Technological Threats, Nationalism, Pseudonymization, and Bavarian Elections

Yuval Noah Harari has been getting a significant amount of exposure recently appearing on podcasts such as Waking Up with Sam Harris and Under the Skin,, with Russell Brand. A general theme runs through his appearances as he promotes his new book 21 Lessons for the 21st Century, which focuses on present day concerns. He states the three greatest challenges facing the world are:

1. Nuclear War

2. Climate Change

3. Technological Development

I want to focus on the third aspect of technological development. Harari raises concerns about AI, machine learning, and algorithms and the role they will play in the future. The argument succinctly put is that computers using machine learning draw inferential decision making from data sets that outstrip human understanding while at the same time automation is increasing making people unemployable.

Paraphrasing Harari, he says: People will become obsolete, replaced in the automation of factories, in the military through cyber war and drones, and eventually in decision-making.

Harari goes on to argue that labor has traditionally been the connecting link between the general population and those in power. The worker derived a sense of value through contribution to the glory of the state and was exalted as such. When the worker becomes expendable, what happens to this relationship.

Harari views hope in our collective identification and myths such as nationalism and evolution of the collective to encompass and care about great numbers of people. In example he draws is the renewed emphasis on Middle America and concerns surrounding the Opioid Crisis and the crisis of meaning.

An excerpt from Vasilay Grossman’s book Life and Fate discusses the mythical creation of the unified collective:

“During the retreat ‘Russian’ had a mainly negative connotation: the hopelessness of Russian roads, Russian backwardness, Russian confusion…But a national self-consciousness had been born and was waiting only for a triumph.” It came in the form of Stalingrad.

“National consciousness is human. It is a manifestation of human dignity, human love of freedom, and human faith in what is good.”

The problem with nationalist regimes is that a unified collective must be working toward a common goal. Occasionally, this takes the form of an internal enemy. The topic of focus recently has been on Uyghur Muslims being rounded up in China. Some estimates put the number anywhere from 120,000 to 1m who are sent to reeducation camps. The existence of the camps is no longer a topic of debate as China recently admitted to the existence of such camps following a lengthy period of denial.

This should not be taken as an argument that mass mobilization around a common identity is always negative. History is ripe with examples of mobilization that raised societies moral consciousness like the Civil Rights Movement and the Gay Rights Movement. I merely wanted to highlight a recent international issue.

History paints the lines of division as black and white or good versus evil. The calculation is never as simple as the narrative suggests.

GDPR and Artificial Intelligence

While Technology, Machine Learning, and AI present challenges, they also present great possibilities in confronting some of our most difficult challenges. The European Parliament itself recognizes the implication of Big Data in solving real world problems in disease, climate change, reduction of energy consumption, improvements to transport safety and the enablement of smart cities while at the same time noting the use of such analytics to bolster competitiveness, innovation, market predictions, political campaigns, targeted advertising, scientific research and influence elections.

Big Data also represents a threat, in addition to Harari’s arguments, with regard to the protection of fundamental rights such as: the right to privacy, data protection, data security, freedom of expression and non-discrimination.

Laws inevitably struggle to keep up with technology. What I want to highlight is the difficulty of the transparency requirement of the General Data Protection Regulation and ways to deal with the processing of individual data emphasizing psuedonymization.

The transparency requirements of Article 22 GDPR are: the right of the data subject not to be obliged to a decision based solely on automated processing, including profiling. In addition Articles 13(2)(f) and Article 14(2)(g) create an obligation to inform data subject about the logic involved as well as the significance and envisioned consequences.

The appropriate information to provide is complicated by two issues: companies claiming trade secrets regarding their data processing techniques and the incomprehensibility of machine learning to be explained in terms of human logic and rationality. The difficulty arises in providing an explanation for predictions based on AI algorithms created through machine learning fundamentally not expressible in human terms.

In important reminder is that GDPR applies to the processing of data subjects. Pseudonymization is a way to get around this. Pseudonymization is a de-identification technique that replaces the all important personally identifiable information fields within a record set with artificial identifiers that allow the continued use of the data. It can conceivably be this simple.

Elections in Bavaria

I’ve recently moved to Germany and will consider some of the events going on in and around Munich. Bavaria recently held elections for the Landtag, or Bavarian Parliament. Bavaria is the state that Munich is located.

The Christian Social Union (CSU) is sister party to Chancellor Merkel’s Christian Democrat Union and dominates the political scene in Bavaria ruling for all but three years since World War II. There was bit of a shake up this election cycle and while the CSU won 37.2% of votes, fringe parties won a large share reflecting divisions over topics like immigration and climate change.

As President Trump claims, immigration has shaken up Germany with Bavaria bearing the majority of immigrants that flooded the country. The results of the election signal a shifting perspective among German voters. While many still favor a liberal immigration policy, there has been a rise in anti-immigrant sentiment.

There are 205 seats in in the Bavarian Parliament and the numbers are broken down as such: 85 seats for the CSU, 38 seats for the Greens, a pro-immigration and pro-environment group, 27 seats for the Free Voters, 22 seats for the SPD who are the biggest losers from the last election while the anti-immigration Alternative for Germany won 22 seats rounded off by 11 seats for the FDP.

The outcome will likely have ramifications in Berlin, the seat of federal power, and it will be interesting to see how the CDU’s annual congress shakes out. The congress is held in December and we will see if Merkel, serving her fourth term, will be able to ward off a mutiny with her coalition.