[MLUG - DISCUSSION] from prodigy to genius?

[Jonathan King <EMAIL:PROTECTED>]

On Tue, 2 Sep 2003, Mike Miller wrote:

> On Tue, 2 Sep 2003, Michael wrote:
> 
> > > Sure, they do all sorts of great things.  Almost no one ever achieved
> > > as much as quickly as the kid we were just discussing.  We'll be
> > > following his story.
> >
> > Examples?
> 
> Dean Kaman
> Eric S. Lander
> W. French Anderson
> 
> come to mind.  all were amazing as teenagers.

OK, so one possible confound in thinking about this kind of thing
(is being a prodigy a good predictor of adult genius) is that
prodigious early achievement might be a very strong predictor, but
most geniuses may turn out not to be quite that brilliant earlier
on.  The people you mention might well have been 1-in-a-million
types as teenagers, of which there are (by definition) about 6000 on
the planet.  I'd be surprised if they weren't at least 1000 times
more likely to reach greatness than people below them on the curve
(as teenagers).  But, of course, there are a lot more people below
them (in fact, a million for each one!) so they might not dominate
the list of top adult geniuses, or Nobel Prize winners, or what have 
you.  

Interestingly, though, these kinds of lists often have subtle
properties.  As far as the Nobel Prize goes, there are (pretty
famously) many, many more people who could have won the prize
compared to the actual prize recipients, and many Nobel Prize 
winners who were not only brilliant, but, quite frankly, in the 
right place at the right time.  

OK, so now it's time for a not-very-random example:

Kary Mullis won the Nobel Prize some years back for inventing the
polymerase chain reaction (PCR).  It's pretty clear now that
*whoever* had done that was going to get a ticket to Stockholm,
since it is a technique that completely revolutionized molecular
biology, and molecular biology is the fastest moving science of our
time.  Now, Mullis did do it first, though, so he won.  

The reason I mention him is that his official biography is not the
kind of thing that you would usually expect to end in "and then he
won the Nobel Prize".  He was 22 when he got his Bachelor's at
Georgia Tech, 27 when he got his PhD in Biochem from Berkeley,
taught for a year, spent a 4-year post doc at the University of
Kansas doing stuff completely unrelated to anything else he ever did
before or since, then did a 2-year post-doc in something else at
UCSF.  He then joined a biotech firm (not usually the fast track to
scientific greatness unless you're the famous founder which he was
not) at age 34, and *finally* worked on a problem that led to
something great...after 4 years of nothing very spectacular.  But
when he was 38, he had one really good idea.  OK, cancel that: he
had one of the best ideas anybody ever had in the 20th century.  
(And, very oddly, the same idea in spirit that many people have had 
in your better first year programming courses.)

And so it was that he won a Nobel Prize.

BIG Digression: 

Here's the idea.  In many situations, it would be wonderful if we
could magically make billions of copies of some piece of DNA;  
perhaps it came from an important piece of crime scene evidence and
would identify the killer.  What to do?  Well, it turns out that we
and all of our cells have enzymes known as DNA polymerases which can
synthesize a new (complementary) strand of DNA given one strand of
the DNA, an adequate supply of the nucleotides that need to make the
copy *and* a short strand of DNA (can be really short) that just
happens to bind to the beginning of the part of the DNA that we want
copied.  That short strand is called a "primer"; basically it primes
the polymerase to get it started copying the DNA.  Here now is the
famous PCR technique:

1) take your *one* sample of double-stranded DNA, and heat it to the
   point where the two strands separate.

2) Cool the strands a bit, and your primer will now stick to the two 
   strands.

3) Now the enzyme "DNA polymerase" will copy the rest of your DNA 
   sample until you hit the end of the strands.  (For fastest
   results, you might need to raise the temperature a bit first.)

4) You now have *two* samples of double-stranded DNA.

OK, so do you see where this is going?  I started with one, and now
I have two.  If I want a million strands, all I have to do is repeat
the four steps above...only 20 times, since 2^20 is about a million.  
If I want a billion samples, I run it thirty times.  All I need is
one piece of DNA, enough primer (easy to get, don't worry), enough
nucleotides (ditto) enough enzyme, and some basic robotic technology
to stir stuff, heat and cool it on a timing circuit.

Now do you get it?

A-yup!  The dude got a Nobel Prize for implementing a simple
recursive algorithm in a test-tube using materials he basically just
found lying around the lab.

Now, I'm sure there's a lesson in here somewhere, but I'm not sure
what it is...

jking

_______________________________________________
discussion mailing list
EMAIL:PROTECTED
http://mlug.missouri.edu/mailman/listinfo/discussion