Tech and Politics

In this episode, we focus on three further contact areas between AI and democracy at different analytical levels:
At the group level, AI impacts equality of rights among different groups of people in society.
At the institutional level, AI impacts the perception of elections as a fair and open mechanism for channeling and managing political conflict.
At the systems level, AI impacts competition between democratic and autocratic systems of government.

Skript: http://digitalmedia.andreasjungherr.de/ai.html#artificial-intelligence-and-equality

One tenet of democracy is that governments should be chosen by those they will serve. Such self-rule is a normative idea about legitimizing the temporal power of rulers over the ruled and a practical idea that distributed decision making is superior to other more centralized forms of decision making or rule by experts. AI impacts both the ability of people to achieve self-rule and the perceived superiority of distributed decision making over expert rule in complex social systems, highlighting potential limits to self-rule in several ways.

Skript: http://digitalmedia.andreasjungherr.de/ai.html#artificial-intelligence-and-self-rule

The successful application of artificial intelligence depends on a set of preconditions. Some are obvious. For example, to be successful AI needs to be able to access some digital representation of its environment, either through sensors mapping the world or through the input of existing data. Where these representations are difficult to come by or data are scarce, as in many areas of politics, AI will not be successful. Other preconditions are not so obvious. For example, for AI to produce helpful results, the underlying connections between inputs and outputs must be stable over time. This points to two problems: unobserved temporal shifts between variables and the dangers of relying on purely correlative evidence without support of causal models.
More important still, especially with respect to democracy, is that normatively speaking the past must provide a useful template for the future. Change is a crucial feature of societies, especially the extension of rights and the participation of previously excluded groups. Over time, many societies strive to decrease discrimination and increase equality. In fact, many policies are consciously designed to break with past patterns of discrimination. AI-based predictions and classifications based on past patterns risk replicating systemic inequalities and even structural discrimination.
Few problems in politics and in democracy more broadly share these characteristics. This limits the application of AI in society and, accordingly, its impact on democracy.

Skript: http://digitalmedia.andreasjungherr.de/ai.html#conditions-for-the-successful-application-of-artificial-intelligence

AI has become a pervasive presence in society. Recent technological advances have allowed for broad deployment of AI-based systems in many different areas of social, economic, and political life. In the process, AI has had - or is expected to have - a deep effect on each area it touches. We see examples in discussions about algorithmic shaping of digital communication environments and the associated deterioration of political discourse; the flooding of the public arena with false or misleading information enabled by generative AI; the future of work and AI’s role in replacement of jobs and related automation-driven unemployment; and AI’s impact on shifting the competitive balance between autocracies and democracies. With these developments, AI has also begun to touch on the very idea and practice of democracy.

Skript: http://digitalmedia.andreasjungherr.de/ai.html

Algorithms are used in ever more societal areas. This raises broad concerns. While in principle algorithms provide a set of clearly defined steps to solve a given problem, their current uses have raised the question whether this is still the case. In the following sections, we will focus on concerns about algorithms trapping people in algorithmically constructed bubbles and loops, the alignment problem, and opaqueness of algorithmic decision making and its consequences.

But these worries should not distract us and blind us with fear. The use of algorithms in society is not optional. So these issues should point us to improving algorithms and their use, not entertain the illusion of somehow avoiding their use. This includes conscious and informed approaches to mechanism design underlying the algorithm and algorithm auditing and forensics.

Script: http://digitalmedia.andreasjungherr.de/algorithms.html#bubbles-and-loops

Algorithms need not stop at providing insight or advice, some also can take automated action. These algorithms are designed to process data, make decisions, and execute actions without requiring human intervention at every step.

Beyond the uses of algorithms, we also need to address risks and fears. One risk is algorithmically enabled unfairness:

Once algorithms start shaping people's option spaces, the question of fairness emerges. Algorithms make, or at least support, decisions about people spanning various areas of their life: they assign people's credit ratings, they evaluate their job applications, they assess the likelihood of them engaging in criminal activity, or administer welfare benefits. These algorithmic assessments and decisions matter for the choices people have and the way they are treated by institutions of authority. This makes it important that algorithms are treating people fairly.

Script: http://digitalmedia.andreasjungherr.de/algorithms.html#algorithmic-action

Algorithms are also used to provide suggestions and advice. This includes algorithms that suggest courses of action or scenarios about future developments based on regularities identified in past data. Examples include algorithms that recommend users of digital media platforms content of potential interest, algorithms that advice police forces on where to expect a concentration of criminal activity, or algorithms that advice doctors on whether specific symptoms indicate a specific disease. Algorithms play a crucial role in decision-making processes across various sectors, helping individuals and organizations to make informed choices by offering data-driven suggestions.
In discussing these algorithms, their uses, effects, and evaluation, it is helpful to differentiate between algorithms providing advice and suggestions to experts and those providing advice and suggestions for lay people and users of digital communication environments and services.

Script: http://digitalmedia.andreasjungherr.de/algorithms.html#algorithmic-support

The uses of algorithms vary widely. Algorithms can identify patterns in data, surfacing hidden connections between phenomena or actions. Similarly, algorithms can also be used to automate actions. Based on rules learned from past data, algorithms can automatically present people with specific options.

Algorithms thus help us understand the world better and interact with it more efficiently. But algorithms are also a source of worry. How can we be sure that patterns identified from data documenting the past should be replicated in the present or future? How can we know what the algorithm learns from data? And how can we be sure that the actions taken by algorithms conform with the goals, with which we have designed and deployed them?

Script: http://digitalmedia.andreasjungherr.de/algorithms.html

The natural corollary to quantification and data-driven insight and capabilities is privacy. The simple formula is the more data, the greater the potential analytical insights, the greater the capacity of organizations, companies, or states to make profits, shape people's option spaces, or the future. People's interests and rights in keeping aspects of their lives, character traits, interests, and behavior private can appear as an annoying speed-bump on the road to greater capacity and profits. This is especially relevant in the collection and use of data documenting people's uses of digital devices or services.
New capabilities in data collection, retention, and their uses raise hopes and desires within companies and governments for access to ever more data on ever more users on ever more aspects of the world. But here, interests between companies, governments, and users do not necessarily align and might in fact diverge. Greater capacity of companies and governments might run counter to the interests and rights of people who find themselves documented in data.

Script: http://digitalmedia.andreasjungherr.de/data.html#privacy

Data provide a reduced representation of the world and promise to uncover hidden patterns and connections between entities that remain invisible in their full, unreduced manifestation in the world. This allows people and organizations to make sense of the world, and increase their level of control. But data do not only allow people and organizations to see the world, they also allow them to shape the future. Building mathematical models of the world, provides new ways to form expectations about relevant future developments as well as effects of specific actions and interventions. Having access to better models about the world allows people and organizations to position themselves with greater foresight in the world regarding future developments and effects of their actions. This allows them to outperform the competition and shape the future according to their interests.

Script: http://digitalmedia.andreasjungherr.de/data.html#observing-the-world-through-data