The GDPR here ensures that an unsuccessful applicant can find out which of his data were decisive for the negative decision. However, most algorithmic decision-making processes still involve people – and then the GDPR does not apply. Industry does not yet find any solutions to ethical problems For this reason, former Federal Minister of Justice Heiko Maas (SPD) last year even October 2018 Calendar Cute proposed an anti-discrimination law for algorithms (AGG) in order to prevent a systematic discrimination of population groups and to ensure “unprejudiced programming”.
October 2018 Calendar Cute
So far, nothing has come of it, also because the economy was running against the project of the Social Democrat Sturm, which from its point of view was “anti-innovation”. In the IT industry, one has recognized the ethical issues associated with transferring many decisions to algorithms. Attempts to get a grip on the dilemma itself are silted up after a new Bertelsmann study.
Although there have been ethical guidelines for computer scientists since 1994, they do not have a binding character and therefore often remain empty claims, the Bertelsmanns complain. Guest Commentary: The Silicon Valley Tech Giants Endanger Fair Competition GUEST COMMENTARY Silicon Valley’s Tech Giants Endanger Fair Competition In order to curb the social power of algorithms and its sculptors in Silicon Valley, SPD leader Andrea Nahles has this week taken up in the Handelsblatt an unusual proposal.
Nahles wants to oblige the big digital corporations to publicly share an anonymous and representative part of their data if they have exceeded a fixed market share for a certain period of time. Among other things, the Social Democrat relies on an idea by the legal scholar Victor Meyer-Schönberger, who teaches the threshold at a market share of ten percent. “If a company exceeds this threshold, it has to share some of its feedback data with all competitors who want it,” the Austrian demands.
On the other side of the Atlantic, Europe’s attempts to tame technological progress are viewed with skepticism. “The regulation of individual algorithms slows down innovation and makes it harder for companies to use artificial intelligence,” warn Darrel West and John Allen of the American think tank Brookings in Washington. Algorithms do not have to be perfect Rather than breaking the “black box” of algorithms, it is better to set broader goals for the IT industry and then promote them with political measures. Although there is particular US-Bundess States such as New York are also trying to check the fairness of public service algorithms.
However, claims to disclose the source codes of digital helpers or directly regulate the development of algorithms have so far not found a majority. Angela Merkel: “Which minister would you like to replace with artificial intelligence?” Image ANGELA MERKEL “Which minister would you like to replace with artificial intelligence?”
Finding the right balance between regulation and freedom from October 2018 Calendar Cute innovation is one of the most important future tasks of politics. It should be based on two thoughts: not everything that is technically possible, must also be made. At the same time, algorithms do not have to be perfect. It is enough if they do their job better than the people. To find out, the inventors need our digital future freedom.