The HLF Portraits: Ronald L. Rivest

Video in TIB AV-Portal: The HLF Portraits: Ronald L. Rivest

19 views

Formal Metadata

Title
The HLF Portraits: Ronald L. Rivest
Title of Series
Author
Contributors
License
No Open Access License:
German copyright law applies. This film may be used for your own use but it may not be distributed via the internet or passed on to external parties.
Identifiers
Publisher
Release Date
2020
Language
English

Content Metadata

Subject Area
Abstract
The Heidelberg Laureate Forum Foundation presents the HLF Portraits: Ronald L. Rivest; ACM A.M. Turing Award, 2002 Recipients of the the Abel Prize, the ACM A.M. Turing Award, the ACM Prize in Computing, the Fields Medal and the Nevanlinna Prize in discussion with Marc Pachter, Director Emeritus National Portrait Gallery, Smithsonian Institute, about their lives, their research, their careers and the circumstances that led to the awards. Video interviews produced for the Heidelberg Laureate Forum Foundation by the Berlin photographer Peter Badge. The opinions expressed in this video do not necessarily reflect the views of the Heidelberg Laureate Forum Foundation or any other person or associated institution involved in the making and distribution of the video.
Loading...
Axiom of choice Statistical hypothesis testing Complex (psychology) Group action Confidence interval Water vapor Inverse element Open set Perspective (visual) Expected value Sign (mathematics) Lattice (group) Different (Kate Ryan album) Feasibility study Hill differential equation Physical system Social class Constraint (mathematics) Gradient Median Complete metric space Wave Arithmetic mean Figurate number Mathematician Spacetime Point (geometry) Momentum Wage labour Transformation (genetics) Amsterdam Ordnance Datum Routing Student's t-test Rule of inference Product (business) Frequency Latent heat Goodness of fit Term (mathematics) Modulform Computability theory Combinatorics Standard deviation Distribution (mathematics) Numerical digit Model theory Content (media) Algebraic structure Cartesian coordinate system Limit (category theory) Quadratic form Faculty (division) Formal power series Spring (hydrology) Loop (music) Universe (mathematics) Musical ensemble Table (information) Family Drag (physics) State of matter Direction (geometry) Multiplication sign Decision theory 1 (number) Hierarchy Set (mathematics) Insertion loss Parameter (computer programming) Mereology Prime number Food energy Order of magnitude Mathematics Coefficient of determination Many-sorted logic Analogy Matrix (mathematics) Position operator Decision tree learning Area Process (computing) Moment (mathematics) Physicalism Price index Flow separation Googol Quantum Right angle Resultant Thomas Bayes Geometry Ramification Classical physics Computer programming Trail Polynomial Functional (mathematics) Divisor Distance Theory Power (physics) Hypothesis Theorem Integer Curve fitting Time zone Dependent and independent variables Fields Medal ├ľkonometrie Graph (mathematics) Calculus Numerical analysis Inversion (music) Number theory Inverse problem Local ring Spectrum (functional analysis)
[Music]
well I'm interested in how you became
who you are today so I want to start at
the beginning and I'm gonna ask the
question based on the lack of knowledge
are you the son of scientists are you
the son of people interested in
mathematics or did you come out of
nowhere in regard to your subject
so I grew up in Schenectady New York it
was born there the Ellis Hospital may 6
1947 and my mother was a homemaker my
dad was a flexural engineer and worked
in the Navy and radar and worked at GE
Research Lab Schenectady as you may know
was a town established by GE for a lot
of production turbines and things like
that too but my dad worked at the
research lab which was and still is
there and he was very interested in the
new things computers as well as a lot of
the the radar and other things that he
were saying you didn't spring from
nowhere intellectually there was this in
the household there was already in the
householder where there was a interest
in science and technology were you an
only child I was the oldest of four so
all this before are you the only one
whose career when somewhat in the
direction of your father's
SoDo Mike myself well I got a brother
was going down top time he oldest yeah
and my brother is a retired marine
biology professor Oh my sister working
pharmacology and my youngest sister is
now a physical therapist in Seattle so
should I imagine a household strewn with
books on engineering and science or
actually he didn't bring it home it's
his particular interest there are a lot
of toys and various things having
educational things I remember the World
Book Encyclopedia you know which we all
looked looked out there and having a
number of electronic gadgets and toys
and that's all into so was that I
remember my mother although she was not
a technologist was always asking
questions so she was very very curious
Jenny she got us in the mode of you know
always questioning we're there at
whatever age you might have felt it
ambitions specific ambitions for the
children are you pretty well allowed to
follow your interests you
we'll follow they we followed our
interest pretty much there was no
expectation that you'd be a doctor or a
lawyer or a journalist or whatever in
fact when I was an undergraduate I
didn't know what I wanted to do and I
was you know using undergraduate program
as it was intended to to explore
different options
well we already see one expectation for
you and that is that you go to college
yes so that at least was in the air yes
yes yeah and both of my parents grew up
on farms in Michigan and and my dad
almost didn't go to college but was
encouraged by his nap teacher to do so
huh and my mom went to college and well
you've you've introduced the subject of
encouragement by teachers so I want you
in school let's say at 11:00
what's the school like what's the
preparation 11:00 would have been the
sixth-grade express like that so sixth
grade we had some excellent teachers I
remember a biology professor there that
was very very good and he would come
around with my abortion so on you
remember is they no no no mr. place I
think was mr. police was his name it was
that's not under them yeah yeah but yeah
the teachers were very good miss Keon I
had an astounding school system and I'm
always meeting people from Miss Kuna
have done well in their lives as the
teachers have been let's talk about that
so it's a suburb of a suburb of
Schenectady it's near the research lab
and it had many PhDs among the faculty
which is unusual for high school in a
school system high school is now in the
50s high schools 50s yeah sorry yes
and is this a time when a lot of them
are not able to get jobs in universities
or was it just the habit of the system
yeah you weren't in that was no task I
didn't I find since I went to high
school public high school in the fifties
although in California that I was
educated by a lot of very intelligent
women yeah who as women could not get
other jobs in the fields that interested
them they have also been the case yeah I
just didn't know what their personal
Sidhu does fair enough
well let's get you to middle school
junior high I'm tracking the point of a
particular spark it may not even happen
through high school but are you are
using a needle so junior high school in
seventh grade
seven yeah that's like seven eight or
whatever we're right we're terrible I
think as they are for many kids you're
going to be really what everybody's
misbehaving so it was nothing
particularly loss period nothing else is
happening right I do remember learning a
little algebra then but that was about
it in terms of so now I'm gonna put you
in high school are you any more mature
in your intellectual development I still
was much more interesting it was a
larger community it was much more
technically oriented a good very good
high school teachers were excellent I
remember some of the teachers teaching
writing teaching mathematics teaching
science I was it was a great experience
just a good high school there are you an
extraordinary student or just a good one
I'm a good student I'm a pretty good are
you good yeah so I remember I remember
being elected as treasurer of the class
because I was so good at math right okay
so I'm glad it's practical already in
your life wasn't calculus or anything
but it was if you're in an American high
school in the fifties you're getting
counseling as to where to go for
university what is that discussion like
I don't remember much of that I didn't
apply very many places didn't I remember
being courted by Michigan State which is
where my dad had gone to because I was a
National Merit Finalist and they were
trying to attract national drag my
enlist there and I ended up going to
Yale forgive liberal education which I
think was a good choice on the end heels
not a bad choice in most cases but one
doesn't even in the 50s I mean
competition every decade has gotten more
and more crazy but even in the 50s
you didn't just waltz into Yale so your
grades must have been pretty good Fred
good grades I had like perfect scores of
the SATs and things like that
okay perfect scores in the SATs some
indicator of something so I was doing
well in a connect academically and you
were doing well verbally as well as
mathematically both both sides of the
exercise yeah I enjoyed I enjoy the
writing side of things as well ah yeah
so that wasn't so usual yeah okay Yale
bids for you tries to interest you or
you just do I think I just went through
the standard application process and it
was except that they they have a
interview process where you know there's
some alum
local tech community interviewed you and
sees if you would be a good fit okay um
the nice thing about undergraduate life
in America is that the first year you
don't have to choose a major yet you're
now you're you're tasting the waters up
tasting the water you're traveling the
waters of tell me how you were beginning
to make decisions about your future
so well the nice thing about a school
like Yale this is a liberal school broad
spectrum of possible career paths and
interest you could take and and so I was
trying to decide I didn't know whether I
wanted to be a technologist or maybe a
lawyer
or maybe a psychologist or something
else but you're supposed to be wondering
it yeah and I enjoy taking the
professor's on film
you know and other things so it was a
good liberal education right I drifted
in the end towards mathematics and part
because the mathematics curriculum was
the least demanding and I could explore
all these other interests at the same
time huh so you know I took some classes
I didn't least demanding to the talented
obviously it was but you mean just in
terms of the course requirements course
requiring with you yeah so the course
requirements were I'm searching for the
first sign of a mentor or a particularly
inspiring figure him or her in the
mathematics department
no actually economics there were there
was a professor there by the name of
Richard Ruggles who hired me to work on
computer programming for the
econometrics Society and so I did that
several summers working on price indices
for Latin America and submitting decks
of punch cards to be run through the IBM
7090 or whatever was we had then to
compute price indices and so it was
technical programming work which I
enjoyed and got me more familiar with
computers um many of the people watching
this won't even be able to imagine the
state of computer life at the time that
you were an undergraduate Yale in terms
of what was available what was being
thought about it can you everybody
didn't exist as a major then so I didn't
have that as a choice if there had been
a computer science major than I thought
would have matrix would have done it
wasn't engineering I took some of the CS
courses that were in the engineering
program computers were just starting to
become part of the curriculum there was
there's a couple of courses in
programming that I took mostly they were
punch card based write punch cards are
hard to find these days my wife who does
a teaching these days was looking for a
deck of punch cards to show her kids
fourth graders today what a punch card
was just can't find that you can't find
them that's hard to find so they're
around but they're a great computer a
Museum in Boston there's like there is a
museum and they probably have some there
but she was looking for something to
bring into yes that they could actually
I would but there's also gonna food
machine with paper tape which you don't
see anymore at all so so should I not be
romantic and imagine you you begin to
have access to computers in this form it
interests you but there's not yet a
Eureka or something about you're sensing
the future it was something I was
drifting into I think because it was fun
it was interesting computers certainly
had a power that you didn't see and
other technologies is directly and
immediately even submit something and
get results back and it related to the
mathematics I was doing but there wasn't
any kinda remember any particular moment
and that's an interesting question but
of any particular moment said yes I want
to be a computer scientist although I
did drift into applying for computer
science graduate school which I must
have made a decision something exactly
if you ask me out what when did I decide
that that's what I wanted to do I
actually don't remember crossing that
threshold or just sort of drifted yeah
I'm doing more and more computer science
this looks like an interesting
discipline why don't I go try it you
wind up going to Stanford and we'll talk
about that obviously but I'm wondering
roughly how long Stanford has even had a
ph.d program in computer Stanford had
just started the PhD for that started so
I finished Yale in 69 yeah I think
Stanford started its program in 65 or
something though so it just have been a
few years have been a few PhDs out
you just sound to me less clueless than
your about what your future is going to
be because it's a fairly bold decision I
mean the other would have been to go to
Matt in mathematics yes well it was
partly tech field it was partly things
like let's go live in California for a
while - okay
Stanford should be interested so
Stanford at that point as you say just
started the the ph.d program in 65
what other do remember programs are
there around because you may have
applied to other programs or maybe
Stanford was one of the only ones to
exist so at that time I think I applied
to MIT as well it did not get in but
that's a good lesson for people to know
but and I think I applied a couple other
places I don't remember MIT
unfortunately has winter if you had
gotten into both MIT and Stanford where
would you have gone it's probably
Stanford anyway yeah probably yeah so
you're in Sanford what is the guidance
such as a PhD student gets to you as you
arrive in terms of this new field at
least field so they're establishing the
curriculum there they're putting
together courses I got to meet the
faculty you know Bob Floyd was my thesis
advisor he taught a marvelous lucky yes
last on algorithms Don Knuth was there
so Herman on many other people were just
just a fantastic group of faculty and so
part of it was the coursework and taking
the the wonderful courses they were
teaching and part of it was trying to
integrate into research and figure out
what kind of a research program you
wanted to do and and there was also at
the time just to set the context there
was also the Vietnam War and so I was
worried about the draft I'm trying to
figure out what I my life might be like
should that become a concern I did end
up working at the artificial
intelligence laboratory up in the hills
behind Stanford it's no longer part of
Stanford campus at least not part of the
teaching campus anyway and we had I
remember there was a cart there I think
it's a computer museum now where they
were
to talk about autonomous vehicles this
was one of the very first the Stanford
card trying to drive around the parking
lot and not hit any of the cars so I was
working on some of the coding for that
which allowed me to get a deferment for
a while anyway on the right the Vietnam
War situation I'm very interested in the
and the past others this the formation
of a community interested in questions
obviously Berkeley is across the bay are
you sensing as a graduate student a
larger discussion going on some of that
yeah there was for example one of the
early papers that I worked on was a
medium fast median finding algorithms
are given a set of numbers how do you
find the median of that yeah that was a
sparked by some insights that Manny Blum
had and Bailey's at Berkeley uh-huh and
connected with Bob Floyd and then some
of the graduate students and bond Pratt
Bob target and myself got involved as
well and so that that there is a larger
intellectual big in the community there
yeah as another example Don Knuth I used
to run I was it was a weekly or monthly
it's fairly often sessions at his house
where he would might speakers in and so
I remember dick carp coming down from
Berkeley talking about NP completeness
and so on to week so there was a
community was starting to it was a bit
of a distance but people would come back
and forth and I remember going up to
Berkeley wants to talk with some of the
researchers there so so there was some
back and forth between Stanford and
Berkeley but computer science was still
very small there's no industry you look
like you have now all right which had
its advantages yes because you could
talk to everyone yeah at that point yes
yes yeah um at the again you're doing
graduate work you're man who likes to
talk about ideas give me some insight
into how you're beginning to think in
terms of his direction to go and maybe
the opportunities are not vast or maybe
they are but how do we get you well
let's just say through to your
dissertation so my dissertation about
the work the research I did ended up
focusing on algorithms and I was
interested in sort of the combinatorics
the algorithmic content of how they got
to get computers to do complicated
things there are certain search
algorithms remember also to bring up a
side directly part of my thesis was on
search algorithms and I remember being
concerned at the time about the ethical
implications of this you know that if we
can make search so much more efficient
you know what what about surveillance
and what could the government do with
this and so yes so we see all this
resonating today with both the
government and Google and other
companies and actually throughout your
career yeah because and we will talk
more about the ethical issue because I
think you have clearly early
demonstrated and interests in the
ethical implications
interesting to me that it happened so
yeah so early in the process
um are you speculating at all this may
be a classic retrospective question
where we now know what's happening at
that time but at the time you may not
have known about the the future of this
that field only but effect on society
these perhaps it's hard to tell them you
would hindsight you can say anything but
yeah yeah you don't particularly
remember feeling you were now rushing
toward the future in a way perhaps one
way of setting context is to make it
just to say that I'm a big fan of
science fiction as well and so imagining
what the future could be like including
the future of computers if you know the
azimoff at Heinlein were big there and
other writers now but you know trying to
speculate as to where the field might be
going as a society or technologically
certainly always been a part of what I
do and so I think that's a part of also
part of the if you will the culture in
which you're operating professionally
you have any US I think I think that
that's for two I mean I think the AI
theme which has always been a part of
what I you - well not as strong as
perhaps some of the other things but
it's you know Frank can you build an
intelligent computer yes what was an
issue that arose certainly in graduate
school people were thinking about those
things and throughout my career I've
bounced back and forth thinking about
these things off and on but
the larger implications of what computer
science might have an impact on society
what can you do with computers is one of
the big questions still of the day right
right
with attendant fears and hopes yes like
are you are you at this point I'm also
interested in the relationship between
mathematics and computer theory but not
in its formal aspects although I'm
interested if you want to talk about
that as in the position somebody who had
elapsed mathematician although of course
mathematics is in computer era choosing
this field whether this was considered
odd or you had gone bad or you were in
worthy of mathematical theory I mean no
I think I think it's not like that at
all I think I think the the there was a
blending of these fields when I was at
Stanford Don Knuth was growing his group
of researchers there and to the people
in particular that I ended up spending a
lot of time with where David Connor was
a professor of combinatorics and washing
Shabbat both was also similarly working
in combinatorics and they taught courses
which related to the combinatorics of
algorithms and graphs and so on but also
talked to the algorithmic side of things
as well so I worked with it so I think
he saw a blending of these keys fields
more than any kind of in a way that that
leads me to the opposite question why
did computer theory break off into its
own program why didn't it stay within
the mathematics so computer theory is in
fact I mean it different places it's
different things I mean when I came to
MIT here there was an effort at one
point by the part of them on the part of
the mathematics department to take the
theory group out of computer science and
that's what I would move it in there we
declined that invitation but it was an
interesting one and nonetheless even so
right now and at MIT in computer science
theory we live within the computer csail
laboratory which is interdepartmental
and has mathematicians and computer
scientists both in it and we were side
by side
all the time so it's it's a it's clearly
a computer science theory as a feel that
spans both computer science and
mathematics right and people are happy
with that back and forth in the blending
your you've already gotten yourself to
MIT but that's fair enough the the the
the the PhD is well-received yeah I
think a PhD MPH sees a demonstration did
you've done some research and right so
that was but it wasn't a breakthrough
and it wasn't a breakthrough
okay it was any waves really good to
know because the stages of a career are
very interesting yeah all the time it
was probably badly written too I think
I've learned to write better since then
so even though you were not a bad writer
as an undergraduate well it did I think
it's the technical writing is hard it's
challenging okay oh why you have to put
your head in the mind of the reader
first of all and you have to be very
clear about the terms are using and that
they're all well defined and the things
that those do building the structure
right the usual challenge of writing is
trying to communicate the complicated
structure to an audience that right
doesn't know any of that the begin with
the goal is that lyricism the goal is
explanation that's right although you
know lyricism in the sense of
conciseness and elegance play a role in
computer science but not not poetry I'm
very struck this is really just aside a
question we don't have to belabor it but
just the use of words that are aesthetic
words that mathematicians and computer
theories which are a version of the same
use elegance of course and there is a an
act abuse of that language absolutely
yeah I know that the simplicity and then
trying to come up with theorems that are
cleanly stated and simple to understand
really matters in the field you can come
up with very complicated structures and
things that are true but interesting
because they're just too complicated and
your head around and so people look for
the simplicity right well you haven't
yet set the world on fire but you're not
doing badly you get your PhD what do you
do next
so after the PhD it was interesting the
ph.d program at Stanford had people from
all over the world including a number of
Frenchmen
and so geo Khan and Shanti malware two
of the colleagues I had there and they
said Ron why don't you come to a postdoc
in France afterwards I said that sounds
marvelous I love to travel I had taken
French in high school and I knew some
French I figured I could get along and
so I'd accepted we went there for a year
lived in Paris this is that inria
Institute Nationale de facie informatica
dominique right look at the accent my
French is mostly guys and so that was a
postdoc working with John female mostly
on algorithms and so on - it was an
interesting challenge one of the
interesting aspects of it was that they
hadn't told me when I was accepted but
the working language would be French so
I were there all day I'm talking French
all day and exhausted it but your French
was up to it got to got to be up to
Augusta and got to be after three mum we
know the language we know that you're
having a marvelous time in Paris
but what is the quality of the inquiry
there at that point so I think it was
good I was afraid the French Research
Lab was a first-rate research lab people
they were doing interesting work largely
theoretical and the work I was doing was
Zhang Jian I was or combinatorial Naga
rhythmic but it was it was good stuff I
was pleased with what we did there it
was gonna be a limited time there
because you you know you had to get on
with your career how do you make that
decision so it was a clearly a one-year
postdoc and so towards the end of that
year I have to go around the u.s. dude
it's like took time off in the spring to
travel all around Seattle and under San
Diego and Carnegie Mellon and MIT and
everywhere big loop around the country
try your what the opportunities were
right and I became persuaded that MIT
was the place education kind of offers
but MIT is its own argument for itself
but what about what they were doing at
that point that might have so there was
a good theory group here at MIT at the
time Albert Meier was was the the person
hide the main contact with Mike Fischer
was here at the time number of other
faculty were we're here then so it was a
clearly a place where a theoretician
could come in
work happily and and I think things
we're exciting at the time we the P
equals NP question had just started
bubbling up with further ways and people
were wondering could we've resolved that
and there was questions of circuit
complexity and just algorithmic
questions in general were we're of
interest of course in the course of your
career which will will will will get you
as it develops you wind up coming up
with insights with colleagues and so
forth which we'll talk about that have
very profound practical application but
the fellow who's just shown up at MIT is
a theorist who is or isn't interested in
the application of his ideas so mit has
a culture which is very much practice
oriented as well so they've played way
and when mi MIT likes theoreticians but
they also would like to see people spam
the bridge so there wasn't the the
practice theory gap and so there was
encouragement to to do that all
particularly the teaching that some of
the introductory and of course is I help
teach we're systems classes and things
like this so computer systems not but
not just the theory classes so there
were there was encouragement in those
directions I wasn't personally in terms
of the problems I was picking at that
time White's oriented towards the
practical stuff as I became later yes
but but it's in the end there was there
was encouragement motivation to think
broadly dinner in part just interact
with your colleagues better write in
part because you know you want to have
impact on society you've already made a
decision and maybe also the time the era
but you'll you'll tell me one way or the
other
not even to consider going into industry
I mean you you sound like you are on so
what an academic to reject yeah pretty
much one of the I mean there were some
research labs and I looked at Sandia
laboratories for example when I my big
tour of the country and there wasn't the
kind of industrial research labs there
are now so really the kind of questions
were interested in asking yeah you would
not have gone to uh there was no place
you know the industry didn't really
exist at the time I mean PC hadn't been
invented yeah it's the year that you
come to a mic so I come to MIT in 74
okay right the PC wasn't gonna happen
for another six years anyway or
something so is this whether we talk
about going to industry there really
wasn't an industry in any sense like
there is now it is just you know many
orders of magnitude different so know
tortured decision-making this was this
was a clear path yeah you know and I had
done teaching at Stanford I thought
summer courses I've done been a teaching
assistant for a number of terms I
enjoyed the teaching as well as the
research so I think the academic path
was pretty much clear we're very close
to a Eureka moment at this point that's
because the work year you're doing with
colleagues and you talked about the the
coming to the the insight in the end
that will mark your career really so so
the the work at MIT here was was
primarily algorithmic and characters are
looking for efficient ways of doing
things and a lot of the work of a
theoretician in computer science is
precisely this trying to determine which
problems you can solve it efficiently
and find a good algorithm for them if
you can and which problems are hard
intrinsically so it said that's
separating that we were there's just
hundreds of problems that you might want
to look at it turns out that you could
say some of them were easy some of them
are hard and some of them you're not
sure but they're clearly the same and
something that just recoding zuv the
same problem so that sifting out of
these various problems was was the bread
and butter of what was going on in
theoretical computer science at the time
yeah trying to figure out which problems
are easy on a computer which problems
are hard and so that was part of what I
was involved in looking for good
algorithms I was also thinking about
things like P equals I think at the time
to try to prove the certain problems
were hard well I didn't have the tools
of the time in fact that's still an open
problem uh very much so
so so that drifted into working with a
number of students on a variety of
things some of which involve things like
one-way functions and some of the
cryptographic things I was working with
a student by the name of Steven boy AK
who was was now
the NSA on you no inverse making matrice
industries and when when our matrix
inverse is easier to work with the
matrix themselves and things like this
so there was a lots of stuff in the air
about complexity applied to computation
which problems are easy which are hard
and then the Eureka moment is as you
said arose when Steven boy actually just
mentioned gave me a paper from Diffie
and Hellman which said you know New
Directions in cryptography and that
paper was really what changed my life in
many ways it said here's a set of
problems that we don't know how to solve
but which looked like they could have
theoretical interest in practical impact
and they're absolutely right they said
this this is a beautiful paper nicely
written and said here's the idea the
vision of a public-key crypto graphic
system and some ideas as to the kinds of
things that relate to that and what how
that might be achieved but they didn't
have a working solution and so that why
did we need a key system why-why-why
public key system yeah yeah yeah so the
vision was that everybody could make up
their own public keys and distribute
them publicly without the need for a
centralized approach it's a little bit
like the appeal that Bitcoin has today
where Bitcoin is a decentralized
cryptocurrency yes without having a
centralized issuing rights all right
well the public key vision is a bit like
that although in fact you need to have
some support for authenticating public
keys if you give me your public key how
do I know it's really you that's giving
it to me right so there's some of that
it aspect to it but it was a
decentralized flavor and it really fit
very well with the about to be born
ecommerce market rather than the
hierarchical sort of military situations
that exists very closely so it was a
different thing and one of the things
that I found most inspiring about the
diffie-hellman paper was their
discussion of digital signatures so you
can take a message and you can append
something that comes from you as clearly
from you can be verified as coming from
you right verifies that it's from you if
your
fyz that it's the content that that
message was was signed by you so it's
the electronic analog of a handwritten
signature and that was really novel I
wasn't just confidentiality it was a
sort of authentication that was achieved
there and that I found to be a exciting
notion as well and ahead as it turns out
many more ramifications down the road as
well so how do you dive into this I mean
how do you is there a particular problem
you embrace so they said basically you
know you need to have inverse problems
you need to have something that's easy
to do but hard to undo right so it's
sort of a one-way function the kind of
thing I've been talking with c-boy yak a
little bit on there and they gave some
ideas for some ideas but the world is
open you can take any kind of problem
you like and in fact we're still looking
at problems trying to see which one's
fit this this model right so you can
take a problem which is hard to hard to
invert basically turn it into a public
key cryptosystem with quantum and
computing and things like that nowadays
the question is what's hard to compute
make it may have changed the ground
rules may have changed over those days
we had conventional computers classical
conveyors at the time I was co-teaching
a class on discrete mathematics and we
were talking about number theory at the
time so that was very much in my mind at
the time and so we were looking at the
approaches that a number theory can
carry they say we so I got ID Jameer and
my Needleman involved uh sooner on this
product they said I'll you know I'd love
to talk with you guys about this suite
of problem you bring the question to
them I bring the question to them
I had this paper I said this is an
exciting paper we should we should think
about these questions and I started with
with Adi and he odd he's always
enthusiastic about new directions and
new problems so we started start with
him and then we brought let into it
audience and given the context of the
discussion we just had about math and
computer science yes Adi and LAN we're
in the mathematics department here at
MIT of time we had offices adjacent to
each other here in the lab for computer
sciences that was called but you know so
I was a computer scientist they were
mathematicians the laboratory was set up
to be interdisciplinary and really
achieved that purpose of it at this
point
mathematicians working together with
computer scientists and some since we
would all say were theoretical computer
scientists right right technically we're
different departments now there's no of
course simple answer to this but I'm
gonna ask the question too because I'm
interested also in just the process of
collaboration yeah how does it work in
this context I mean it's the three of
you that in the end came up with the
direction that Dell yeah no
collaboration as usual I mean
collaboration actually it's interesting
to talk about collaboration over the
decades because just a step back up yes
it's always about it one of the works I
did at Stanford as I mentioned earlier
was this fast median finding algorithm
there were five of us on the paper and I
remember the program committees saying
what is this you can't all be co-authors
you know you're just trying to get
travel money for the graduate students
involved or something like this it
really was caught you know co-authored
in a collaborative way but that was
unusual then collaboration is nowadays
much more a thing and much more common
and routine than it was back then at the
time Lenin Adi and I started working the
other three of us working together
that wasn't so uncommon but it wasn't I
mean single authored papers were perhaps
much more common then than they are now
so how do you how do you work on a
problem together you sit down you talk
about it what what are the constraints
what do we know what are the approaches
we generated lots of ideas about could
we use number theory could we use some
lattice based kinds of things can we use
some other kinds of thing other
constructions that we come up with there
are lots of ideas about how you might
try to do this so we just sort of like
generating ideas as to what might work
and mostly they didn't work this is a
silly question but it it hits on a more
important one and that is obviously
there's trust in the collaboration I
don't mean in terms of glory-seeking
although that's a factor it's in human
nature but separate from that does
somebody come up with an idea and I'm
simplifying it of course and the others
say that's the stupidest thing I've ever
heard but nobody disparages ideas you
know that's probably the wrong way to
put it but they're excited by that and
we'll
say you may have something cryptography
is interesting is you don't know whether
an idea is gonna work or not users say
here's a construct it looks you know
here's the way you would encrypt
something here's the way you would
decrypt something and then the key
question is can the adversary also
decrypt without the knowledge of the
private key right and so that's a
computational question which you may not
know the answer to and you may not so
even when we publish to the RSA paper W
you know we didn't know whether it would
ultimately be secure or not because you
know the key problems involved were open
problems and a cryptography is very much
like that the difficulty for the
adversary in breaking these schemes are
generally open problems they may be
problems that have been studied for a
while and look like they're harvest
Hoos are hard we don't know how to prove
problems are hard very well yet so
nobody says it's stupid or somebody says
now here's a way that you can probably
break it now you do this inadequately
it's breakable and when Adi and LAN and
I had this you've probably heard this
story before but you know we had sort of
a role different roles to play audio I
would come up with ideas more often and
Ladd would be the one to sing you know
that's not gonna work guys here's a way
to break it so he was expert at breaking
schemes that we had so we went through a
whole number of different ideas but not
even I did not to break the month so how
to take a part that's so it helps me
understand yeah how you would work
together also I guess implicit in what
you're saying is maybe one day with
different stage of computational
capacity and Sun and on what is now
unbreakable may turn out to be breakable
one yes absolutely the things change so
back back to the collaborating think for
a second I mean that's one of the
lessons that was learned during World
War two apparently was in Germany the
cryptographers who were making codes in
the Crypt analysts who were assessing
their security and daring to break them
were in different buildings in different
camps and they didn't communicate enough
and had they done so they might have
realized that their Enigma was was was
breakable Wow so I think we learned
today that it's very helpful to have
people working both sides of the fence
working together to build codes so
anyway that that was and then we're
talking about
I gotta track it with that well we're
really just developing the idea that
will make the difference yes we had to
we had a number of different ideas that
we looked at and the thing that turned
out to be RSA was not the first it was
you know the 40th or something right and
eventually you know I think it was me
that put the with the pieces together in
the particular way that ragaar say but
there were pieces that we'd all studied
carefully different framework something
so put it together so yeah this might
work so again it was another proposal
that one of you know 40 wanders right at
that right and it could have fallen down
like the others it was nice and clean
and simple so it seemed like it had a
nice structure to it that you know could
have turned out to be another something
for the Dustin if I had seen a way to
break that but you're developing
confidence as a group that this this is
something worth testing worth writing up
so that was writing yeah the way these
things work is you have an idea in the
field and cryptography these days you
publish it and you say you know we don't
see how to break this it seems like it's
here's what we can figure out about its
strengths it relates to these other
computational problems like factoring or
something like that and you say these
are this is what we know but you know
it's an open problem to assess the
security yeah it can you as a community
now figure out the imitation
Invitational to tell you the limitations
yeah yeah yeah yeah and in fact the
process for coming up with the new
crypto systems has changed over the
years to be one that's very much focused
on standardization and a community
effort so at the National Institute of
Standards technology has done a
marvelous job of running competitions
for the submission of it cryptographic
algorithms and and having community
conferences and community efforts to try
to break things and so that's what's
what's needed to really assess the
strength of it because we don't know how
to prove things are hard that that's the
problem with cryptography is we don't
know we don't have the technology it's
one of the big open problems in computer
science how do you show that a problem
is really
hard yeah let's go ideas on how to kind
of maybe do it fast but it's it's a
can't show that it's gonna you know
how's it hard why is it hard my word
case where skates how do you there Holly
it's hard on the average how do you this
paper works omitted it actually is an
answer that people have been looking for
it seems to seems to have worked well
still stood the test of time so the
paper came out it was it was a proposal
everything yes you know we said this is
well we know this is what we think we
can do with this it doesn't seem to
answer the question the Diffie and
Hellman raise and they're wonderful
paper New Directions cryptography it
interested in the framework that they
gave and it really said now there are
new things you can do in cryptography so
here's one that's an idea that's based
on factoring that allows you to achieve
a public key cryptosystem not only the
encryption side but also the digital
signature side would write which was
really relies on particular did you hear
from them by the way we talk we talk to
them and then they they were you know
interested in appreciative right
supportive things I need to get you now
from this point well to what we've been
talking about the realization that this
might actually be a solution to the
implementation and the effect of the
field yeah so it was the the
implementation issues were interesting
because at the time computers were slow
they were very much slower than we had
today I mean other 10,000 times slower
so you know and so the process of
finding prime numbers even which is an
essential part of the IRS I think when
you multiply two large prime numbers
together yeah finding those prime
numbers could take a half an hour on a
computer or something like this yes it
was Norman that wasn't the numbers that
are considered short by today's standard
so you know those have changed over time
with Moore's law and computers getting
faster but at the time the
implementation was actually a serious
issue this was a proposal that in some
sense it was feasible because of the
algorithms involved were all polynomial
time
if you look at how much time was
actually required by the polynomials
you know it took too much time to be a
really a practical interest yet that
that would soon change with as computers
get faster well one of the things we did
early on was to get interested in chip
design so MIT at the time was in the
process of building up a capability in
VLSI fabrication so designing and
implementing large-scale chips to do
computations of various sorts so we
actually put time in about 1980 to
design and implement a prototype or a
chip that would do the RSA computation
huh would find a large prime numbers
that would do the encryptions onto with
special-purpose circuitry like that you
can actually make it very feasible in
the end of snow most special-purpose
chips aren't needed so much because the
chips are pretty fast now but it was it
did the job
I mean it demonstrated feasibility the
time difference it was also an
interesting just intellectual experience
for us - how do you design a circuit
chip Justin I'm gonna get real
practically practical here and ask about
the economic implications of this
because they feel and the have proven
profound in terms of your even patently
this this concept and perhaps presiding
over in the private as well as the
public world its implementation how do
you think about that so at the time we
didn't know much about startups I think
the culture of computer science
departments has changed hugely since
then at the time there were a few
startups but not many if you try to find
colleagues who had done startups or
things like that right it's not like
Stanford where everybody's out of their
office golf during there's your star
today or something so but we said yeah
this looks interesting it looks usable
and MIT was supportive of efforts like
this so they said well this sure will
pad it and then we said well the set up
a little startup company - yeah there
might be some applications I never
application sketched in the new
directions paper we were thinking of
some - but mind you the
worldwide web and not yet been invented
and so all of the applications that flow
out of that
were still in the future and during the
80s when when the company was trying to
get going it was all very difficult
because there were no large-scale
applications like you would get by
having Amazon sales or something like
that or ecommerce in general so from the
time that we invented the system through
the 80s up until the invention of the
World Wide Web it was pretty much a
barren market markets had to be created
when the web happened things changed
like that it was it was an amazing
transformation all of a sudden everybody
was using computers in the web to do
commerce transactions had to be verified
they wanted to be encrypted often you
wanted to authenticate things you needed
digital signatures so things changed
remarkably with the invention of the web
if you read David Kahn's book about the
history of cryptography he talks about
one of the large impetuses to
cryptography earlier which was the
invention of radio and during World War
1 radio was used for commanders to talk
to their jerseys - but everything was
broadcast so people could listen in you
had to have cryptography so the
invention of radio was was a first
impetus for cryptography the world wide
web is really the second that sort of
the same kind of area nice parallel so a
short layman's version of this is you
didn't own the rights at a point where
it was paying off in a big way right
right and you lose the rights and the
natural course of a patent right that
expired right housing or something like
that so it's it's gonna allow me to
introduce a very important thing in your
life which is even the ethics of this
much of your approach in general anyway
has been little fasten flowers bloom you
know let as many people into the use of
this as I can that I prepared to say
yeah I think so I mean I think for this
particular scheme we founded a company
and there was a patent because I think
that's the only way to
get the energy in the investments and
the food right to get that out there but
then again the goal was to market that
did you say a lot of thousand
applications being yeah with the support
of this company and the software
provided by this company you can have
lots of applications using it it wasn't
trying to be restricted was trying to
get things out there I really want to
belabor because we don't have a lot of
time this Labor's not the right word the
the ethical questions that come into a
researchers life your kind of research I
guess part of that is the whole question
of who has access to who can use it but
also in the course of your life you come
up with ideas that you've deliberately
not taken possession of so to speak that
you've wanted used and make decisions
that allow it to happen widest use can
you talk about that and because it is
ethical look the question is well so I
think the ethics of cryptography are
very interesting mostly the community
runs in an open manner like you're
talking about where people publish
designs and so on - occasionally they
will patent and so on - but I think that
the effort in cryptography these days
has been pretty much through the
standardization route people won't use
cryptography unless they're standardized
and and so that that's the route that
people take now that's the recommended
path you want a cryptographic algorithm
you look for the US national standards
you say what what are the algorithms are
or published well that said back to the
ethical issues of cryptography one of
the big issues then and even now is
government access to encrypted data and
so I've been involved with that debate
since the early days in the 90s the
government tried to encourage everybody
to use a certain encryption chip which
wouldn't not only encrypt the data but
also allow them access to the plaintext
right right and that didn't go well and
I was happy the other side to encourage
that it not be adopted and today we see
the US government pushing for access to
iPhones where Apple and Facebook are
saying you know this if you try to put
backdoors
- these phones you will weaken the
overall infrastructure be worse for
security over all the benefits you get
aren't equal to the losses that you will
see you will suffer and so the ethics of
cryptography these are important policy
issues they deserve debate their
profound but but you know I think the my
taster the answers are pretty clear that
you we need good technology we need good
strength particularly with foreign state
actors attacking us left and right not
only an efficient space but election
space and so on - having good technology
to protect us without the backdoors
built-in this is what we want to be the
last question I'll ask because there's
so much in your career we could talk
about but you've tried to contribute to
voting questions as well in the in
ethical yeah perspective can you just
end the conversation with a little bit
so I think I drifted in the direction of
looking for questions where technology
comes to play and policy is relevant so
this encryption debate is one of the
areas encryption sorry elections are
another one where technology is coming
out we can see how to protect elections
better than we used to know how to be
able to do things like risk limiting
audits the use of cryptography turn out
to be powerful tools for securing
elections yes and we'd like to see more
of that so and that involves not only
technology but also policy because
people have to understand adopt and
implement these kinds of things and and
as you see on the books here at the
table I guess that you know I get I
guess I I'm attracted to areas where
there's interesting policy questions
climate change being the latest area
that I've been looking at and it's not
you are you getting a response I guess
with the distinction of what you've
achieved people do tend to listen a bit
but you have to say no are you
encouraged by I think it helps there's a
lot of pushback on the part of the
fossil fuel industry and so on to saying
you know this is a hoax or whatever but
I think that by and large the science is
winning this debate and
I'm happy to participate in the sciences
winning this debate is a great way to
have thank you very much
I did it dog with you
you
Loading...
Feedback

Timings

  949 ms - page object

Version

AV-Portal 3.16.0 (9cfa3864b8acb689056f9c67aa39bc8ec4c75d58)
hidden