IP: An interesting viewpoint ..on the net and peronal interactions (was on Hellmouth)

[Dave Farber <farber () cis upenn edu>]

> From: shapj () us ibm com
> X-Lotus-FromDomain: IBMUS
> To: farber () cis upenn edu
> Date: Wed, 28 Apr 1999 11:35:03 -0400
>
> > That is what is missing from families. the basics of respect and caring
> > for/about others. For if the parents don't care about kids, where do kids
> > learn to care about others?
>
> I agree completely.  I'ld only note that the phenomenon is not limited to
> families.
>
> Also, I wonder sometimes whether technology helps or hinders this.
>
> A somewhat disjointed anecdote from my own experiences as a contributor to
> Netnews:
>
> Just about everybody now reads newsgroups in one form or another.  You see
> netnews content quoted in this list and in newspapers. A long time ago, I
> contributed to this project.
>
> After early deployment at Bell Labs and a few other places the original netnews
> software spread quickly.  People installed new phone lines to get their feeds.
> The first 1200 baud modems at such sites invariably went to the netnews lines
> because the traffic took *hours* to move (and was tiny compared to today).  At
> one point in 1981, a machine at Bell Labs was running up $250,000+ in phone
> bills per MONTH moving this traffic.  In retrospect it is amazing how many
> companies footed bills that large for something that provided no measurable
> commercial benefit.
>
> In this day of high speed backbones, fast modems, and ubiquitous ISPs, It's hard
> to imagine (even for me, and I was there) how that system felt.  Conversations
> would stretch over weeks because the messages had to go from one machine to the
> next, each step involving a several hour delay.  Material to Australia went by
> air freight for a while -- there were no 10 cent a minute international phone
> calls then.  Sort of like wrapping paper around a rock and using a ballista to
> talk to your neighbor. [Air mail for cave men?]
>
> But the slow speed had a moderating influence.  If you typed something in anger
> it would be weeks before the victim saw it.  Since the response time was so slow
> it just wasn't very satisfying to be obnoxious.  A few people were anyway.  As
> the message transfer time got faster this changed, and the flame rate rose.  As
> the price went down, more and more people got on who weren't part of the
> original techie community, and these people brought their own way of doing
> things.
>
> Needless to say, netnews was a runaway success.  The problem was that it wasn't
> scaling.  As more people used it, more traffic got delayed or lost, and the
> phone bills went through the roof.  It got to the point where several of the
> backbone sites (including, ultimately, that one at Bell Labs) would get shut
> down due to costs and the whole exercise would collapse. Administrators started
> selecting which groups their site would carry and filtering the rest just
> because of the cost. This led to the features that make netnews so hard to
> censor today.  It also led to the B-news project.
>
> The B-news project marks the second (maybe the third, depending on who's
> talking)  "phase" of netnews development. It was the first time (to my
> knowledge) that a large piece of software was successfully developed by a group
> of geographically distributed collaborators under the open source model (it
> wasn't called that at the time).  Most did not meet each other until years
> later.
>
> There was a sense of mission about B-news. To some of us, the goal was to
> perform a social experiment: could we build a tool that would allow the whole
> world to take part in discussions? Nobody really appreciated the degree to which
> the impersonal nature of the medium would bring out the worst in the
> participants.  We knew that the nerds would write flames (obnoxious messages,
> often containing personal attacks).  We didn't really pay attention, because
> nerds flame in person too.
>
>
> Fast forward 17 years...
>
>
> Netnews is alive and well. Nowadays messages move almost instantly.  Flame wars
> are constant, and can go on for months.
>
> In a fit of annoyance one day I set out to compute how many dollars are lost to
> this behavior per year.  There are 400 million users we can identify, Figure 10
> minutes per working day (low), so a total of  1,000,000,000,000 minutes of
> reading time per year world-wide.  Assume that 1 message in 25 is a flame (this
> is also low), so 40,000,000,000 minutes per year are spent reading them, and
> that the average reader makes $30,000 per year, or $0.25/minute.  This works out
> to $10,000,000,000 of time wasted per year.  It's a simplistic calculation, and
> it isn't all of the story -- there is lots of good stuff too -- but that's a BIG
> number.
>
>
> Mostly, we forgot that the computer is an impersonal medium.  You can say
> anything you want, and there are very few consequences. Under these
> circumstances, people say all sorts of things they would never say to your face.
>
>
> Netnews was a social experiment.  Like Robert Morris's internet worm, it got out
> of the lab and got away from us. Today, most computer literate children in the
> world have used this software and been exposed to its attributes, both good and
> bad. In the end, I'm less concerned about the $10,000,000,000 per year than I am
> about the lessons those kids bring back to real life from their interactions
> with our software, and more broadly from interactions with the computer in
> general.
>
> People say "don't worry," but most of them don't understand the cognitive
> processing that young children do.  Children don't have the tools to
> discriminate between good content and bad content as adults can -- heck, a lot
> of *adults* can't do this.  But they do learn.  Efficiently, and not always what
> we want them to.
>
>
> As a result of the netnews experience, I've learned first hand that technology
> has an enormous impact on human behavior, and not always for the better.  It's
> not the games or the pictures.  It's the nature of the medium.  Video and voice
> are starting to put a personal element back into the picture, which is
> important.  As with the original netnews system, the delaying factor is cost.
>
> Lots of forces in the world work to help us objectify each other, and the more
> we see people as objects the easier violence becomes.  Advertising does it
> deliberately.  Netnews did it by accident (and with balancing benefits). I am
> left to wonder what impact on people's behavior our work has had.  Perhaps more
> of us, as techologists, need to wonder about such things.
>
>
> Jonathan S. Shapiro
> IBM T.J. Watson Research Center
> Email: shapj () us ibm com
> Phone: +1 914 784 7085  (Tieline: 863)
> Fax: +1 914 784 7595