« VoIP Security Trends for 2007 | Main | Coverage of Top Ten Predictions »

January 30, 2007

Comments

David Schwartz

I couldn't agree more. All this talk of Turing tests makes me ill. Besides the annoyance factor, how do you deal with languages? Will this become like calling my bank? "For English press 1 - para Espaneol 2"... Give me a break. I just want to talk to my friend. I too do not believe this will happen.

On a more fundamental level, however I have my doubts about many of these systems. Forget the Turing test for a moment, how do they differentiate "valid" spam (e.g. my local book club meeting) from "invalid" spam? Unless I am required to manually "white" and "black" list it won't work and the complexity of "manually" anything will force this approach to follow in the footsteps of the dinosaures. In addition, for fear of "false positives" most people will disable the feature altogether(just look at how many folks disable personal firewalls due to similar problems).

The only viable solution IMHO is using bidirectional pattern analysis coupled with a social networking framework. (I just happen to working on such an system now ;) )

D.

The comments to this entry are closed.

My Photo

Search Blog

  • Search Blog
    Google

    WWW
    voipsecurityblog.typepad.com

Become a Fan

Telephony Denial of Service (TDoS)