Talk on the privacy bargain, big data, and human sensors versus human barcodes

Here's the video from the talk I gave last week at the O'Reilly Strata conference on "big data" in NYC. The talk is called "Designing for Human Sensors, Not Human Barcodes," and it talks about the philosophy underpinning the "privacy bargain" we strike online when we trade personal information for access to services.


    1. Renny: His glasses are made of wood! WOOD!

      Great talk as usual, Cory. I’ve followed pretty much all your iterations through this topic and you’re getting better (pithier, more succinct, more impactful, more affective) every time.

      Incidentally, in the UK we have something of an antidote to the dead-kid anecdote called the Baby Mailing Preference Service – like the normal, general-purpose MPS, it lets you opt-out of baby-related mailings, and any sensible marketer will pay attention to these preferences so as not to upset the bereaved. It’s not a perfect or even a great solution by any means, but it’s a start.

  1. cory, i wonder do you have any thoughts on the new privacy bargain facebook is proposing with frictionless sharing and the implications sharing everything automatically has for the human persona?

  2. Really excellent work.  Any recs for other recent talks on the subject?  The tech and social aspects are moving SO fast.

    Edit to add…

    It’s such a relief to hear some technologically savvy ways to empower the users. Usually we feel pretty swamped by the resources of the money machines. The creation and monitoring of a well-lighted and honest marketplace in which we make these transactions seemed like a hopeless dream but you give it a real chance of working. Having our own browsers refuse transactions that don’t take place in that honest marketplace, and having ways to eject dishonest brokers from that market just may give us a chance to take back some ground in the perpetual fight.

    And the browser agent that spoofs the information? Awesome. An additional possibility would be to choose specific patterns to the spoofed data to achieve unrelated goals in aggregate with millions of other users of the app.

  3. Excellent talk, Cory. Ironically, I can no longer “like” BB items without logging into a social media site. Oh well, thanks for a great lunch break.

    FWIW, I’ll be sharing this with several analytics developers much smarter than myself to see what suggestions they might have.

    For my part, I think a user-defined privacy scheme that worked on at least three layers of detail would appeal well beyond technically savvy circles. Imagine an agent that ran a wizard asking a few simple questions to establish what data the user wanted to release. Below that, more customizable preferences could let the user see what the wizard learned and tweak it in more detail. An advanced preferences dialogue could get down to the nitty gritty of which sites should be trusted, which sites should have to ask permission, with the remainder handled by the user agent.

  4.  Ghostery is the closest thing I’ve seen to the cookie management that
    Cory suggests. It essentially does what he wants, by blocking cookies
    from hundreds of known data miners, but not legitimate sites that you
    want to be signed in to (and you can selectively allow advertisers if
    you like).

Comments are closed.