[FoRK] Super-Intelligent Humans Are Coming

Stephen Williams sdw at lig.net
Fri Oct 24 12:02:27 PDT 2014


On 10/24/14, 6:09 AM, Dr. Ernie Prabhakar wrote:
>> On Oct 23, 2014, at 9:58 PM, J. Andrew Rogers <andrew at jarbox.org> wrote:
>>
>> Compounding this, humans can rationalize almost any action as “moral” if placed in the proper context. A super-intelligent human could be expected to be superb at constructing the necessary context. Normal humans are able to do this regularly.
> Yeah, that’s why it is not at all clear to me that having more super-intelligent humans is a big step forward in any meaningful dimension.  I’m more worried about making ordinary humans “smarter”, in the sense of giving them the cognitive and computational tools to understand new contexts more quickly.
>
> Haven’t you all seen: The Curse of Smart People?
> http://apenwarr.ca/log/?m=201407

Thanks for that link!
Excellent insight, up to a point.  But wrongheaded too: In my world view, studying a pretty wide sampling of the tech industry 
and various other industries, that working at Microsoft isn't that smart overall, which means the people working there have a 
certain blindness.  Or they are taking the best, most convenient offer without regard to the larger picture.  Certainly if you 
live in Redmond and need to maximize salary, you're going to lean toward rationalizing that it is your best choice.  And 
financially it has worked out for plenty of people.

He is clearly describing people with high IQ in certain areas, but low meta-intelligence and low EQ.  You've heard "A players 
hire A players and B players hire C players."?  I posit that "High EQ players try to hire the optimal mix of high IQ / high EQ 
while low EQ players somewhat randomly hire mostly low EQ players."  Most low EQ people are EQ blind, so low EQ hiring managers 
who tend to unknowingly offend or discomfort high EQ candidates and fail to put any value on their high EQ are not going to 
build a healthy organization.

Perhaps this describes the problems at Microsoft.  It fits what I know so far.

Part of being high EQ is being highly self-aware and detecting and understanding the real motivations of others.  Being high IQ 
allows you to convincingly argument in logical ways, pulling analogies, examples, and predictions from knowledge and experience. 
Meta-intelligence includes the ability to do resilient QA on your own logic and decisions.  OK, I want to make this gut decision 
on something.  Why?  What could be wrong with that?  Would that argument hold muster on FoRK?  Could I convince everyone that it 
is the correct decision?

I have argued points with many people in many situations over a lot of years.  Sometimes I was wrong, and I learned from it.  
Often, whether right or wrong, I was surprised by other people's reasoning.  Arguing a wrong point to conclusion is in fact a 
demand to be debug and learn what was wrong with the reasoning.  Just recently I had to confront and work through a problem with 
an engineer being obstinate about not changing an interface that someone else needed to be changed.  They had locked onto one 
principle as being the deciding factor and had an emotional stake in withstanding any other opinion.  Only by identifying the 
logic, boxing it in, and proving that it could be wrong for key requirements through heated discussion could I get this person 
to reconsider alternate ways of looking at the problem.  For a few days, another engineer said he felt like I was bullying him 
and he felt like quitting, then later after overcoming the impasse he seemed very happy about his new design.  Any culture of 
absolute command and control, avoiding all conflicts, equality of all opinions, and similar would have prevented getting this 
and other things right, leading to a quagmire.

I am really bummed that there was no debate team where I grew up (I had never heard of it).  And I'm bummed that I couldn't get 
my daughter interested as it seems very good general training.  I got to judge a few competitions though at Stanford which was 
interesting.

Impostor Syndrome is real.  I'm glad someone finally named and widely described it.  I suppose many of us wondered if others 
felt the same thing.  Repeatedly renewing myself and operating as a consultant for half of my life, I suppose I am used to the 
feeling in new situations.  Being self-taught, I probably had it worse for longer at times than most, but it spurred me to dig 
in.  As he indicates, it is not necessarily good to resolve the feeling early.

A different, seemingly opposite syndrome is when you meet a group who is very sure of themselves and their decisions, including 
what outside influence to ignore or even remain completely ignorant of, but where they are completely wrong, inefficient, and 
woefully behind the times.  You know you are right in what you are thinking because you have validation of the cutting edge of 
tech and you have plenty of direct experience.  But you're in an environment where the local peer pressure and sentiment is 
completely against you.  A local knowledge minima, trying to change your PH by clique clack.

A Cobol or VB programming shop a few years ago.  .NET development now.  Businesses with no modern IT understanding at all, 
virtual cargo cults.  People still using relational databases now (heh, got you on that one didn't I?).  You get the feeling you 
just stepped into the past.  You sort of want to help, but you are uncomfortable and know that there will be endless cajoling 
needed to drag people into the present who refuse to change and are adept at using bureaucracy.  Bringing up teenagers all over 
again.

Really, in so many ways, the recent burst of maturity and evolution of the web, browsers, mobile operating systems, servers, 
virtualization, languages, libraries, etc. is so nice.  I'm sure plenty are bewildered, but providing everything as services on 
the Internet shields all of those stuck people from even being able to decide against progress.  The Internet has routed around 
the damage once again.

>
> — Ernie P.

sdw



More information about the FoRK mailing list