David Isenberg and The Rise of The Stupid Network

Over one weekend in May of 1997 David Isenberg, who then worked at AT&T Labs Research (nee Bell labs), wrote a paper called The Rise of The Stupid Network which explained (and still explains) with breathtaking simplicity why the Internet is superior to the “intelligent networks” favored by traditional telcos and was about to crater the value of these expensive networks.  The paper is especially relevant now at a time when telcos and organizations like the ITU (a communications arm of the UN) are trying to use regulation to reconstitute the outmoded intelligent network regardless of the demonstrated success of the stupid network in doing everything the intelligent network was designed to do and much, much more at  a fraction of the cost (could that be their problem?)

I was in my last few months of being an officer at AT&T myself at the time and I unintentionally helped to shorten David’s career there as well.  A copy of the paper found its way to me (I was sort of the Internet guy among AT&T VPs) and I thought the paper was brilliant: it gave words and names and reasons to what us Internet enthusiasts sort of thought. In my soon-to-end quest to convince the rest of AT&T that we ought to go full speed ahead into VoIP, I forwarded David’s paper enthusiastically to other AT&T execs and told some reporters who asked me about it how right on it was.  AT&T released the paper onto the Internet, perhaps without fully understanding its import.

I left AT&T that summer to found VoIP company ITXC.  Hell, if AT&T and the telcos weren’t going to do what was obviously the future, what a great opportunity for a startup.  David’s paper and the consternation it was starting to raise in AT&T were certainly part of my decision. 

David hung on for a while but was increasingly muffled and become an independent “prosultant” and dynamite conference organizer in January of 1998.  He is still both of those things as well as a blogger (isen.com/blog). Shortly after he left, AT&T asked David to remove the paper from his website.  However, it is still available on other sites including the one I linked to above.  It is just about as current as the weekend it was written.  Probably just a few things David would change if he were rewriting it today.

OK, you ask, why is a stupid network better than an intelligent one?  In case you don’t have time to read David’s paper now (it is both short and approachable), I’ll give you the Readers Digest version.

The intelligent networks have lots of computers inside them. Those computers and their software were designed to support the kind of applications that the builders of the intelligent network thought everybody would want.  Not surprisingly at the time, they  thought that these were all applications built around voice and the routing of voice calls.  The intelligent network knows all about voice.  It amplifies it; it de-noises it (“you can hear a pin drop”); it supports all sorts of not very imaginative but once-lucrative voice services like 800 numbers, 900 numbers (mainly for porn), 500 numbers (don’t even ask; nobody wanted them).

Trouble is that you can’t build any applications for the intelligent network that the designers didn’t have in mind at the time they put the “intelligence” in.  Moreover YOU can’t build any applications for the intelligent network at all because the intelligence is all in phone company computers that YOU CAN’T TOUCH.  Nor can any other innovators except those who work for phone companies.  I’m sure there are some left but not many.

The Internet is stupid.  It’s made up of a bunch of routers that just know how to get packets where they're going cheaply and more or less reliably (and despite catastrophes like hurricanes, earthquakes, and terrorist attacks which give the intelligent network a huge headache).  It also has some DNS servers (technically at a higher level in the protocol stack) which translate human readable web and email addresses into routable IP addresses.  The brilliance of the stupid network is it doesn’t make any assumptions about what are in the packets.

So, since the stupid network doesn’t know or care what you’re up to, you can put your computers out at the edge of the stupid network and write almost any application you can dream of.  And, since you, the innovator, add the intelligence on your computer, there is no one to tell you “Hey, little boy or girl, leave that computer alone.  You can’t do that on OUR network.”

In David’s words:

“The Stupid Network would let you send mixed data types at will - limited only by the knowledge and imagination of the application programmer community. One way voice messages, multi-way voice conferences, two-way video, email, documents, audio and/or video entertainment, whatever, could be mixed and interspersed at will, within and between sessions. You would not have to ask your Stupid Network provider for any special network modifications - its only function would be to, ‘Deliver the Bits, Stupid.’”

If it seems obvious now, remember he wrote this in 1997.  And remember that the telcos still don’t “get it” – perhaps willfully.

The designers of the Internet way back over thirty years ago had no idea that it would be used for voice, email as we know it, and certainly not for the WorldWideWeb.  But the stupid Internet, fortunately, doesn’t know enough to get in the way of new applications.  It keeps getting cheaper and fatter bandwidth and new applications are developed to take advantage of fatter and cheaper bandwidth and are also built on top of old applications.

This is important now because the telcos in the US (and perhaps the cablecos) would like to use their monopoly (or duopoly) status and help they get from the now telco-friendly regulators at the FCC to take us back to the bad old days of the very limiting intelligent network.  That’s what all this noise about charging Google and Vonage extra for “using my pipes” from at&t CEO Ed Whitacre and others is really about.

You’ll can hear all about that in a conference David Isenberg and pulvermedia are having April 3 and 4 in Silver Spring, MD  called F2C: Freedom to Communicate.  It’s all about the many issues – many of them government issues – involved with NOT stifling innovation on the Internet narrowly and people’s ability to communicate broadly.  I won’t agree with everyone there; neither will you: but this is the place where the right communication issues get discussed by people who mostly know what they’re talking about.  If these issues are important to you, you might want to try to go.  You definitely want to hear some of what comes out of there.

I’ll be speaking on a panel  at 2:15 PM on April 3 – mainly to promote the petition Jeff Pulver and I filed with the FCC to try to mitigate the effects of the next disaster-caused telephony outage.  But, even if you’re tired of hearing me talk about that, you’ll meet lots of smart people trying to preserve the benefits of the stupid network – and you’ll get to see David Isenberg in action as a conference organizer and very clear thinker.

A related post on an F2C issue - Should the Internet be Lawless - is here.