[Infowarrior] - Insightful: Former DARPA Director speaks

[Richard Forno]

The IT Godfather Speaks: Q&A With Charles M. Herzfeld
Gary Anthes
 http://computerworld.com/action/article.do?command=viewArticleBasic&article
Id=9035398&pageNumber=1

September 24, 2007 (Computerworld) Is DARPA still funding the kinds of
research that made the U.S. an IT leader? Charles M. Herzfeld, a senior
fellow at the Potomac Institute for Policy Studies in Arlington, Va., has a
few thoughts on the matter.

Herzfeld was hired by the Advanced Research Projects Agency (later renamed
the Defense Advanced Research Projects Agency) in 1961 to head up research
in ballistic missile defense, and he became ARPA's fifth director in 1965.
He also served as director of Defense Research & Engineering, to which DARPA
reports, from 1990 to 1991.

What was your introduction to computing?

When I was a graduate student at the University of Chicago, in 1948 or so,
John von Neumann came and gave three seminars on electronic computing. He
was instrumental in getting the ENIAC built, and he came to tell us about
it. It was hugely important stuff, and it changed my life absolutely.

Then, before ARPA, J.C.R. Licklider gave two or three lectures at the
Pentagon, and I remember those vividly. He said, "The way we were doing
computing is really pretty stupid. I think there's a better way." He was a
brilliant man, and I became a disciple of his.

And a few years later, you and Licklider would end up at ARPA, with
Licklider the first director of its Information Processing Techniques
Office.

Yes. IPTO was one of the things at ARPA that I became godfather of. I was
the go-to guy if it got into trouble. [The IPTO] directors changed the
world, but I claim to be the godfather, not the father. And as godfather, I
took their message to Congress.

What else did you do as godfather?

I signed the first two or three ARPA orders in 1966 and 1967 as director. I
said, "Do that -- build a network, however small and crappy it is." Lick was
gone by then [he went to IBM in 1964], but I had recruited Bob Taylor as the
follow-on.

One day Taylor dropped into my office, and he got $1 million in 20 minutes.
He acts like I was sitting in my chair handing out million-dollar checks,
but not so. I was sure that networking computers would change computing. I
do not claim to have foreseen what happened, but I knew Licklider was on to
something.

Did you casually hand out big sums like that very often?

Whenever it was needed. My secret was that I always had money because there
was a long list of things we were doing that we didn't have to do. I was
ruthless about that.

What else did IPTO do in those early days?

We created the whole artificial intelligence community and funded it. And we
created the computer science world. When we started [IPTO], there were no
computer science departments or computer science professionals in the world.
None.

Do you agree with some today that DARPA has pulled back from the long-range,
high-risk projects?

There certainly has been a change, and it's not for the better. But it may
be inevitable. I'm not sure one could start the old ARPA nowadays. It would
be illegal, perhaps. We now live under tight controls by many people who
don't understand much about substance.

What was unique about IPTO was that it was very broad technically and
philosophically, and nobody told you how to structure it. We structured it.
It's very hard to do that today.

But why? Why couldn't a Licklider come in today and do big things?

Because the people that you have to persuade are too busy, don't know enough
about the subject and are highly risk-averse. When President Eisenhower
said, "You, Department X, will do Y," they'd salute and say, "Yes, sir." Now
they say, "We'll get back to you." I blame Congress for a good part of it.
And agency heads are all wishy-washy. What's missing is leadership that
understands what it is doing.

The Washington Post [on Aug. 13] ran a Page 1 story saying that the FBI had
given emergency responders $25 million in "computer kits" for exchanging
information on suspected explosives, including weapons of mass destruction.
But, The Post said, many of the kits didn't work and some were just
abandoned. What do you make of that kind of report?

We are becoming incapable of handling a technology challenge of any major
magnitude. We are losing the ability to do big, complicated things. In your
example, nobody thought that someone had to organize a maintenance space for
repairs, spare parts and so on. They only thought about buying the radios.

Is it partly a failure of technology?

Absolutely not. We have technology on the shelf we don't know what to do
with, and we are buying more every day, to the tune of billions of dollars a
year. What's missing is leadership that understands what it is doing. The
whole thing is just off the rails.

What's the story at the National Science Foundation?

My friends complain that they have to submit 10 proposals to get one funded.
Cuckoo. And it's tremendously demoralizing and very inefficient. The process
is too risk-averse. But doing really good research is a high-risk
proposition. If the system does not fund thinking about big problems, you
think about small problems.

Could there be another Sputnik?

Yes, I expect it. In the biological world, it may be an accident. Someone is
doing virus research and comes up with something that spreads easily and
kills a lot of people. There is terrorism. It is absolutely thinkable that
these guys will steal a nuclear weapon, have some technical help and blow it
off in New York Harbor.

Gary Anthes is a Computerworld national correspondent.