Even though this meant that a person's actions in the future could not be predetermined in a lab, it did not rule out that they were, in principle, decided well in advance, possibly at the time of his birth. There were a few proposals to solve the problem of the existence of free will (which was, it may be recalled, "Does free-will exist? If so, then where does it reside, and if not, then which side of the bed am I going to wake up on tomorrow morning?"). The most important pro-free-will theory making the rounds was that conscious decisions were made at the time of unexpected stimuli. It was fairly clear to those handling the neuro-imagery that in periods of little stimulus, such as sleep or intense meditation, neural turbulence set in late - the simulated data would stay accurate longer. But at the moments of unexpected stimuli, such as on unexpectedly bumping into an old friend, neural data suddenly needed to be refreshed. It was entirely possible that we did not have sufficiently sensitive devices to decode thoughts during such moments, it was also possible that there were some neuron-level laws that were still undiscovered. The unexpected stimulus free will theory, claimed that it was at these moments that free will manifested itself. Of course, without an understanding of what consciousness means, the free will question also stayed unresolved. The AI programs that were developed either did not accept unexpected stimuli (they only accepted input when they asked a question, or the like) or were designed to respond randomly to such events. In this latter case, they were observed to behave surprisingly well, but nothing could be said in the absence of further inquiry. However it was indeed observed, humorously, that there was never a computer-rights movement, with the computers claiming that since they were also conscious, they should also be provided civil rights.

Laconia 1 - The free convicts

1. My first, and foreseeably only foray into science fiction.
2. A most interesting thing happened a day after I penned this and chapter 2 - I read this article in the January 2010 issue of the American Scientist. This stimulated a few more ideas - they'll find mention somewhere in chapter 3.
3. The conversation at the close of this chapter sounds lame - I know.
   

(Present day, AD 2010, National Law School of India University, Bangalore)

The moot court passionately debated the legality of the new clause in the sentencing rules of the IPC. The "rogue" clause had caught criminal lawyers unawares, disconnected as they were from developments in the relatively obscure field of neuro-electrics. The few that had heard of them while watching Discovery Channel probably dismissed them as pure fantasy. It must be said to their credit, however, that similar was the reaction of much of the scientific community.

Neurons form the basis on which the nervous system works. Each neuron, scientists knew, was basically a simple little cell, with connections to several other neurons. Based on the electrical signals transmitted to it by adjacent neurons, and its internal state, the neuron decides to itself send a signal to those around it. Much research had already been conducted in the computing sciences on the behavior of cellular automata, and it was just a matter of time before somebody connected the fields. Devices were built that could measure the electrical activity at the synapses and deduce the state of individual neurons. The neuro-electrics community claimed that these could be used to measure the "state of the brain," and thus provide a map of a person's thoughts at some instant. Other scientists were skeptical of such claims - since nobody knew how the neuron states mapped to specific thoughts, to do any such thing, one would have to treat the nervous system as a collection of a hundred billion neurons - if the state of each neuron were quantized into 8 bits, and assuming that the "brain clock" ticks 100 times every second, there would be 10 terabytes of data to analyse each second. And nobody knew whether 8 bits were sufficient to represent neuron state, or whether the brain indeed worked at 100 Hz, although some (controversial) experiments on human reflexes suggested some nearby number.

Like much scientific research nowadays, it was not a brilliant genius who came and provided the answer. The problem was worked on for several years by neuro-electrics groups around the world, and some three hundred research papers and seventeen hundred patents later, a workable system emerged. It employed several advances in machine learning, used some complex technology called thought recursion (observing the local brain response to determine the importance of neuron states), and a final few patents on "knowledge compression at source". About 100 MB of data needed to be downloaded each second now, sufficiently low to allow practical use - scientists were now able to observe a subject's thoughts with reasonable "resolution." The system was called "Brain-Tap," a harsh reminder of what this device actually was.

"It's a gross violation of a convict's right to privacy. With this the government has access to his political beliefs, his personal life,... even his toilet habits!" argued Mahesh, as the court observers erupted in laughter.

"In jail, the convict would have little privacy too. And since he's sentenced for life, he might as well get used to it the easy way," countered Yogi Bhat, his classmate in the third year of the law programme.

"Ok boys. It is time," said the criminal law professor Dr. Acharya. "Your debate today was commendable, but unfortunately for you, Mahesh, the Supreme Court has already passed its ruling on the constitutionality of the new clause. Boys, I suggest you read the full text of the ruling, which you will find in this month's Bar Gazette. I suspect you'll be seeing a lot of applications of this one clause later in your legal careers."

The said clause stated that a criminal sentenced to life in prison could escape incarceration if he underwent a month's psychiatric therapy, and willingly had Brain-Tap installed on him. Large scale strikes had been called by lawyers and civil rights groups across the nation. The matter had been referred to the Supreme Court, which was urged to declare the clause unconstitutional. Unfortunately, the court cited the overcrowding of jails, and in a one-thousand-page report overturned the petition. The new provision had come to stay.