Algorithmization of justice
The amount of information available for processing and analysis using computers is growing like a snowball. The data from surveillance cameras, GPS trackers, mobile phone sensors, financial transaction records, and history of web page visits are increasingly influencing decision making. And the more this data, the more it is necessary to rely on their automatic interpretation. The inevitable consequence of this is the emergence of “computer justice” systems that, without human participation, reveal violations of laws and rules. Fines for speeding, issued automatically on the basis of data from video cameras and radar or content analysis system on Youtube, which is looking for copyright infringement, is already an everyday reality.
A group of American scientists, uniting lawyers, linguists and programmers, conducted an interesting experiment in this area. In the course of the experiment, 52 programmers had to compile a program that would analyze the data from the GPS tracker installed in the car and issue fines for speeding in accordance with the New York State Traffic Regulations. This turned out to be a very difficult task - even in the most law-abiding states, laws are never executed literally and 100%. Some violations go unnoticed, some too insignificant for law enforcement officers to pay attention to them. Computers do not forget anything and do not miss anything. Thoughtless application of rules and algorithms leads to unnecessarily harsh punishments and ridiculous errors such as blocking video with wind noise for violating copyright.
Experimental data consisted of a GPS track, obtained during a real trip by car, information about the speed limit of different sections of the route and the actual code of the rules of the road. The participants in the experiment were divided into three groups. The first group wrote the program, strictly and literally following the text of the rules. The second group had to follow the spirit, not the letter of the law, and its participants independently decided in which cases the speeding was worth the fine. Programmers from the third group were given a specification compiled by experimenters, where the permissible limits for violations and sensor errors were described in detail.
')
According to the vehicle speed chart, it is clear that the driver generally followed the rules - minor peaks that go beyond the official restrictions were most likely associated with overtaking on the track or changing the terrain. At the same time, the programs that followed the letter of the law, on average, absurd 498 fines were written out during the trip. In the group that followed the spirit of the law, on average, the violation was recorded 1.5 times. The group that wrote the code according to the specification made by the experts did not find any violations at all.
In the course of writing programs, many factors have emerged, which the law does not say a word, but which have a huge impact on decision-making. What is the sensor error? Is breaking a speed limit of 5 miles per hour? And at 10? If it lasted only a few seconds? Or minutes? Or repeated several times within an hour? The legislator leaves these and other issues to the discretion of the performers. A lively police officer is unlikely to penalize a driver who has exceeded the speed for a few seconds while overtaking (at least if he is interested in real compliance with the rules, and not in the execution of the plan on penalties). The program does not have common sense, so the programmer must take into account all the nuances and special cases.
Interestingly, students from different groups reacted very differently to the possibility that their program would analyze their own ride. In the group that followed the letter of the law, only one student agreed with this, but only on the condition that there would be a “backdoor” in the system, which would allow him to avoid punishment. Similarly low were the results of those who wrote the program according to a strictly specified specification, and this despite the complete absence of fines in their version of the program. Only among those who themselves decided how the program should work, 37.5% were ready to submit to its decision.
The authors of the experiment believe that the algorithmization of justice has a great future, and the practice of using computers can have a strong influence on the laws themselves. At the junction of jurisprudence and programming, a new, highly sought-after profession may arise. Computers and the Internet have irreversibly changed many spheres of human activity. The revolution in education is already in full swing, the Watson supercomputer makes diagnoses no worse than professional doctors, and perhaps there are big changes in jurisprudence.
A group of American scientists, uniting lawyers, linguists and programmers, conducted an interesting experiment in this area. In the course of the experiment, 52 programmers had to compile a program that would analyze the data from the GPS tracker installed in the car and issue fines for speeding in accordance with the New York State Traffic Regulations. This turned out to be a very difficult task - even in the most law-abiding states, laws are never executed literally and 100%. Some violations go unnoticed, some too insignificant for law enforcement officers to pay attention to them. Computers do not forget anything and do not miss anything. Thoughtless application of rules and algorithms leads to unnecessarily harsh punishments and ridiculous errors such as blocking video with wind noise for violating copyright.
Experimental data consisted of a GPS track, obtained during a real trip by car, information about the speed limit of different sections of the route and the actual code of the rules of the road. The participants in the experiment were divided into three groups. The first group wrote the program, strictly and literally following the text of the rules. The second group had to follow the spirit, not the letter of the law, and its participants independently decided in which cases the speeding was worth the fine. Programmers from the third group were given a specification compiled by experimenters, where the permissible limits for violations and sensor errors were described in detail.
')
According to the vehicle speed chart, it is clear that the driver generally followed the rules - minor peaks that go beyond the official restrictions were most likely associated with overtaking on the track or changing the terrain. At the same time, the programs that followed the letter of the law, on average, absurd 498 fines were written out during the trip. In the group that followed the spirit of the law, on average, the violation was recorded 1.5 times. The group that wrote the code according to the specification made by the experts did not find any violations at all.
In the course of writing programs, many factors have emerged, which the law does not say a word, but which have a huge impact on decision-making. What is the sensor error? Is breaking a speed limit of 5 miles per hour? And at 10? If it lasted only a few seconds? Or minutes? Or repeated several times within an hour? The legislator leaves these and other issues to the discretion of the performers. A lively police officer is unlikely to penalize a driver who has exceeded the speed for a few seconds while overtaking (at least if he is interested in real compliance with the rules, and not in the execution of the plan on penalties). The program does not have common sense, so the programmer must take into account all the nuances and special cases.
Interestingly, students from different groups reacted very differently to the possibility that their program would analyze their own ride. In the group that followed the letter of the law, only one student agreed with this, but only on the condition that there would be a “backdoor” in the system, which would allow him to avoid punishment. Similarly low were the results of those who wrote the program according to a strictly specified specification, and this despite the complete absence of fines in their version of the program. Only among those who themselves decided how the program should work, 37.5% were ready to submit to its decision.
The authors of the experiment believe that the algorithmization of justice has a great future, and the practice of using computers can have a strong influence on the laws themselves. At the junction of jurisprudence and programming, a new, highly sought-after profession may arise. Computers and the Internet have irreversibly changed many spheres of human activity. The revolution in education is already in full swing, the Watson supercomputer makes diagnoses no worse than professional doctors, and perhaps there are big changes in jurisprudence.
Source: https://habr.com/ru/post/190212/
All Articles