Yup: When I was looking for VC funding,
lots of VC Web sites claimed to like
innovation. Eventually I concluded much
the same as in Phil's essay.
The nutshell description I formulated was
the VCs want a startup with a
product/service easy enough for any
grandmother to understand, traction
growing up and to the right, in a huge
market, with five founders, with all
credit cards maxed out, with at least four
of the spouses pregnant, about to go
under, and desperate.
For one step more, they are looking in the
rear view mirror and want to bet on
patterns.
For one more, they want to fund the
company to go to "the next level", i.e.,
the big go to market effort.
The VCs in effect believe that any startup
that is successful will, on the way to
being a big success, say, worth 1+
billion, at some point nearly necessarily
be desperate for some venture funding:
Then that's where the VCs want to write a
check and get on the BoD.
Even if the VCs are willing to do seed
funding, they believe that for each
level more VC cash will be essential.
Then they hope to be the best positioned
to make the series A round with the up
and to the right and go to market
effort.
Here is where I see the VCs failing:
First, the patterns they are looking for
are too rare to be promising. Instead,
for the future, the main pillar of the
success will be something new, powerful,
and valuable for a large market. For
this, it will be necessary for VCs to work
effective with things that are new,
powerful, and valuable. This is not
always easy. Relatively few VCs are able
to do that. Even if they are able, the
limited partners of the VCs won't like
that.
Second, there really is a good pattern
with fantastic evidence from the past. It
is just that the VCs don't see this
pattern.
Third, let's look at this pattern with its
fantastic evidence. At first, we will
just use some examples and then later will
explain with much more substance and
meaning.
(a) In WWII, the US had big need for some
better weapons. The usual VC approach
would have the US go to the designers of
the 16" battleship guns and ask for 18" or
20". Well, the US was smart enough not to
do that. Instead, Szilard, Teller,
Wigner, Einstein, Fermi, etc. led the US
to the A-bomb that ended the war in the
Pacific in about a week.
(b) Teller saw that fusion had promise of
much more yield and in a few years had
Castle-Bravo with 15 megatons of yield.
(c) The WWII aircraft engines went from 9
cylinders in one row to four rows of 7
cylinders per row and extreme efforts with
supercharging, intercooling, magnesium,
huge propellers with tips moving close to
the speed of sound, etc. But a few people
in each of Germany, England, and the US
saw the promise of gas turbines, and we
got some fantastic jet engines. Four of
these on an airplane quickly put the
Atlantic passenger steamships out of
business!
(d) A few people saw that reproducing the
marks on a sheet of paper would be a big
advance in office work and ... we got
Xerox with nearly a license to print
money.
(e) Ike wanted to see what the Soviets
were doing. So, Kelly Johnson at Lockheed
put some long, narrow wings on an existing
jet airplane and gave Ike the U-2. When a
U-2 got shot down, soon Kelly showed up
with an armload of engineering drawings
talking about titanium and a special
engine that Pratt and Whitney had (a turbo
jet up to about Mach 2.5 and then a ram
jet to Mach 3.2 or so). Kelly got his
check, and the US CIA got their SR-71 --
speed 3.2 Mach, altitude 80,000+ feet,
range on one load of fuel ~2000 miles, and
never shot down.
(f) Word whacking was a pain, really
expensive. The IBM correcting Selectrics
were only a little better. But an early
microprocessor, a simple operating system,
some simple word whacking software, a
floppy disk, and some simple printers
quickly eradicated the typewriters. We
got Intel, Apple, Microsoft, and more.
The examples (a) -- (f) are only a few of
the dozens we could list from military
projects, biomedical projects, various
tools, various services, etc.
So, for each of (a) -- (f), at the
beginning, the project had to be
evaluated. The rear view mirror as VCs
apply it would not help. Instead, the
need was to work effectively with ideas
that were new, powerful, and valuable.
Fourth, let's look at some "substance"
supporting the patterns illustrated by
examples (a) -- (f). What is needed is a
way to evaluate ideas that promise to be
new, powerful, and valuable for a large
market.
Okay, one more example: GPS. If the USAF
charged a penny for each commercial use,
they'd soon own the world and everything
in it???? Well, they'd have a lot. Uh,
at one time I was working at the JHU/APL
lab in the group that did the first
version of GPS, for the US Navy,
especially for the missile firing
submarines. I heard the stories. The
project was first on the back of an
envelope. With not much more in ideas, it
got funded and the rest is history.
So how was the evaluation done? Sure,
first some physics. Next some math. Next
some engineering. Money allocated. Work
done. Rockets launched. All done.
So, here it is: The foundation of the
pattern for the evaluation to confirm new,
correct, and significant was some science
and math that could be checked. That's
the pattern: The crucial core of the
project is an idea based on some science,
math, and engineering on paper that can be
checked. That's the pattern.
In the more advanced countries, there are
plenty of people who can generate such
ideas, do the science, math, and
engineering (e.g., the LHC) and plenty of
people who can check the work with high
accuracy. E.g., there are problem
sponsors at the US NSF, NIH, DARPA, ONR,
etc. that can do the checking or manage
the checking done by others. Research
university Ph.D. committees do such
evaluations around the world daily.
Surprise. Please sit down for this: US
technology VCs just will not, Not, NOT, no
matter what, not even under promise of
$trillions, not with water torture, not
with anything, do or manage such
evaluations. Feet locked four feet deep
in reinforced concrete, they won't do it.
Results: First, from all I can see, the
most promising path to the future is new
ideas, presented just on paper, as
science, math, and/or engineering, that
are powerful and valuable for big parts of
our economy and that can be evaluated
accurately. Did I mention "just on
paper"?
That's what the JHU/APL did for the first
version of GPS -- on paper. That's what
Kelly Johnson (there's a picture) brought
to the CIA for the SR-71 -- an armload of
engineering drawings. On and on.
The B-29 is what carried the A-bombs that
ended the war in the Pacific. That plane
cost ballpark as much to develop as the
bombs! A joke about VCs is that they
would say "You build it, drop the first
bomb, and we will fund the gasoline for
the flight of the second bomb."
For entrepreneurs, paper and pencils are
cheap! Can buy a lot of computing for
$2000. A lot of Internet bandwidth is
cheap. So, do the math, etc., write the
software, build a server, plug it into the
Internet, go live, get users, run ads,
make money. To me that's the most
promising path to success. But VCs will
have nothing to do with it.
For the VCs waiting to fund the big go to
market effort, in the past a lot of
companies that were successful grew just
from retained earnings and never took
equity funding. Now with cheap paper and
pencils, computing, the Internet, etc.,
that path seems relatively promising now.