[IP] more on Fortune Fast Forward- on Gates & spam

[Dave Farber <dave@farber.net>]

From: Thomas Day Newbold

Hi David Kirkpatrick,
CC: Dave Farber, for IP list if you choose (if so, please remove my email
address. Thanks.).

Thanks for the report from Davos.  A fascinating range of discussion.  On
the question of spam, I think Gates is off base here.  A global adoption of
a universal email "bonding" scheme adopted at the consumer level seems
pretty far fetched (even as far out as two years from now!).  I can
understand his viewpoint, though: hey, we control [95%] of the operating
system market, and Outlook, Outlook Express, and Hotmail account for [80%]
of email clients, so we can just add this feature/service to
Windows/Outlook/Hotmail and it will be adopted.  The password scheme is
flawed for exactly the reason you point out.

The real battle in spam is in the economics.  Today, spammers make more
money from the response to spam than it costs to send the spam (mailing
lists, software, bandwidth cost money).  The victims of spam spend lots of
time (and therefore, money) identifying & deleting spam, which is an extra
cost to business/society.  Widespread adoption of an effective spam
filtering solution would re-balance the economic playing field by reducing
the response rate to spam (effectively raising the cost to send spam by
decreasing the economic returns that keep the spammers in business) and by
reducing the time cost for us dealing with spam.  The webmail providers and
AOL are already doing this at the server level (but at least with AOL, with
some false positive problems).

On my PC, I use a freeware spam filter product called SAProxy (built by a
company called Stata Labs which is now offering a commercial version
http://www.statalabs.com/products/saproxy/overview.php), which is an
implementation for Windows of an open source Bayesian spam-filtering package
called SpamAssassin.  The genius of SpamAssassin is that it is uses a
combination of a rules based filter (to which you can add custom rules- e.g.
dkirkpatrick in the subject=spam), with an adaptive learning filter that
learns from your spam and your own personal ham.  In my primary email
account, I get 25-100 spams a day (I know that's a pretty light load).  My
customized version of SAProxy catches them ALL (once or twice a week, one
might slip through).  It is really quite an amazing product.  I have not
tried the Pro version, since my old freeware is performing so well (not sure
the free version is still available).  I have no involvement with Stata
other than as a happy user of their software.

To my thinking, an adaptive, easily-adoptable client-side spam-filtering
solution could really put a dent in the economics of spam.

Best Regards,
Tom Newbold 

-------------------------------------
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/