Hacker News Reader

Comment by dredmorbius | Hacker News Reader

Have you made any attempt to determine how effective such "personal responsibility" methods are at actually avoiding surveillance, and/or do you have any thoughts on how you might go about finding out? A follow-up on this topic, perhaps interviews with people who'd know, might be of interest.

I do like the piece, as a demonstration of the difficulty / cost / inconvenience of actually opting out of pervasive surveillance. As a work of performance art your blog post is excellent. (A point several people commenting in this thread seem to have spectacularly missed.)

View on HN

Thanks! You totally grokked what I'm doing with this--I appreciate it!

Yes I do think a lot about the effectiveness of personal responsibility, and its limits. Certainly I could request my logs and data from every company, although that takes more time than opting out. In the meanwhile, I don't block ads as those as my only clues as to what a corporate database thinks it knows about me (obviously I use blockers to stop info flow from me to them, it's just that I want to see what they think they know). I also use lots of addresses and CC numbers so I can trace data migration and can identify the culprit if so ( https://www.optoutproject.net/what-email-address-should-i-us...)

As for if it's working... There are more quantitative ways to tell for sure. In the meanwhile, my Pregnancy Experiment started in 2013 and in the intervening ten years I have seen zero online ads for anything baby- or child-related, and I have only received an unsolicited catalog in the mail every three years (thanks to the above technique, though, I always knew who the weak link was). When I tell other moms that, their jaws drop.

Logs are one option, I've been increasingly leaning toward various canaries / telltales myself, which ... it looks as if you're doing through email at last in part, and credit card numbers, which would of course have been my next suggestion. Variant spellings of your name, or if you're ambitious, different postal mailing addresses, would be another option.

(Curiously, a lot of the methods for tracking surveillance ... are also used by marketing entities to track campaign effectiveness and such.)

On top of these, unused and obscured URLs might also be used.

I'd especially like to see what dot-connecting capabilities various entities have, and am thinking of how these might be tested, e.g., by having two not-directly-linked tokens used and seeing whether or not there's plausibly been a connection made between the two.

I'm realising that there's actually a substantial site and project behind your work, I'm going to take a look at that before I say too many more stupid things ;-)

NB: As one who read economics at Uni, I'm developing a much-belated mad respect for sociology.

Yes, I regularly use unusual spellings, different postal addresses, pseudonyms, CC's, the whole nine yards.

Nod.

If you happen to know of any good references / sources on methods --- for both individuals and organisations --- or even just the right keywords to use for researching the literature, I'd appreciate it. That's information which could stand to be more widely known.

(My search-fu is usually pretty good. There are times when I'm stymied simply not knowing the language / terminology used within research / technology areas.)

Reply on news.ycombinator.com