Generational Dynamics World View

[Warren Dew]

(05-11-2020, 09:27 PM)John J. Xenakis Wrote:

(05-10-2020, 02:34 PM)Warren Dew Wrote: >   Exactly.

>   I actually made a living for three years making true predictions.
>   They weren't generational predictions, but they were system
>   dynamics predictions about the behavior of large groups of people.

>   It was really tough.  What I learned was that people are far more
>   willing to pay for predictions that what they want to happen will
>   happen, than for predictions that are true.

>   I think this is just how the oracle business works, going back at
>   least to Cassandra.

I've never heard anyone say that they've been making predictions based
on system dynamics, besides myself.  What kinds of predictions were
they?

And you're right about people hating predictions that are true, rather
than what they want to hear.  It freaks people out, and turns some of
them into loudmouth assholes, because of cognitive dissonance.  The
ancient Greeks were brilliant in coming up with the Cassandra myth as
something that describes humanity.

I did the work in 1992-1995.  It was work predicting the future of the telecom industry.  We would put together system dynamics computer models, and sell them to various players in the telecom industry, mostly to the baby bells.

The first such model was a bit primitive, but we eventually developed a pretty sophisticated model which we sold to a couple of companies.  Our most sophisticated model handled the interplay between consumer preferences, infrastructure deployment, and technological improvement, and predicted uptake of wireless services - at the time, remember, few people had wireless phones - along with associated financials.

Customers put in their own assumptions, but we also kept an in house scenario that had my own best guess assumptions in it.  Those guesses were pretty good, because some of our customers shared a lot of data while we were developing the model.  We had lots of data from which we could back out a demand curve for communications services, and lots of data on how long it would take to deploy infrastructure capacity and how much it then currently cost.  Technological improvement was handled by a module that had been proved out in an earlier model in the petrochemical industry.

That model correctly predicted the now historical gradual uptake of wireless phones as a supplement to wireline phones until about 2003, at which point it correctly predicted that a noticeable number of customers would start dropping wireline services in favor of a wireless phone only.  It basically made an accurate year by year prediction for 10-15 years into the future.

By that time, of course, I was long out of the business.  Nobody wanted a true prediction that their favorite project would lose money or barely make money when there were plenty of consultants that would happily predict the project would make unrealistic scads of money, thus justifying their getting corporate headquarters to waste tons of funding on it.