Felix Stalder on Sat, 9 Jun 2018 11:28:01 +0200 (CEST) |
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
Re: <nettime> Complexity, epistemology and power in the digital condition |
Hi Brian, sorry for the late reply. I think you are right that there are different configurations of how science done and how its value is established, based on how it's integrated into the state/commercial apparatus. > Then after the world-shaping ethical-practical development of the > disciplines and professions in the late nineteenth century, that > epistemological period collapsed into wat and a whole new figure of the > state emerged in mid-20C. It was able to synthesize the different powers > of knowledge into programmatic collaboration over continental spaces and > multi-year intervals, particularly due to the mental discipline > cultivated in places like academia and government research. You know how > they thought, that was our era, it was yesterday. The most recent social > form - the hypermodernizing technocratic state - was able to bring the > scientists, the professions and the military into unison with the > corporations, and by so doing it ignited the Great Acceleration around > the world. Yes, and this is both an acceleration of trends and a break with it. At the core of this, enabling what you call the Great Acceleration and which lead me to this great chart http://www.igbp.net/images/18.950c2fa1495db7081ebd1/1421396650502/GreatAcceleration2015igbpsrclowres.jpg is machine learning. Epistemologically, this is a real break with how we come to know something about the world, yet, it is put into service the sustain growth, necessary for capitalist expansion. In terms of epistemology, you look at the major (postmodern) critiques of science -- von Förster's insistence that the observer is part of the system (1972), Lyotard's turn from truth to instrumentality (1979) and Latour's argument that the natural and the social cannot be separated (1988) -- these are all more valid today than they were back then, and they all apply to machine learning to the extreme. Here, facts are not discovered but made, not from the outside, but the inside of the problem, and the focus is not on truth but on utility in terms of achieving greater ability to manipulate data-streams and the connected physical dynamics over relatively short time, thus pushing the system into another place where it can be observed again and the whole cycle starts again. At least in some instances of this, the process of knowledge creating is truly random and directionless. The method, enabled by cheap computing power, is simply to exhaust the problem space, trying out every possible combination and then see what works. Then, when something works, build a new problem space, exhaust it again through random methods find something that works etc. Layer, by layer. This is how depth comes into deep learning. Any solution is fine, as long as it "solves" a predetermined problem in the short term. All post modern critiques against industrial science are still valid from the point of view of modern science, but but now they are the new method of machine learning. Of course, the rub against the old-established methods of science (objectivity, distance, causality) and this creates serious confusion. I think something similar is happening in the political space, as you point out. Here its the nation state against, I would say, global (or at least transnational) protocols, which are, in a way, a directionless form of social governance which is neither disciplinary nor operating on the level of the imagination. It's pure instrumentality. All the best. Felix ||||||||||||||||||||||||||||||||| http://felix.openflows.com |OPEN PGP: https://pgp.mit.edu/pks/lookup?search=0x0C9FF2AC
Attachment:
signature.asc
Description: OpenPGP digital signature
# distributed via <nettime>: no commercial use without permission # <nettime> is a moderated mailing list for net criticism, # collaborative text filtering and cultural politics of the nets # more info: http://mx.kein.org/mailman/listinfo/nettime-l # archive: http://www.nettime.org contact: nettime@kein.org # @nettime_bot tweets mail w/ sender unless #ANON is in Subject: