[Tfug] [Bulk] Re: Stallman vs Ubuntu

Bexley Hall bexley401 at yahoo.com
Fri Dec 14 19:18:23 MST 2012


Hi Zack,

On 12/14/2012 5:59 PM, Zack Williams wrote:

>> I worked for an HMO about 15 years ago.  I built a model that determined
>> the risk of the groups that were utilizing under the different plans.  This
>> info was used by underwriting and sales and service to set pricing.
>
> What's funny is when this fails.  I have a male friend who happened to have
> a "conventionally female" name who recently started getting mailings for
> diapers and formula from a big-box retailer.

It's only funny "from the outside".

I *consistently* get pulled aside for a "pat down" when flying.
This is done solely based on *appearance* (since the folks
doing this haven't seen my boarding pass).  It's only a mild
inconvenience (and I agree wholeheartedly with the *intent*
even if the faults in the "model" are disheartening) but I could
imagine folks subjected to even more intrusive or inconvenient
interactions.

And, what do you do when models have significant *flaws* in
their suppositions?  E.g., if your parent(s) died of coronary
disease at a young age, then *you* are at increased risk for
a similar fate (think: not all children share their parents'
genetic makeup!)

Returning to my home automation issue...

If you can see child A's bedroom light going on at 1AM (long past
their nominal bedtime -- which you can deduce from watching when
the bedroom light went *off*, originally) followed by a bathroom
visit.  Then, the behavior repeating a few hours later (same child).
And, this happens routinely, you could deduce the child has a
greater chance of being diabetic (than children exhibiting "nominal"
sleep/bathroom patterns).

[This is pure fantasy on my part.  I'm just trying to imagine
how detailed knowledge of a person(s)' behavior can "leak"
information that one might not want leaked!]

<frown>




More information about the tfug mailing list