Authors@Google: Kurt Beyer


Uploaded by AtGoogleTalks on 20.05.2010

Transcript:
>>Presenter: Welcome everybody. Before introducing our, today's guest. I want you to picture
a different time. I want you to forget about these; which hold a million or so gates on
them, called CPUs. I want you to think in these terms. Vacuum tubes. This is probably
half a dozen logical gates in this module alone.
[pause]
Megabytes of memory, forget about that. We're talking visible bits here. You're free to
come later after the talk to see the bits. Each single bit of memory.
[pause]
And as far as input goes? Forget about this. And get used to this. Patch panels.
It is my distinct pleasure to introduce today Kurt Beyer, who is going to transfer us to
this time of early pioneering computer science giants, and take us on a little journey about
Grace Hopper, whom you all know about. So let me just read a little bit about Kurt.
So, Kurt grew up in a blue collar immigrant family in Huntington, Long Island. Kurt's
dad Karl was a baker and his mum, Ann a nurse. Kurt was a captain of the baseball and basketball
teams at John Glenn High School and an accomplished trumpet player. And received his nomination
to the US Naval Academy from the late Senator Daniel Patrick Moynihan.
While at Annapolis Kurt played baseball and in senior year, was named brigade commander.
In charge of the 4500 person brigade of midmanship, midshipman. He graduated Annapolis in 1990
and was commissioned an officer in the United States Navy. Before attending flight school,
Kurt continues his education at the University of Oxford for 2 years. At Oxford he completed
a Master's degree and rowed for Oxford where his crew competed in the finals of the Henley
Royal Regatta in 1991. He also played on the university basketball team which won the British
university championship in 1992. Following Oxford, Kurt headed to Pensacola for naval
flight school where he graduated first in his class. Kurt flew F-14 Tomcats and was
assigned to a fighter's squadron at Naval Air Station Oceana in Virginia Beach. Injury
cut his naval career short and Kurt was honorably discharged receiving a Navy Commendation medal
and National Defense Service Medal. In 1997 Kurt moved to California to convalesce and
complete a PhD at the University of California; Berkeley.
Kurt really immersed himself in the Bay Area Dot-Com revolution, co-founded a digital media
start-up, and married the beautiful fourth generation San Franciscan.
That's why it's important to move to California. Listen to us guys, out there.
The tragedy of September 11 changed Kurt's path and he returned to Annapolis as a civilian
professor to help create the Naval Academy's new information technology major and lecture
regularly in the process of technological innovation. He served on the academy faculty
3 and a half years and helped direct the international scholarship program. During this period the
naval academy had the most British scholarship winners of any American University, including
8 Rhodes, 3 Marshall and 4 Fitzgerald scholars. In January 2006 Kurt returned to the San Francisco
area to head up full-time a digital media start-up and co-author, multiple patents pending,
on high speed digital data processing. Kurt currently works at Morgan Stanley Smith Barney
and advises start-ups and executives in Silicon Valley. He lives in Mill Valley, California
with his wife Johanna and two sons Charlie and Gus. Please welcome Kurt Beyer.
>>Kurt: Thank You Boris. I'd also like everyone to forget about Larry and Sergey and think
of a woman named Grace Hopper who actually I feel is one of the key founders of the computer
industry. And also influenced the direction of that industry. So I'm assuming in this
group people have heard of Grace Hopper. Great, I see a lot of nods. Often times when I'm
giving talks I just get blank stares. And those blank stares were actually one of the
reasons why I decided to write the book. I first met Grace Hopper when I was 13years
old. I went to my sister's graduation from the College of William and Mary.
And Grace Hopper was the commencement speaker. And I remember that day; 2 things really stuck
in my mind that day. One; I saw this old frail woman; she looked like that in 1983, who was
sitting there knitting while the president of the university was introducing her. And
I thought that was a little strange. And then she stood up and I really thought; what am
I going to learn from this grandmotherly figure? And she started talking about a technical
age that I couldn't even imagine at that time. And that's really strange right? Usually your
grandmother is telling you about the past and how hard it was in the past and how easy
you have it now. So Hopper was talking about the future and a future that we would all
inherit one day. The next time I ran into Admiral Hopper was plebe year, freshman year,
at the naval academy. And what was really neat about my experience in the mid 80's at
the academy, is Grace Hopper arranged for all incoming students to receive their own
computer. Which was a really big deal in the 1980's for an eighteen year old. We could
bring that computer up to our room, we could plug it in to something called the Ethernet
which was connected to MILNET and we could schedule our classes, email our professors,
our medical or dental records were all digital. And this was all in the 1980's. And she came
back to the academy and addressed my entire class, and she said we're giving you these
tools so that you can help lead this digital revolution. And so I never forgot that. When
I came to Silicon Valley in 1997 I was shocked first of all, how everyone was telling me
how they were inventing this thing called the Internet and the World Wide Web. And we're
going to be able to do things like have digital medical records online, and email each other,
and I thought, you know, there's an 85 year old woman that explained all that to me about
12 years ago. And I couldn't believe that no-one had heard of this woman. So when I
started researching the foundations of the computer industry when I went back to the
naval academy, one thing that struck me was, this woman, a much younger vibrant Grace Hopper
who was responsible for so much. So I decided to tell the story of that first 30 years through
her career. It's hard to know when to start a book. Especially
a book about an industry. Most people would start with the most obvious thing. The first
computer. And that didn't seem right, because one, which computer do you pick? Is it an
English computer? Is it a US computer? Other people would argue, well you can't start with
a computer; you have to go back to calculating machines prior to that like Charles Babbage.
So I decided to start the book, not with a piece of technology but with an event. My
argument is that all radical technical change needs some type of crisis to bring it on.
And so that crisis was December 7th 1941. Within a week's attack of Pearl Harbor, you
saw the US government go to war, start changing it's policy of research and development and
start funding exotic projects. Some of those being in calculating machines. But these types
of events also have an effect on the individual. Think back to our own experience in 9/11.
Most of us remember where we were and for many of us our lives were changed forever
because it makes you reflect on your own path, and maybe you take a new path after the crisis.
That happened to a woman named Grace Hopper. We would never have heard of Grace Hopper
if it weren't for Pearl Harbor. On December 6th, Grace Hopper was a tenured track professor
at Vassar College. She had her PhD in mathematics, she was married. She lived in New York City
with her family, ah, all living on the same street. They had a summer house in upstate
New York. She had a very normal, stable life. Within 6 months of Pearl Harbor she quit her
job, left her husband and tried desperately to join the Navy. The problem was there were
no women in the Navy at this time. And I think this gives us the first insight into Grace
Hopper's personality. She didn't let simple things like there were no women in the Navy
stop her from doing something. Well sure enough, the rules were changed, the WAVES were formed
and Grace Hopper got her wish. The Navy didn't know what to do with a 5 foot 2, 105 pound
PhD in Mathematics, so they sent her to a top secret project up at Harvard, called the
Harvard Mark 1. Let me read to you a little about her first day on the job.
"Hopper reported to Howard Aiken's Harvard command in the basement of Harvard's University's
Croft Physics Laboratory on 2 July 1944. After being escorted down to the basement by an
armed guard. If this wasn't disquieting enough, the presence of her physically intimidating
boss was. 'I was a little bewildered and at that point, of course thoroughly scared',
Hopper recalled. Years later she distinctly remembered the first words uttered to her
by Aiken. 'Where the hell have you been?' Aiken had expected Hopper months earlier and
had not seen the point of sending a female mathematician to midshipman school. Aiken
proceeded to give his new recruit a tour of the automatic sequence control calculator.
Hopper had arrived before IBM had installed the smooth steel casing, so the machines thousands
of moving parts were full exposed. 'All I could do was look at it. I couldn't think
of anything to say at that point'. Aiken ordered his new assistant to put off finding a place
to live for another day and to get to work immediately. Her first task was to compute
the interpolation coefficients for the arctangent to the accuracy of 23 decimal places within
in a week." [Pause] Pretty simple project, right?
The problem was; it couldn't be done in a week. And she was told she was in charge of
this machine. A machine that had never existed before. A machine that she had never seen
the likes of. 81 feet long, 10 feet high, 750,000 moving parts, 5 miles of wiring. So
what did she do? She went to the manual of operation. There was no manual of operation.
This is the first computer and she's in charge of it. Think about how disquieting that could
be. The machine actually was developed like most machines were of the day. For a singular
purpose. And this case, this was a machine to create ballistic tables. So that's what
Hopper had to do initially. Was figure out how to get the outputs to be what the purpose
of the machine was. The navy, once they started hearing that they
had some type of high speed calculating machine, that could, in fact, do 3 calculations per
second, which was revolutionary at the time. They started sending Hopper and her crew other
problems to solve. And this kind of put Hopper in a conundrum. They had to continue to start
reconfiguring the machine each time a different problem was given to them. The most famous
problem that was given to them in the fall of 1944 was brought by a man named John Von
Neumann, a famous mathematician, and he wanted them to figure out how to cause a sphere to
implode upon itself. And they weren't sure where to put the charges to create this sphere
implosion at a specific rate. It was a problem that Von Neumann and his team couldn't solve
with computers. Well what was a computer? A computer in that day was a person with a
piece of paper and a pencil. And they would have rooms of computers. Often times they
were women. And you would bring the problem, like you'd bring a shirt to the Laundromat.
You'd hand it in, they'd divide the problem up, and people would actually solve it with
paper and pencil. They didn't have enough time; this was a critical war problem. So
it took Hopper and her team four months to solve the problem and of course we all know
what the problem was used for. The "Fat Man" was dropped on Nagasaki, it used that implosion
problem, and 6 days later the war ended.
[Silence]
Now Howard Aiken was in charge of the Harvard Mark 1. He has an interesting story in and
of his own right. In 1938 he was a graduate student at Harvard. But he had spent many
years working in, for power plants for industry. And he got the idea for a calculating machine
initially to help him do some of his PhD work. Harvard saw itself as a [Labrard] school and
was not willing to support Aiken's effort. And Aiken tried for 5 years, excuse me, for
3 years to get their support. It wasn't until Pearl Harbor which opened the door for Aiken
to receive funds to try to build his computer. He had been a Naval reserve officer, and so
he bypassed Harvard administration, and received funding directly from the US Navy, rented
the basement of the Croft Laboratory, built out the computer and then staffed it, not
with Harvard students and professors, but with naval officers and enlisted. He would
not allow any of the Harvard professors who wanted to run their problems on this machine,
he would not allow them to use the machine. People offered compare him to Charles Babbage
and Aiken was the first to compare himself to Charles Babbage. He saw himself as the
second coming of Charles Babbage. And here you have Aiken and Hopper reviewing parts
of the original difference engine that Babbage had created. In effect Aiken saw himself as
Babbage and Hopper, his assistant, as Ada Lovelace, if any of you are familiar with
that history. Aiken was a hard man to work for. He decided to run the computation laboratory,
not as a civilian laboratory but as a military ship. And the computer was their ship. Everyone
called him commander. They operated the facility 24 hours a day, 7 days a week. They had watches,
they had logs, they had an officer of the deck. But the advantage was that he got things
done, and he got them done rapidly. One, they built a prototype and got it functional during
the war, and it did some significant work including the implosion problem for the war.
That type of innovative environment was good for getting things done, but it has its shadow
sides as well. And let me read you a piece from Hopper to describe what it was like working
in this environment. "Early programming innovations grew out of
the intense pressure generated by the war. Hopper compared the experience to working
in a long tunnel with only the problems to solve ahead of her. Her and Richard Block
were constantly searching for ways to increase the speed of the coding process and their
solutions evolved out of the expediency rather than intellectual curiosity."
So what did they create during those war years? Well they came up with the concept of subroutines.
They would use a tape to program the machine, to code the machine. And what they realized
was that once they had a piece of tape that was cleared of errors, why not use that tape
over and over again. So they started building a library of tapes, and then they could bring
and put those tapes together when they were working on another problem. They didn't call
it subroutines at the time but in effect it became the basis for some of Hoppers work
later on. They developed a concept of flow-charting, to chart out the problem they were working
on, and to allow people who were working on the problem to know what the person had done
before them. So you start seeing the rudimentary operational system behind programming also
developing in '44 and '45. The most famous problem that they solved, or at least identified
during this period, was discovered one night when the machine stopped working and they
inspected the machine and realized that a moth had been captured in one of the relays.
So in the log book, they taped the moth down and under it they wrote that "we've de-bugged
the computer and it seems to be working fine now". And you actually can find the original
computer bug at the Smithsonian Museum. [Pause] A lot of people feel that Von Neumann is the
one who came up with the stored computer architecture. We even refer to it as Von Neumann architecture.
And a lot of historical background is based on a lawsuit between Von Neumann and Presper
Eckhart and John Mauchly who are here in the middle who built another computer down in
Philadelphia at the time, called the ENIAC. What I discovered in the research was that
before Von Neumann went down to the ENIAC project, he had been working with Hopper for
four months at the Harvard project. The ENIAC didn't even have programming tape; they just
had plug boards that they would reconfigure the machine to do a problem with. So it's
my feeling that Von Neumann was the one who got to see what they were doing up at Harvard
and then look at the Vacuum tube technology that they were using at ENIAC and then write
his famous spring of 1945 paper 'First Draft of the EDVAC' to kind of build the case for
the stored program architecture. These ideas don't come out of the blue so I think we need
to give credit where credit's due and if you actually read the original paper, he even,
in his footnotes thanks Aiken and Hopper, but never mentions the ENIAC project.
[Pause]
The hard work at, um at Harvard, started to take its toll on Hopper. I think during the
war, during that mission she was fine, but then once the war was over, she started having
road blocks to her career. She wanted to stay in the military but once the war ended, the
military said "we're going to go back to our pre-war policy and not have active officers
as women". She wanted to stay on at Harvard, Howard Aiken was made a full professor, but
Harvard did not hire women as professors. So they gave her a 2 year part time research
grant to continue to help Howard Aiken, and then they released her. Now let me read a
little bit of what happened to her next. "On a cold night in November 1949 only 6 months
after leaving Harvard and joining the Eckhart and Maulchy Computer Corporation, Grace Mary
Hopper found herself behind bars at the Central Philadelphia police station. The programming
pioneer was arrested at 3am for drunk and disorderly conduct. She was eventually placed
in the custody of Pennsylvania General Hospital for treatment. Hoppers' life was unraveling.
At the age of 43 years old she had accomplished much, yet her growing dependency on alcohol
was jeopardizing her career and her relationships. As winter approached, she attempted to commit
suicide 2 different times." I wanted to include that in the book because
I think it's important for us to realize that pioneers and innovators are human. And Hopper
went through a lot during those years. She accomplished much, but it had a grave toll
on her personally. And interestingly enough, it's her crew from Harvard, her team from
Harvard that she built and worked with during the war that came to her rescue during those
years. Got her sober and got her back on the right track. And so she started working on
this project, the first computer start-up company, the Eckhart and Maulchy Corporation.
And they build the EDVAC which Von Neumann mapped out, and so this would be America's
first functioning stored program computer. Now Hopper didn't have much to do in 1949,
1950 and 1951 because no one really understood why you would want one of these. Howard Aiken
was asked one time, "How many computers do you think the country needs?" And he thought,
well, about 12 and that would handle all the computational needs of the nation. Well there
was an event that occurred which I think epitomizes the change in perspective around these machines.
Someone in the marketing department at UNIVAC, got the idea to use one of these computers
during the first televised presidential election; Eisenhower vs. Stevenson. They would collect
data from previous elections; voting patterns, enter those into the UNIVAC, and then with
1% of the vote in, they would try to predict the outcome of the election. They couldn't
fit a UNIVAC on set next to Walter Cronkite, so they built like a Star Trek version of
it with just some flashing lights next to him. And with 1% of the vote in, the UNIVAC
was saying that Eisenhower was going to win by a landslide. 490 electoral votes to 89.
The Roper poll was saying that it was going to be a close race. So what did Walter Cronkite
do? He lied to the American people and said the UNIVAC says it's going to be a close race.
Well sure enough, when the votes were in, 491 to 88. The UNIVAC had missed by one electoral
vote. A UNIVAC representative went back on TV and told the American nation that in fact
the UNIVAC had guessed what the outcome would be. And you can imagine every CEO in the country
who was watching that night started thinking these machines can predict the future, we
need to get one of these. Now there's other reasons why orders started
taking off in 1952, 1953, but sure enough, during this period the UNIVAC was so well
known and so popular amongst business that it became their Coca Cola or Kleenex, and
we may have called computers UNIVAC's if that company had, was able to beat IBM during the
1950's. Because of the UNIVAC's popularity, Hopper now had a crisis. She was in charge
of getting these machines to work. And you can imagine each company that bought a UNIVAC
wanted it to do something specific to that company. So if an insurance company bought
one, they wanted it to do actuarial tables. GE wanted it to do payroll. Hughes Aircraft
wanted it to do engineering problems. So Hopper was in charge of getting the systems running
and configured for each of these companies. The problem was there weren't many programmers
at that time. The concept, the profession was just starting to evolve. So out of crisis
comes innovation, and Hopper, goes and, over an 8 month period invents a critical piece
that she would call the Compiler. The compiler would allow the computer to start helping
the programmer program itself. She tells the world about the compiler in a very funny titled
paper 'The Education of a Computer' and in that paper she maps out how a Compiler can
pull from a library of subroutines, um, take care of many of the administration functions
of the programmer and create automatically, machine code. She, 6 months later, produces
another paper, called 'Compiling Routines' and this one really is the interesting paper
because she starts explaining the concept of source code. She calls it Pseudo Code at
the time. And what you'll be able to do with source code, or pseudo code, is the programmer
or programming team that creates these higher end languages can now be linguists. And they
can produce programming languages that you don't need a PhD in mathematics to be able
to use. It was key to liberating the technology. It was the key to solving the staffing problem
at many of these companies. The next step, was she had to convince management
that this was a good idea. So she decides to create a program; one in German, one in
French, and one in English. And all 3 programs would produce the same machine code. And she
demonstrates it to management. And management is furious. "We're having a hard enough time
helping the American companies program. Now you're wanting to help French and German companies?"
She said, "No you're missing the point. The point is we can make source code or pseudo
code anything we want it to be. We can make it mathematic based, we can make it business
oriented." That was the point she was trying to make. One thing that Hopper was good at
was knowing her audience and talking to them in a way that they can understand and get
motivated. So she started talking about savings. The savings we would have if we could make
programs that high school students could program or undergrads could program. Um, these companies
would stop poaching our programmers. And that's when they started getting the value of what
she was talking about. This automatic programming concept. So they ended up giving her a budget
in 1954, and creating the Division of Automatic Programming at UNIVAC. And Hopper was put
in charge of it. She became the first director of Automatic Programming at UNIVAC.
You know a lot of people ask me, how did Hopper achieve so much in the 50's at a time when
after the war there were so many efforts to try to get women out of the workplace? And
one of the answers was that she invented her positions. That's one way to break the glass
ceiling, right? Is to invent the job, and then assign yourself as head of the division.
So that's what she did. Hopper, during this period, becomes a sales person for this vision
of automatic computing. Believe it or not a lot of programmers, so there was a community
of programmers by the mid '50's, were against her efforts. They wanted to be the High Priests
of these machines; they wanted to keep them complicated. And Hopper's pretty much trying
to undermine their new profession. Well I think it teaches us an important lesson about
how technology evolves. I think a lot of times we just think that technology is just inevitable.
It just happens. It happens to be more logical than other things. When in fact, an inventor
has to be a good sales person. So what are the tools of a sales person? What did Hopper
do during the 1950's to start spreading this concept of Automatic Programming? Well, first
of all, she started employing as many people as she could in the design process. So of
course people on her staff at UNIVAC worked on the first compilers and the first programming
languages; Math-matic and Flow-matic were the first two that she built out. But she
also incorporated people from other companies, and the users from the companies that were
using the machines. So in effect, what we would call say, open source programming today,
was something that was the norm during the '50's and something that Hopper helped orchestrate.
She also used a variety of institutions to help push her vision forward. So one of them
was a computing group called the Association of Computing Machinery; the ACM. Many of you
probably are a part of it now; it's the largest organization within computing. It was actually
founded by the group that worked at Harvard. Harvard's crew. Ah, Hopper's crew at Harvard.
And by the 1950's they started having their own annual meetings. And Hopper put herself
in charge of two important committees. One was Senior Editor, so she got to pick the
types of articles that were being published for the ACM. And the 2nd one I thought was
very interesting also. She put herself in charge of the Nomenclature Committee. Every
profession has its own language. And that language is in essence, is power. If you can
define the language of being a Doctor, or being a Lawyer, being a Military Officer,
you're a pretty powerful person in that profession. Hopper realized that. And so she was the head
of the Nomenclature Committee during the early 50's as well which started defining the language
of computer programming. She also relied on her friends at Department of Defense. Even
though she had to leave the military, she always remained part of the military during
those years. She was a reserve officer, and every 2 weeks she would work at DOD, and get
a sense of what their computing needs were. So during the 50's she started having DOD
pay for different workshops on Automatic Programming, on different trainings, and so this way she
started getting some of the largest consumers of these new technologies on board with her
vision of programming. I think this salesmanship, this method of inventing in an open way, culminates
in 1959. She goes to DOD and suggests that we should all work together to form a common
business language. And so she sets up a meeting for 2 days in the spring of 1959 where they
pull in 7 government agencies, 11 computer users, and 10 computer manufacturers. And
for 2 days they brain storm what would the perfect, the ideal, business language would
be, what would it look like? They also form 3 committees. A short term committee, a medium
term committee, and a long term committee. The short term committee is going to take
the suggestions and map out what the standards will be for this language. The intermediate
term committee then would continuously improve it. Well, the short term committee happens
to be made up of people that used to work for Hopper. And interestingly enough 50% of
the Short term committee were women. And through the fall of 1959, they build out the standards
for what they named during the fall; COBOL. The language is COBOL. Hopper being the senior
tech person of the committee signs off on the standards of the new language in January
of 1960. They launch the language and everyone hates
it. It's terrible. It's too verbose. It takes up too much memory. It's slow. So the question
is; how did COBOL become the most successful programming language to date? 70% of all the
active code today is COBOL code. How does this happen? Well, Hopper understood something
about innovation; that the most logical innovation isn't always the best innovation. You have
to build into the technology itself the different groups that will have a hand in pushing that
technology forward; there are economic parts to a technology, there are political parts
to a technology and there are social parts to a technology. So in essence Cobalt ended
up being a second best solution that pleased no one and everyone.
But that salesmanship comes back again. You can build the prototype but then you have
to push the prototype out. And so for much of the 1960's Hopper dedicates herself to
getting COBOL used by a variety of different companies and organizations.
She used different techniques to push COBOL. You're probably wondering what does Marilyn
Monroe have to do with COBOL? Well I'll tell you. As the 1950's came to a close, Barry
Rand's, Director of Programming Research, Grace Hopper, conducted an experiment to demonstrate
how far the programming language had advanced in the 15 years since she wrote her first
code. Hopper would attempt to turn a trim attractive blond into a computer programmer.
Hopper's subject; Marilyn Mealy, Marilyn Monroe, Marilyn Mealy, a 19 year old blond high school
graduate from the Mayfair section of Philadelphia, was prettier than average and liked to swim
and listen to HiFi records. According to Hopper, Marilyn was much like other young women who
window shopped during their lunch hours and looked forward to their evening dates and
dancing. And sure enough, in 6 weeks Marilyn Mealy became a COBOL programmer. And Hopper
would parade Marilyn around and say "Look what you can do with this language". She used
other techniques also. More hardball techniques like she convinced her friends at DOD that
they would not buy any more computers after 1961 that did not have an operational COBOL
compiler. So there's other ways to push a certain technology forward as well.
In 1967 Hopper retires from UNIVAC and she could've had a very nice relaxing retirement
from that time forward. She had accomplished a lot. And people in the industry understood
how important she was. The ACM actually named her the first Man of the Year that year, that
she retired, which I thought was funny. They also named after Hopper, kind of the Nobel
Prize for Computer Science, the Grace Hopper Award in 1971. And many of the luminaries
here in Silicon Valley have won the Hopper award. But Hopper wasn't done. In '67 she
decides to go back into the Navy and help them standardize COBOL. It's a 6 month project,
20 years later, she's still on active duty and she's the first woman to become an admiral.
And that's when I got to meet her. She was at the end of her career and I was at the
beginning of mine. So I want to finish today with, what are some lessons that we can glean
from Hoppers experience and how can we apply those lessons in our own work today?
Hopper used to say, "You manage things and you lead people". So what did she mean by
that? You manage things but you lead people? Part of leading a group means that you need
to provide them with a purpose, with urgency, with a vision. During World War Two that urgency
was provided to Howard Aiken externally. People worked night and day because of the war effort.
Interestingly enough once you took the war away, the process of innovation collapsed
on Howard Aiken because he didn't know how to lead people. Hopper was so good at painting
a certain type of computer future for people and then pushing them all in that direction
and giving them a sense of purpose. One little trick she would always do, is she would hand
the most difficult problems off to the youngest, most inexperienced people on her staff. And
she didn't think that they'd actually solve the problem, but what she would say is they
didn't realize that they should fail and that's half the battle. She'd watch them; how they'd
approach the problem and they'd usually hand off the problem to a more senior person but
maybe with some insight into a different way of solving it. Something that Hopper was very
good at also, which served her well in the 1950's was learning what I call learning from
the margins. She had this peculiar habit when she was a college professor. So she taught
mathematics, but each semester even though she was teaching a full load of classes, she
would audit 2 classes. And those classes had nothing to do with Mathematics; they ranged
from architecture to philosophy to history. I think what she realized was experts sometimes
can't think out of the box. Their expertise becomes an intellectual prison. So one way
to try to keep the pathways open is to constantly learn about different fields. And think how
well that served her during the 1950's. So she would go visit Hughes Aircraft. And in
just a few days she had to figure out how the company worked, what the business model
was, and what types of applications they would need. And then she'd go to another company
and be completely different. And in a way her mind became like that general computer.
Right? It was malleable; it could change to fit the situation. And I think it's Hopper's
habit that she already had in the 1930's which helped create the uniqueness of the computer
industry during the '50's. Because at the end of the day, what is unique about computer
technology? Any technology before computers had its own Talis, its own purpose. So you
design and build a lawn mower and you use it to mow lawns. A hammer hammers nails. A
computer can become whatever you want it to become; and I think its epitomized best by
Smart Phones. Now, this one is manufactured by another company that will remain nameless.
But when I wake up in the morning; I check my emails, the temperature, I hand it off
to my son, he plays a game, he hands it back to me, I look up some stocks, I make a call.
That's the power of computers. It's the fact that they can evolve and be anything that
you want them to be because of the software, because of the programming. And when I was
reading other histories of computing, it always shocked me that they concentrated so much
on hardware and hardware development, when in fact I think programming and programming
development is the real key to the industry. And so we have to thank Grace Hopper for that.
We talked about her ability to sell the vision. So you've got to sell it internally to your
team, to motivate them. But then you have to sell it externally, and that means you
have to be able to talk to a variety of different audiences and know what that audience needs
to hear. And think about what Hopper did throughout her life, she always kept one foot in the
military, one foot in academia, and she was a business woman. And so she knew how to talk
to each group. She knew the important things that they needed to hear and she could morph
her vision to their needs. And finally, I think the main point that I
want to leave with everyone today, is that Technology is not inevitable. I think we fall
into that trap sometimes. We think that it's supposed to develop that way, or of course
it evolved that way. In a way that's sad because it takes away Human Agency, right? Human beings,
especially in the early part of a technology's development can influence the direction of
that technology. You take out Grace Hopper from the equation; you get a different computer
industry. It evolves in a different way. To remind her of that, and to remind the people
she worked with of that, during her navy days she kept a clock on the wall, and the clock
had all the numbers backwards and the hands ran backwards. And her point was; we could've
designed a clock any way we wanted to design it. It didn't have to be designed the way
that eventually there is a consensus around. And every technology is like that. So, remember
that as you work on your own technologies. Finally, I , I think, you know, in a kind
of a post financial world, Grace Hopper serves as a wonderful role model for all of us. Um,
she, she worked and created and innovated, not necessarily for the stock options or for
the financial payout, but she really was enamored with the technologies that she was working
with, she cared about the people she worked with. And she had a sense of duty, to self
and country as well which I think is not typical sometimes for Silicon Valley. And I really
admire her for that as well. That's probably one of the reasons why people have never heard
of her, because she actually didn't make much money during these years and so we have to
thank her though, for being such a strong pioneer and setting the foundation for the
rest of us going forward. So I'd like to take some questions at this point.
[Clapping from audience]
>>Person 1 from audience: So you know today we think of compilers compiling into an assembly
language which is some representation of the underlying machine codes. Um, how did things
look in the '50's to then, 'coz you haven't mentioned the, you know, machine codes and
assembly languages and things like that? And also what was the correlation, if she had
any, to John Backus who did the Fortran Compiler?
>>Kurt: Sure, sure, yeah. Um, yeah there were 2 paths that were happening. Hoppers was the
compiling path. Ah, there were other languages growing, mainly out of MIT which were taking
a different path. She was very friendly with John Backus actually. When I talked about
how she built this team that went beyond the borders of UNIVAC, Backus was actually working
at IBM. And one of the things that motivated Backus was how slow MATH-MATIC compiled. So
when he was designing what would become FORTRAN, the main thing he was trying to solve was
to figure out how to increase compiling speeds. And so, very friendly with Hopper. Hopper,
there were, software wasn't monetized at that point so she had no problems working with
him, making suggestions, they all knew each other at the different conferences that they
were going to. So it was a much different time in terms of development back then.
>>Person 2 from audience [European accent]: I thought that was an interesting comment
on Open Source in the early days. I think, ah, you just said it, which is that software
wasn't monetized in those days. Are there any lessons that, in like the discussions
around software patents for example should take from that era when there were no patents
so to speak of? And anything else we could learn about that?
>>Kurt: Yeah, yeah. I think one could argue maybe you had a faster rate of innovation
during that period. You start seeing a shift because of the IBM Anti-trust case in the
late '60's and so IBM is forced to stop giving the software away for free with the hardware.
The beginnings of the software industry are already developing in the '60's, but that
really gives the main boost then in the 1960's, ah, 1970's. I think it's no wonder that, you
know someone like Bill Gates really helps redefine the business model around software,
in that his dad was a lawyer, not an engineer. And, I mean, one could make the case that
also that bringing a group of smart people together and having them work on a singular
project and then patenting the outcome is an effective way of innovation as well. So,
I guess both and, both ways of doing it are both viable and we've seen it work, um, but
open source isn't something new. Open source was actually the way things were originally
done.
[Silence]
>>Person 3 from the audience: OK, Very good.
>>Kurt: Thank you.
[Clapping]