Massively parallel carbon-unit-based voice pattern matching

[Ernest Hua <hua@chromatic.com>]

> > "Why the hell, when you have a tape of the perp. saying something,
> > do you not air it in the middle of a city chock full of people
> > who could very well know this guy's voice/mannerisms?" Why didn't
>
> This is the single best suggestion/question I have heard about this
> entire case!

It may not be as good of a suggestion as you might think.  If the
number of false positives are so great that there is no reasonable
way for the FBI to go after everyone of them (that is, say, they
could not accurately weigh the likelihood of any particular siting
as having a reasonable chance of being worthy of investigation),
then such a act is worthless.

Also, the FBI, no doubt, has voice analysis systems much more
capable of detecting subtleties.  100,000 enthusiastic people
trying to recognize their friend/relative/co-worker's voice is just
not consistently accurate enough to give the FBI good leads while
avoiding bad ones.

Now, that said, I suppose the FBI could try something like this:

    Have some randomizing automated test which everyone can call
    into to test their hearing abilities before they can tell the
    FBI their possible lead.  Well, I take that back.  May you test
    them AFTER they leave info on their lead so that they don't get
    offended by the testing.

Well ... I guess I'll leave the exercise up to the FBI (to make the
phone-in more user-friendly).

Now what if the FBI had voice detection systems just like the ones
used in "Clear and Present Danger"?  Now THERE'S a good reason for
tapping 1% of every phone switch ...

I wonder if the public would mind having a MACHINE do drift net
matching on voices on all public phone switches in the U.S. just to
find the Olympic bomber ...  What if the specifications of that
machine were made public and the machine placed under Congressional
oversight?  (At least members of Congress would never get tapped.)

Ern