Rezensionen und Kommentare

Ann Blair: Too much to know (2010, Yale University Press) 3 Sterne

The emergence of reference books and their differentiation into various genres of encyclopedic literature is much more fascinating than I would have thought. Anne Blair takes a very close look at (surprisingly large) collections of scholarly excerpts and notes in print during the 16th and 17th centuries but she also puts them in the context of the developments before (classical antiquity and middle ages) and after this period. This comes with many anecdotes about scholarly lives and practices, e.g. about a cosmology containing a little construction kit for readers to build their own model of the planetary system or a carpentry-based early platform for collaborative research. The book does not belie its character as a piece of scholarly work with meticulous discussions of various aspects of the matter resulting in much repetition. But it remains most readable and often entertaining. Working in research data management myself, the read has changed my reflections on contemporary developments and the hopes and worries related to them.

The emergence of reference books and their differentiation into various genres of encyclopedic literature is much more fascinating than I would have thought. Anne Blair takes a very close look at (surprisingly large) collections of scholarly excerpts and notes in print during the 16th and 17th centuries but she also puts them in the context of the developments before (classical antiquity and middle ages) and after this period. This comes with many anecdotes about scholarly lives and practices, e.g. about a cosmology containing a little construction kit for readers to build their own model of the planetary system or a carpentry-based early platform for collaborative research. The book does not belie its character as a piece of scholarly work with meticulous discussions of various aspects of the matter resulting in much repetition. But it remains most readable and often entertaining. Working in research data management myself, the read has changed my reflections on contemporary developments and the hopes and worries related to them.

Dieses kleine Buch erreicht sein selbst gestecktes Ziel, den Begriff "Künstliche Intelligenz" zu entzaubern, mühelos. Angefangen mit linearer Regression werden einige Methoden einfach und anschaulich vorgestellt. Dabei stellt sich heraus, dass Maschinen eigentlich ganz ähnlich aus Daten lernen wie Menschen. Die grobe Funktionsweise von KI wird anhand von Alltagsbeispielen einer Identifikationsfigur dargestellt. Zum Glück wird erst gar nicht versucht, diese Episoden lebensnah zu erzählen. Auf diese Weise bleibt der Rundgang durch die Methodik unterhaltsam.

Ein separater Teil widmet sich ethischen und gesellschaftlichen Implikationen. Auch diese Kapitel bleiben dicht an der technischen Umsetzung, was die großen Probleme notwendiger Weise ausklammert. Viele Aspekte werden lediglich angerissen und die Diskussionen greifen notwendiger Weise zu kurz. Allerdings bleibt dadurch das Konzept, sachliche Denkanstöße zu liefern, gewahrt.

Die Einführung wendet sich an die breite Öffentlichkeit ohne Vorkenntnisse. Als Einstieg in die Fachdiskussion ist sie weder gedacht noch geeignet. Das Buch ist eine gute Grundlage für bessere Diskussionen über sinnvolle und bedenkliche Einsatzgebiete künstlicher Intelligenz.