Название: The Democratization of Artificial Intelligence: Net Politics in the Era of Learning Algorithms
Автор: Andreas Sudmann
Издательство: Transcript Publishing
Год: 2020
Страниц: 332
Язык: английский
Формат: pdf (true)
Размер: 10.1 MB

The recent explosion in interest in both the democratization of data and artificial intelligence (AI) brings important questions to the fore: How democratic is AI? How to explore the political aspects of modern machine learning if many computer scientists consider AI technologies a black box fundamentally opaque to human understanding? And how does this affect questions of accountability and political agency?

This timely collection discusses the complex political dimensions of internet and AI technologies, seeking to elucidate possible avenues of democratizing AI. This includes technological strategies such as an 'Explainable AI', but also the questions of political and/or critical concepts which guide the technological process of making modern AI technology more accessible as well as suitable ways of democratizing AI beyond a purely abstract understanding of transparency and accountability.

When people talk about AI these days, their focus is mostly on so-called Machine Learning techniques and especially artificial neural networks (ANN). In fact, one can even say that these accounts are at the very center of the current AI renaissance. Sometimes, both terms are used synonymously, but that is simply wrong. Machine Learning is an umbrella term for different forms of algorithms in AI that allow computer systems to analyze and learn statistical patterns in complex data structures in order to predictfor a certain input x the corresponding outcome y, without being explicitly programmed for this task. ANN, in turn, are a specific, but very effective approach of machine learning, loosely inspired by biological neural networks and essentially characterized by the following features:

1. the massive parallelism of how information is processed/simulated through the network of artificial neurons
2. the hierarchical division of the information processing, structured in learning simple patterns to increasingly complex ones, related to a f lexible number of so-called hidden layers of a network
3. the ability of the systems to achieve a defined learning goal quasi-automatically by successive self-optimization (by means of a learning algorithm called “backpropagation”)