Are You ready for algorithmic regulation?

Are You ready for algorithmic regulation?
Are You ready for algorithmic regulation?

Have you heard of “algorithmic regulation”? This is a new approach that is being suggested for governance by technologists in the USA. What this concept means is an alternative form of government where the advantages and usages of computer algorithms are applied to regulations and law enforcement. Written laws are not replaced but stressed to test its efficiency. The term was first coined by Tim O’Reilly. But first of all, what is an algorithm?

So, could you imagine a world ruled and governed by algorithms? Tim O’Reilly who has a lot of influence in the technology market is an important proponent of algorithmic regulation. According to Tim O’Reilly writing for Beyond Transparency, the types of regulation that could be achieved through an algorithmic approach all have several factors in common. One is that they have a deep understanding of the desired outcome. The second is real time measurement to see if the outcome is being achieved. A third is algorithms, a set of rules that are able to make changes based on new data. The fourth is the possibility for deep, regular analysis of whether algorithms are correct and performing as they should or could be.

In the words of O’Reilly:

“We are at a unique time when new technologies make it possible to reduce the amount of regulation while actually increasing the amount of oversight and production of desirable outcomes.”

It is argued that there are already several projects that try to link consumer reputation systems – (specifically their reach and user friendliness) with government data. One example is that of the LIVES standard which was reportedly developed by San Francisco. Another is Code for America, and Yelp. This links health authority inspection data with the information on Yelp, providing better and more advanced information to customers to help them to make better decisions.

Sensors are going to have a role to play in algorithmic regulation, and there is an underlying factor to consider, which is the types of data that need to be and are collected about people to be able to achieve algorithmic regulation. This presents particular challenges for privacy which need to be considered and overcome so that this type of governance can become acceptable to people. Some people may find the types of regulation being discussed quite disturbing and perhaps an infringement on their privacy, but nonetheless they do have the potential to happen anyway. Yet there are also distinct advantages to the whole concept. One example is provided of congestion pricing to better manage traffic. Currently in London the system is the license plate is read and payment has to be made by the customer. In the future, this could be replaced by automated billing.

One of the first tasks in getting to this point is figuring out measurement. There are plenty of different types of data that can be drawn on to achieve this. There is also a need to overcome the risks. One worrisome risk is explained to be scope growth. Data may be gathered for one activity and then used for other purposes, not initially considered. It is argued that the way around this is to still collect the data but to be very clear about its use and to make sure that there are safeguards to prevent people from using data in ways that it was not initially collected for, to make sure that there is no misuse or abuse of data. There is much to be done, but algorithmic regulation appears to be well and truly on its way.

Writing for The Observer, Evgeny Morozov (2014) explains that algorithmic regulation has a lot of different applications, particularly with regard to governance of the roads. As Morozov explains, these applications could be quite far reaching, such as one being considered in Europe:

“European officials have considered requiring all cars entering the European market to feature a built-in mechanism that allows the police to stop vehicles remotely.”

This is not as far fetched as you might think. It is explained that Ford already has the ability to know what people are doing in their cars and when they are breaking the law, because they have GPS tracking in the cars. While they do not supply that data to anyone, there is data there that could potentially be used in such a way, under the right conditions. As roads are built to include more smart technology, something that is already happening in the UK, there is a definite potential to offer law enforcement that is immediate. Roads that have sensors are able to send signals to smartphones in cars, putting in place a system of speed limits that are varied, and diverting traffic away from congestion.

Traffic law enforcement is not the only way that algorithmic regulation could be used. There are opportunities to use smartphones and wearable technology as well.