Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- p.77: [..] we make no specific legislative proposals. Here are a few guiding principles:
- 1. The law should be solidly grounded in today’s technical realities. Simply trying to extend the concept of “dialed phone number” to the Internet doesn’t work. At the same time, it is crucial that the law not focus too closely on current technological paradigms. In the brief time in which this paper was written, notifications as communications went from an April Fool’s example—Yo!—to a serious set of products.
- DB: It's what I call the materialist mindset. I tried to explain the challenges a bit in 2001. http://thedartmouth.com/2001/10/03/computer-networks-changed-the-game/
- As of yet, I have no better general shorter term solution other than tackling point 2 first and having the mindset changes as the lessons sink in.
- 2. The consideration of the appropriate level of privacy protections that should be afforded to various kinds of communications information must account for the existence of “big data” analyses. Indeed, the momentum and analytical capacities driven by big data is changing even faster than technology in general.320 While the law does not generally regulate how information is analyzed once lawfully collected, the revelatory insights afforded by big data should give rise to new and stronger privacy considerations for non-content.
- DB: IMHO, there has to be a concerted push to socialize the inference models (graph-based) used in Big Data analytics to the staffers of policy makers in particular. The goal here would be to convey a solid visceral intuition of the adage "If it looks, quacks, hangs out with Ducks, has Duck genome, ducky friends, etc - > it's likely a duck".
- Specifically, they should be made to realize that with data analytic advances, it is now straightforward to identify criteria that mediately & predominantly target any given cohort in the US. This means that any type of discrimination (good, bad, be it Affirmative Action, economic stimulus, health care allocations etc) can be surreptitiously encoded in innocuous measures that rely on seemingly neutral, objective criteria. The intended discrimination and decision ‘unwrapping’ is hidden in the correlations of the stipulated criteria and can not be easily ascertained by end result inspection or even black box algorithm inspection. This was of course tried manually before, but now we have the data and the analysis power to do this at scale, systematically and 1, 2, 3, nth orders etc orders removed.
- Lastly, the probable outcomes of chosen courses of actions (legal, regulatory etc) could be demonstrated visually by eg monte carlo simulation.
Add Comment
Please, Sign In to add comment
