Automated decision-making systems (ADMS)
Systems that use automated reasoning to aid or replace a decision-making process that humans would otherwise perform. Despite this, humans are ultimately responsible for how these systems receive inputs, how they are used and all system outputs. In keeping with the definition proposed by AlgorithmWatch, we consider ADMS as socio-technological frameworks that encompass a decision-making model, an algorithm that translates this model into computable code, the data this code uses as an input – either to ‘learn’ from it or to analyse it by applying the systems or users in specific ways. A bot usually performs predetermined tasks. When used to automate customer services or social media responses, bots can be designed to answer simple queries. In the context of disinformation, bots are often designed to emulate real users.
Chapter 1. RisCanvi (I): the jail algorithm
(In spanish) For more than a decade in catalan prisons, a decisive algorithm has been used for the passage of any person who enters prison. In 2008 a group of violence researchers began designing RisCanvi.
Chapter 2. RisCanvi (II): Can the next crime be predicted?
In the first episode of this podcast we learned how RisCanvi works. These algorithms caused quite a stir when it was discovered that in some cases they discriminated in their decisions.
Chapter 3: BOSCO and the bonus to pay for electricity
The rise in the price of electricity has become a real problem for many families in Spain. In order to help the most vulnerable households, the Government created a social bond years ago that subsidizes, in part, the payment of the electricity bill.
The use of predictive algorithms in the criminal and penitentiary field generates many doubts among some experts
The issue of transparency (or the lack of it) is another of the big issues when talking about Artificial Intelligence.
The allocation of public aid is one of the areas of the public sector where the use of automated systems has grown the most.