[extropy-chat] Automatin of information workers.

[Dan Clemmensen]

Dustin Wish with INDCO Networks wrote:

>I hope they every get to programmers....lol
>
>But I think the pointy heads in the upstairs offices will never want to or
>that can't program their VCR, much less database driven apps. Long life the
>Dilberts.
>  
>

>  
>
>Dan Clemmesnen wrote:
>
>It's all about the increasing ease of finding information. I believe 
>this trend will continue and that it is in fact at the heart of 
>acceleration toward the singularity.
>
>As software becomes more sophisticated, it will eventually replace more 
>and more of the information manipulation tasks now done by humans. 
>Telephone operators went first, then clerks, typists, and secretaries. 
>now ticket agents and travel agents are feeling the heat. Eventually we 
>will get to programmers and lawyers, but we have a way to go yet.
>
>  
>
You are quite correct. The PHBs are in fact being automated out of 
existence on many companies. There are far fewer paper-pushing middle 
managers per "individual contributor" than there were in 1980. That's 
because an individual middle manager can push more "paper" now and this 
increase in productivity is faster than the increase in paper per 
subordinate. This in turn has led to a gap in the traditional promotion 
trajectory whereby an individual contributor (i.e., worker bee) becomes 
a middle manager and finally moves into upper management. Different 
companies address the problem in different ways. I think that this 
phenomenon was a big hidden driver of the dot-com bubble, as senior 
techies with no road to promotion decided to start companies. The 
Darwinian selection of competent managers in a new-start is a lot 
harsher than under the old model.

The meta-lesson here is there is a lag in the adjustments that society, 
including businesses, make when relative efficiency changes. But 
exponential technical progress will cause the effect of the lag to 
increase. Ouch.

Of course I'm having fun spouting a theory here, but a serious theory 
should have predictive power and should make falsifiable assertions. 
I'll try that later. I spent long day today as an individual contributor.