3 ms·
The Panopticon Singularity (2002)
- notahacker 9y agoHow well do these predictions stack up? DRM obviously existed to an extent at the time, and now millions of people use locked down mobile OSes as their everyday computer, but to an extent some of its more intrusive practices have been rendered obsolete by the software provider taking total control of how, when and to whom software/content is served via the SaaS model. Gait analysis is one which obviously has thus far failed to be anything more than an unreliable supplementary information source with dreadful test performance even amongst small samples not trying to disguise their walk[1], it's not widely used outside the clinical context of helping patients to walk, and given how much other surveillance technologies have improved it's difficult to imagine it ever being that important. Is anyone actually doing anything with "celldar"? [1]see e.g. https://royalsociety.org/~/media/about-us/programmes/science-and-law/royal-society-forensic-gait-analysis-primer-for-courts.pdf https://royalsociety.org/~/media/about-us/programmes/science... is a discussion recommending scepticism about its use as
- jonathankoren 9y agoI think looking at this from specific technology predictions is a bit myopic. Gait analysis is/was simply a technology for identifying and tracking individuals. The fact that facial recognition won out[0] for this task is irrelevant. The more important issue is pervasive mass surveillance that can be leveraged against you, not just by the state but also by private parties that can be leveraged by the state. [0] https://www.theverge.com/2018/3/12/17110636/china-police-facial-recognition-sunglasses-surveillance https://www.theverge.com/2018/3/12/17110636/china-police-fac...
- denom 9y agoAnd as Cambridge Analytica has shown, mass surveillance can be leveraged against you by private parties for a pittance. There are now vast numbers of small private entities dabbling in what was the exclusive domain of the state.
- jacquesm 9y ago> Is anyone actually doing anything with "celldar"? We can do much better than that now: http://www.mdpi.com/1424-8220/15/12/29896/pdf http://www.mdpi.com/1424-8220/15/12/29896/pdf