October 13, 2019 fkirschner

The new academic year has started last week. As is custom, here is our weekly digest of noteworthy things we would like to share this weekend:

Part of the fallout of the horrible attack in Halle is a re-emerging discussion of radicalization in online communities and through streaming media.

Yasmina Banaszczuk wrote a well argued piece (in german) in taz.de:

” Der Täter von Halle versuchte offensichtlich, in seinen Notizen zur Tat ein ähnliches Element zu schaffen, dachte sich verschiedene Achievements aus, die hier nicht wiederholt werden müssen. Die Gamifizierung von Hass und Terror ist ein dankbarer Mechanismus, um Menschen an rechte Gedanken und schließlich an Taten heranzuführen. Es ist jedoch vor allem eins: Teil einer rechten Propagandamaschine. “

Hartware Medienkunstverein recently ran an excellent exhibition on the qualities and effort that goes into fascist and alt-right online communities. Here’s an interview with the curator Inke Arns on bento.de:

” Wir brauchen neben politischen Antworten auch gesellschaftliche. Es gibt ja schon Initiativen wie #Ichbinhier, “Hass hilft” oder Reconquista Internet von Jan Böhmermann. Davon brauchen wir noch mehr. Am Ende geht es schließlich darum, wem wir das Netz überlassen. Wenn wir sagen, dass es längst zu unserem Alltag gehört, sollten wir es auch verteidigen. “

No idea how to best switch topics. But the AI Now Institute has published its year in review and it is an excellent overview of all the governance and communal happenings around AI this past year:

” There has also been wider use of affect recognition, a subset of facial recognition, which claims to ‘read’ our inner emotions by interpreting the micro-expressions on our face. As psychologist Lisa Feldman Barret showed in an extensive survey paper, this type of AI phrenology has no reliable scientific foundation. But it’s already being used in classrooms and job interviews — often without people’s knowledge. “

In case you want to know more about computational neural networks, how they work technically, why they are not really neural nor really networks, and why training data matters so much, here is a series of videos that explain really well what they are and how they work mathematically:


By loading the video, you agree to YouTube's privacy policy.
Learn more

Load video

And to close this first weekend reading of the new year, here’s Nathaniel Comfort in Nature.com talking about How science has shifted our sense of identity:

” Across the arc of the past 150 years, we can see both science and scientism shaping human identity in many ways. Developmental psychology zeroed in on the intellect, leading to the transformation of IQ (intelligence quotient) from an educational tool into a weapon of social control. Immunology redefined the ‘self’ in terms of ‘non-self’. Information theory provided fresh metaphors that recast identity as residing in a text or a wiring diagram. “