00:00welcome to the a 16z podcast I'm Michael
00:02Copeland getting denied another round of
00:05NSF funding in the early days of mosaic
00:07turned out to be a huge catalyst to
00:10start a company around the fledgling web
00:12browser says Marc Andreessen that
00:15company was Netscape at the time and
00:18Driessen was still at the University of
00:20Illinois and he wanted the NSF money to
00:22help build what amounted to a customer
00:24support team that wasn't the NSF's
00:27business since and recent mosaic days
00:30calibrating the interplay between
00:32academia government and the private
00:34sector has gotten if not easier less
00:37exotic with schools like UC Berkeley and
00:39Stanford setting the standard for
00:42providing students and faculty with a
00:44clear path forward in this segment of
00:46the a 16z podcast Marc Andreessen and
00:49Chris Dixon discussed the role academia
00:51plays in the startup world from picking
00:54the right classes to picking the right
00:56institution from which to turn research
00:58into a company the conversation you're
01:00about to hear was recorded at our 2015
01:04academic roundtable Chris Dixon starts
01:07things off so let's let's talk about
01:10them so you you way back were a
01:13recipient of government grants right
01:16Netscape came out of that how can you
01:20talk about sort of how your I guess how
01:23you see the role of academia as it
01:25relates to startups venture capital and
01:28how that's changed over the last over
01:30your over your career
01:31yeah so first of all it's great great to
01:33see everybody back this year for those
01:35of you who are returning we're thrilled
01:37you're all here so so my work at
01:39University of Illinois around mosaic
01:41which later became that scape was was
01:43all and it was NSF funded and so I owe
01:45the NSF a huge you should have gratitude
01:47for that I also owe them a huge debt of
01:49gratitude for something else which was
01:51turning down the additional funding that
01:53we requested and one of my fondest
01:57mementos is the is the cover sheet of
02:00the NSF proposal with the decline on it
02:03and it was literally four by they were
02:05by the way they were completely right it
02:06was a we we literally had so many people
02:09early on using mosaic that we were dying
02:11support load you have mean customer
02:14support he's an overstatement cuz they
02:15were paying for it so this it was user
02:16user support and so we applied for you
02:20know being in university it's like whoa
02:21go get more funding so we applied for
02:23NSF funding but basically build a
02:24customer support team and they informed
02:26us very kindly but that wasn't part of
02:28what NSF funds so that was a good
02:31catalyst to to go start a company which
02:35was which was very helpful so you know
02:37there's been obviously just you know a
02:39lot of things have changed in the in the
02:41last you know this is 23 years ago now
02:4322 years ago so so many many things have
02:45changed probably my single biggest thing
02:48that I think has changed that we see
02:49with academic computer science and
02:51venture capital probably the biggest
02:53change and I think this holds generally
02:55I know this holds for my alma mater and
02:57I think this holds more generally when I
03:00was getting my computer science degree
03:01at Illinois in the late 80s early 90s
03:03the department sent just an overpowering
03:06message to the undergrads that the
03:08purpose of the department was to MIT
03:10PhDs and future professors and that
03:13industry was a very kind of you know
03:15lower-class you know sideshow
03:18you know dead-end kind of thing and they
03:22were very very clear on that and that I
03:24think had a lot to do with obviously the
03:26selection of the material you know with
03:28a with a much greater focus on theory
03:30than practice and then I think I believe
03:32it even had a selection on things like
03:33programming languages for the coursework
03:34I think the faculty actually went out of
03:37their way to pick languages that would
03:39never ever be useful in a production
03:40environment and so I knew Pascal and
03:43scheme really well it's sure not to be
03:48so the biggest change and I guess you
03:51know people could probably argue both
03:52sides of this but the the biggest change
03:53is just we have a sense for a lot of
03:55universities we deal with if the
03:56computer science departments have a much
03:58I mean they still want to want to breed
03:59PhDs and professors but a much bigger
04:02focus and practical impact in industry
04:04and you know the students that we see
04:05coming out generally have a large amount
04:07of practical skill in addition to
04:09theoretical skill which i think has been
04:11a dramatic change what do you think
04:17that people say that Silicon Valley
04:20isn't working on big problems that maybe
04:23you know I don't know only you know we
04:26used to have the space program we used
04:28to have we were promised self-driving
04:31cars and we got Twitter you went to the
04:33moon yeah what's fun so what do you
04:36think about that do you feel like the
04:37computer I guess both the computer
04:39science academic community and the in
04:41industry is tackling big problems yeah
04:44so there's kind of two critiques right
04:46now in the media and it kind of popular
04:48discussions around this topic but apply
04:49to tech critique number one is techs are
04:52not working on big problems you know
04:53it's basically all you know everything
04:54is Silicon Valley is these silly little
04:55apps and like why don't we take on the
04:57hard problems and of course the other
04:58critique is we're like tech is having
05:00way too big an impact on our culture and
05:02society and like throwing everything
05:03about people and destroying all the jobs
05:05and reordering all the industries and
05:06changing the culture and just having
05:09this disastrous impact and you know the
05:10impact needs to slow down
05:12nobody attempts to ever reconcile those
05:16two critiques and the same commentators
05:18will literally write both critiques in
05:20like different columns you know two
05:21weeks apart without ever attempting to
05:23reconcile them and so and when I call
05:25them on it they basically say yeah yeah
05:27but it's all consistent because you the
05:29tech industry is having a huge industry
05:30it's just all negative which I think is
05:33maybe just slightly too cynical of a
05:34view so I would say I'm kind of a little
05:38bit of a split mindfulness myself which
05:40is I do think it's it's unfair and
05:42inaccurate to say the tech industry and
05:44computers in computer science as a field
05:45it's not tackling big problems and I
05:48think that it's just it's obvious you
05:49know when you when you look around you
05:51know the nature of a lot of the things
05:53the people are working on are you know
05:55really go after you know really kind of
05:56foundational things like I think
05:57communication is actually a really
05:59foundational thing in terms of how our
06:01civilization works you know obviously
06:03money financial services that's being
06:05reordered with technology now logistics
06:07how the real world works transportation
06:09real estate are being reordered you know
06:11culture society you know it's being
06:14deeply affected by technology and so I
06:15think that there are there are a lot of
06:17big problems actually being tackled at
06:19the same time I think it's a fair
06:21critique and I think this is the part of
06:23Peter Thiel's critique that I agree with
06:24it's a fair critique that there are many
06:26fields that are not moving as fast as
06:27they could and in particular fields you
06:31occasions in computer science or other
06:32fields of engineering that intersect
06:33more in the world of atoms as opposed to
06:35the world of bits and so you know drug
06:37discovery is an obvious one
06:39you know advances you know in mechanical
06:40engineering advances in you know space
06:42travel you know lots of different areas
06:45where you would say you know Trenton
06:47cars where you would say boy you know
06:49you would think that we could make more
06:50rapid progress I'm a glass half-full
06:53kind of guy so I look at that and I kind
06:55of say we now have the opportunity to
06:56make progress in many of those fields
06:57and in fact the best way to make
06:59progress in many of those fields is to
07:00apply more computer science and so I
07:02count myself as an optimist but you know
07:05nevertheless I do think there you know
07:06there is something to the critique like
07:08people are quite capable of looking at
07:09the huge advance that has been made in
07:10their smartphone over time and then
07:12looking at the advance that's been made
07:13in their kitchen or in their car and I
07:15think you can kind of see that there's
07:16been a difference in the rate of
07:17improvement and it's an exciting
07:19prospect to be able to go tackle some of
07:20these other fields so going back to the
07:23academic question so state obviously
07:25Stanford and Berkeley and some other
07:27schools have been especially successful
07:28at at seeding startups that go on to be
07:34very successful why do you think that I
07:36mean obviously some of its proximity to
07:38industry and being here but there seems
07:40to be other factors like can you and I
07:42think specifically also for people here
07:43who who are thinking about how they can
07:45make their own universities more
07:47entrepreneurial like what do you think
07:49the key lessons are yeah so I'm you know
07:51I'm a I'm a case study of this in a lot
07:52of ways you know in theory I could have
07:54started my company out of Illinois and
07:56practice I think that was Impractical at
07:58the time may or may not still be
08:00impractical but was definitely practical
08:01at the time so I'm a M&E important you
08:04know I'm a classic import to Silicon
08:06I speak with the zealot Ori of the
08:07converted you know I've adopted Stanford
08:10for a lot of the work that I do out here
08:13and so I you know sort of things that I
08:15have a reasonably good understanding of
08:16the the Delta between the approaches I
08:19think a lot of it has become obvious
08:20over the last couple years I would
08:22highlight a couple things that I think
08:24Stanford and Berkeley do particularly
08:25well I think one is there is and again
08:29for better for worse but I think for
08:30better there's very deep connectivity
08:32between Stanford and Berkeley and them
08:33and then the valley at a ground level
08:35and so there's just a very porous
08:38barrier for the professors and for the
08:41students and for the administration
08:44and you know administration is an
08:46example like we all you know people like
08:47we all know the the leadership is
08:49Stanford they spend spend a lot of time
08:51in fact you know it's not an accident
08:53that the Stanford picked is their
08:54president some time back you know John
08:55Hennessy who in addition to being a
08:58legendary theoretical computer scientist
09:00it was also himself a former company
09:01founder and so and you know maybe the
09:04best university president anybody's ever
09:07you know that's worked very well the
09:10other thing that I think Stanford in
09:11particular has done really well is
09:13Stanford has been what I would consider
09:15to be the most enlightened in the sense
09:17of understanding the full kind of cycle
09:20of life of ideas being incubated in your
09:23university and then companies being
09:24formed and then in the fullness of time
09:26the wealth creation that can happen
09:28through company formation and success
09:30and then the philanthropy they can flow
09:32back into the university and if you walk
09:34around the Stanford campus you know it's
09:36not an accident that there was some Jim
09:37Clark building right and there's the
09:39Jerry Yang Building and there's the but
09:40you know there's the Bill Gates building
09:41and you know it's just building after
09:43building after building the Hewlett you
09:44know the Hewlett Packard guys did a lot
09:46my father-in-law actually not in
09:49computer science in real estate but is
09:50another example of that went to Stanford
09:52on a scholarship my father-in-law went
09:54to Stanford on a geography scholarship
09:56the last year Stanford taught geography
09:59he cancelled the major after he
10:02graduated but it was a good idea to keep
10:05the major until he graduated because he
10:06went on to become a very successful real
10:07estate developer and has donated what
10:09has to be at this point more than a
10:11billion dollars back to Stanford
10:12personally and there's you know 150
10:15buildings in the Stanford campus that
10:16he's paid for and so Stanford has a very
10:19kind of deep understanding that's
10:21evolved over time of the fact and I
10:23would say in contrast universities where
10:24the IP licensing office is dominant and
10:26where I think it's really a different I
10:28would describe it as a different
10:29business model and know people people
10:31really understand that which is give it
10:33away in the beginning and then make the
10:35money back on philanthropy as opposed to
10:38I don't know I won't say which
10:40universities but like some other ones
10:42that I have friends who've interacted
10:44with that they have a very they're they
10:47think the value lives and patents and
10:49transactional and licensing
10:52I mean we like we you know for example
10:56when we make investments we ask about
10:58patents but it's like a side item like
11:00it's never actually Accord any
11:02investment because anyone who's built a
11:04company knows it's a dynamic process
11:05you're constantly building and any tech
11:07you start with it's probably completely
11:09different in four year or five years
11:11it's all about who you recruited and the
11:14people you recruit tend to not like
11:15patents they like open source software
11:17they like they you know they like you
11:21know cool open innovation or something
11:23like this and I don't know so I I it
11:26feels to me like a very I don't know
11:27like like a lot of people just really
11:29misunderstand how it works yeah and it
11:31feels like a lot of em you know there's
11:32there's there's legislation there's
11:34things if they forget the name there's a
11:36you know there's a law on this that
11:38people are grappling with and so it's a
11:39complicated topic but you know my
11:41University of Illinois is an example my
11:43alma mater I think in the last twenty
11:44years it's made significant progress and
11:46they now have the the results to show
11:48for it so the the first thing was the
11:50the Beckman Center which appeared when I
11:51was there which is kind of it's just
11:53like absolutely amazing complex up on
11:55the north side of campus from Arnold
11:56Beckman who was a legendary company
11:58founder who came out of Illinois decades
11:59ago and then Tom Siebel you know has
12:02basically rebuilt the Illinois
12:03engineering campus in a very similar
12:04model and so I think you know this is an
12:07area in which success should lead to
12:08success as you have more and more
12:09examples of how this happens you should
12:12be able to you know to be able to see
12:13things I think also the the model that
12:15works in medical in the medical sphere
12:16and in pharma licensing is more
12:19transactional patent oriented and does
12:21seem to work and is very different in
12:23computer science and that throws people
12:24off but yeah computer science education
12:28so New York City just announced that
12:32they're going to in the next ten years
12:34start teaching computer science and all
12:37I think I think elementary and high
12:40schools for the most part it's very I
12:42think it's it's a computer science is
12:44rarely taught at the pre-college level I
12:47think and it's as I understand you know
12:51the the the we were constantly
12:53complaining about there being a shortage
12:54of computer science expertise what do
12:58you think's going on there and what can
12:59we do to fix it yeah so I mean there
13:02there is a cynical view by the way I
13:03mean by default you would say we never
13:04kid get taught computer
13:05Sciences of plus there is a cynical view
13:07that says that if they teach computer
13:08science to school they'll beat curiosity
13:10on the topic out of the students just
13:12like they do every other topic they
13:13teach and so there is there is a
13:16potential dark side I hope that's not
13:17what happens the positive side seems you
13:21know overwhelmingly positive which is
13:23you know they talk about like a
13:24foundational technology of our time you
13:26know to learn about how software works
13:27and how to build software you know it's
13:29really fundamental so it you know it's
13:31incredibly exciting obviously it's a big
13:33dependency it's good for kids to have it
13:36you know familiarity with it it's you
13:37know you then get an interesting
13:39questions about how many you know how
13:40many computer science students can
13:41university to take right and then what's
13:43the dropout rate along the way and you
13:45know you have all these problems around
13:46underrepresented groups you know with
13:48computer science degrees and so if you
13:50still have the drop left take place in
13:52the university you know then you may not
13:54you may not fix that problem but it's
13:56certainly progress and then I think
13:58there's a deeper idea that I think is
14:00we're thinking about which is the the
14:01impact that computer science is going to
14:03have in many other fields both in the
14:05Academy and also in industry over time
14:07and something I think a lot about
14:09there's a famous essay on by an
14:12Englishman named CP snow from the 1960s
14:16I think it's a famous essay you can do
14:19but if you haven't seen it it's it's
14:21it's about what he calls the two
14:22cultures it's CP snow was this very
14:25interesting character brief digression
14:27CP snows interesting character because
14:28he was both a chemist and a novelist and
14:30so he kind of was right brain and left
14:32brain and he understood both worlds and
14:35he wrote this essay kind of at the
14:37height of the Cold War and at that
14:39specifically at the height of physics
14:40being kind of the top scientific field
14:42of its era and so you know nuclear
14:44energy and space travel and the atomic
14:46bomb and the hydrogen bomb and all these
14:47you know incredibly central topics
14:49around physics and he wrote this essay
14:51if you read the essay it's like you know
14:52sixty years later if you read the essay
14:53and if you just substitute physics for
14:55physics you just take physics out you
14:56put computer science in it reads exact
14:58right and he talks about the two
15:01cultures he basically says there's the
15:02I'll adapt that he says there's the
15:04computer science culture of the
15:05engineering culture which is kind of
15:07ascendant culturally because this
15:10science in this engineering you know is
15:12reordering the world and having a huge
15:14impact and all the physics people then
15:16computer science people narrow all cocky
15:18and aggressive and confident
15:19say all these bold things and then
15:21there's the other culture which is the
15:23artistic culture the liberal arts
15:24culture and the liberal arts culture you
15:27know as you know art literature music
15:28and philosophy and political science and
15:30all these things sociology and and
15:32they're very much on the defensive and
15:34they feel very you know attacked and
15:36victimized and you know are the days of
15:40liberal arts over because engineering is
15:41just gonna take everything over and so
15:43he comes at account from a sociological
15:45standpoint and it's a it's a it's a fun
15:47diagnosis to read but then he proposes
15:49what he calls the third culture which
15:51are the people who can bring the two the
15:52two cultures together and the people who
15:54can bring you know physics or in our
15:56computer science into liberal arts and
15:58the people in liberal arts who can learn
16:01and understand even if they're not
16:02engineers can learn and understand how
16:03engineering works and how computer
16:05science works and he basically proposes
16:07that the third culture will be able to
16:08do things that each of the cultures by
16:10themselves will be unable to do and I
16:12really think we haven't we collectively
16:14have an opportunity to do that with
16:15computer science we have an opportunity
16:17for computer science to have a hugely
16:19positive impact on many other fields of
16:20human activity and we have the
16:23opportunity to have computer science be
16:24something that is open and accessible to
16:26people who aren't going to be full-time
16:28programmers but who are gonna be able to
16:29learn about this understand the
16:31mentality and then be able to you know
16:33really understand what's happening and
16:34be able to contribute and so my hope
16:37would be that that's what will you know
16:39flow from these investments in earlier
16:41CS education and if and if you had to
16:44pick areas just some examples of what
16:45computer science might you know because
16:48of computer science plus X what would X
16:50be yeah I mean so the you know the
16:52obvious giant one right now is by all
16:53these biology and life sciences and it
16:55just it just seems to us like there's a
16:57revolution afoot in a fundamentally new
16:59way it's just extraordinary exciting you
17:02know no you know it's not really two
17:04cultures thing cuz you know biology is
17:05also you know about as close to
17:07engineering than engineering the liberal
17:09arts to start with you know but for sure
17:11that and then I think in Liberal Arts I
17:13think more and when you see a lot of
17:14this on you know Stanford is doing a lot
17:16of this a lot of other universities are
17:17as well but you know there's it's
17:19everything it's you know literature
17:20there's there's new ways of thinking
17:21about literature the written word
17:23there's no ways to think about music
17:24there's no ways of thinking about arts
17:26you know one of the really interesting
17:27things happening right now are the
17:29attempts to digitize you know things
17:31like ancient ruins and
17:32works in in in in regions of the world
17:34that have lots of warrant conflict so
17:37even in the worst case if they get
17:38destroyed we're able to have like a
17:39complete 3d recreation of like an
17:40ancient city and so there's there's a
17:43potential to you know kind of really
17:44advance cultural knowledge and
17:46understanding entertainment you know is
17:49obviously a straightforward one
17:51education itself you know software
17:54driven education you know with all the
17:56tools and techniques we have in computer
17:58science applied to education you know it
17:59seems like a huge opportunity let's
18:01let's talk my real bio is something that
18:03we spent a lot of time on lately like
18:05what you specifically you know is
18:09happening there that makes it it makes
18:11it an exciting time yeah so I think the
18:13core foundational thing that's happening
18:14is something very subtle and very
18:16important and this is happening other
18:17fields as well but we're definitely
18:18seeing it in bio which is you know
18:22biologists like physicists or chemists
18:24or a lot of other you know sort of
18:26highly advanced specialists in different
18:29different areas of science and
18:30engineering you know up until ten years
18:31ago if you would meet you know kind of a
18:33state of the art you know world-class
18:35research biologist odds are they weren't
18:37very comfortable computers and odds are
18:40they had never really programmed and in
18:42fact in physics in example my job in
18:44college my first job in college was
18:45actually a write computer code for
18:47physicists who hated computers and it's
18:50been the same thing of biology you know
18:52for a long time and you meet a lot of
18:53senior biologists who are just still
18:54sort of fundamentally uncomfortable with
18:56a lot of this stuff or they have a grad
18:57student who writes you know who writes
18:59writes the code for the lab you know the
19:01big thing has happened is there's just
19:03so many now young incredibly smart you
19:06know biology phd's doctors chemistry
19:08PhDs who are coming out and it turns out
19:10in addition to being fully qualified in
19:12those fields they've also been
19:13programming in many cases since age 10
19:16and so they've got that same foundation
19:19in software and computers you know they
19:20had a PC in the house growing up you
19:22know they've had a smartphone you know
19:23they're really young ones avetis you
19:24know it smartphone for most of their
19:26life and they've got the same kind of
19:29you know foundational knowledge about
19:31computer science and software that a lot
19:33of the kids in computers and computer
19:34science have because they started
19:36programming it at age ten and so I think
19:38it's actually a first and foremost is to
19:40change in the field which is the the
19:43consequence in generational change the
19:46other fields it's just going to change
19:47as a consequence of that and then you
19:50know you'd have to add to that the the
19:51more formal efforts front
19:53interdisciplinary research at the
19:54university level and then there's the
19:56big macro trends actually happening in
19:57the science and so you know the
19:59realization of genomics is a mature
20:00field you know the the you know the the
20:03enormous advantages of cloud computing
20:04and big data you know now being able to
20:06be applied to biology you know the you
20:10know computational biomedicine all these
20:11sort of fundamental areas of things that
20:14can be done now quite you know quantify
20:15itself as an example we think is going
20:17to evolve into you know a cornerstone of
20:19algae in the future you know we you know
20:21we one of our theories is we think we're
20:22living in the Stone Age today in the
20:24sense of like we really don't know
20:25what's happening on our bodies like we
20:26don't know what's happening in our
20:27bodies until something goes wrong and
20:28then we get whatever tests we get and
20:29then if something goes right again we
20:31kind of go back to our state of
20:32ignorance whereas in the future I just
20:35think it's gonna be standard that you're
20:36gonna know everything is happening in
20:37your body and you're gonna know your
20:39blood work and you're gonna know your
20:40genome and you're gonna know your biome
20:41and you're gonna know your MRI and
20:42you're gonna know all these things all
20:45and so there's you know an enormous turn
20:47that's gonna happen as a consequence of
20:48that so just to get specific if you're a
20:53professor in computer science and you
20:55want to encourage you want to get more
20:57involved in entrepreneurship or
20:59encourage your students to or your
21:00school to you know what can we mean it's
21:02part of what we're doing with this
21:03conference is to try to kind of increase
21:06the communication I guess but what what
21:08can what I guess advice would you have
21:11or kind of suggestions or etc for people
21:16in the audience yeah so you know in any
21:17given I was a couple things sort of in
21:19it any give an error in the valley there
21:20are a set of venture capital firms that
21:22are kind of front and center and you
21:23know we've kind of try to make ourselves
21:24you know one of those but you know there
21:26it will gradually concede that there are
21:28others and you know and you know there's
21:32there's you know three four five six ten
21:34whatever the number is but a firms that
21:35are kind of on the leading edge funding
21:36the next generation of interesting
21:38companies and so in those firms as we do
21:41tend to have a very panoramic view of
21:43what of what's happening in industry and
21:44so I think we're you know I think we're
21:47thrilled to have you guys here but I
21:48think we're you know we're able to be a
21:49resource on that and you know certain
21:50extent the other firms may be and then
21:53there's a set of companies you know
21:55there's there's there's a like to say
21:57like they're you know there's a lot of
21:58there's a lot of tech companies
22:00in any given generation of companies
22:01there's three or four or five that are
22:03kind of clearly top of the heap you know
22:06kind of have you know they're kind of at
22:08scale and are doing very interesting
22:09important things I have not become kind
22:12of classic big companies they just kind
22:14of drift along you know sort of
22:15companies that are still alert and alive
22:17you know often still remember their
22:19founders and are you know hiring
22:21voraciously right and so they're just
22:22you know they're just a sink for talent
22:24you know coming out of all of your
22:25programs and so I think between you know
22:28the top handful of venture capital firms
22:29the top handful of tech companies I
22:31think you know these days it's it's you
22:33know the those are the key things that
22:34you want to really be you know even if
22:36there's no formal conductivity just you
22:37know kink is associated you know
22:39understand know people let those at
22:41those firms and at those companies and
22:44then you know the other I'm sure this is
22:46obvious but just works it works really
22:50is just the natural flow of students
22:52coming out in the industry and then
22:54retaining conductivity back into their
22:55programs and you know being brought back
22:57to campus you know bringing the
22:59learnings back you know telling the next
23:00generation of students you know what the
23:02opportunities are you know being guided
23:04by you know the professors who are up to
23:06speed on what's happening out here about
23:07about where to go and what to do and we
23:09you know we say we see lots of success
23:11stories of how this works and works
23:13really well and then you know we do see
23:15cases where you know there are programs
23:17where there's just they're just
23:17completely you know still isolated after
23:20all this time and the students really
23:21have very little idea of what's
23:22happening out here or in industry
23:23broadly and so for those of you who are
23:26doing that well it's it's going great
23:27I think that mentality could be applied
23:30more comprehensively in the field at a
23:32lot more schools a lot of the best tech
23:34companies in the world today flow
23:35directly out of I mean my company's
23:37example this a lot of the best examples
23:39flow directly out of universities and so
23:40and we have a whole generation these
23:41companies you know that have done this
23:43and so they're there there is a very
23:45good success kind of model it works
23:47incredibly well there's a very clear
23:49failure case which is the dominant
23:51failure case that I think that we see
23:52which is the professor has the idea the
23:57professor starts the company but with no
23:59intention of going full time at the
24:01company the professor runs the company
24:03or gets it started for a year or on an
24:05interim basis around a part-time basis
24:08gets associated with some seed investor
24:10or some second-tier venture capital firm
24:11that goes out and finds a professional
24:13mediocre professional CEO the professor
24:17then is like okay my job is done
24:19professor goes back to teaching and then
24:22the company just drifts and ultimately
24:23falls over and dies and so and so what's
24:26missing there what's missing there is is
24:28basically this is just a lesson we
24:29learned over and over again like there
24:31is no substitute in these companies for
24:33having the core team and in particular
24:35the CEO the person who's gonna run the
24:38company the people who are really strong
24:41and really sharp and really you know
24:43clued in on Mondays to happen and then
24:45are really full-time like really able to
24:47be full-time and be able to be full-time
24:49on a sustained basis isn't it to me the
24:50professor that often it's a it's a it's
24:52like the key grad students right I mean
24:53exactly exactly so that so it's gonna
24:55say is that that's that's exactly right
24:57so that that's the model the the
24:58variation that that works really well is
25:01when there's one or ideally a set of
25:03students set of grad students who really
25:05understand this and really want to do it
25:07and the professor basically sponsors you
25:09know if it's sort of informally sponsors
25:11the creation of the company or helps
25:13with the creation of the company and
25:15then the students actually run the
25:16company the professor can be involved as
25:18an advisor or can be involved you know
25:19the board of directors or whatever the
25:20right thing is or can just go back and
25:22you know on teach you know but and go
25:24back and do more research but where the
25:26company has really formed around around
25:28the students and you know Silicon
25:30Graphics was a you know back back of it
25:32which was a huge success
25:33you know what this model was was an
25:34exact example of that you know a
25:36Netscape was another there's been a
25:37whole bunch of examples like that and so
25:40it's almost the professor as the Sherpa
25:42and as the as the advisor and it's the
25:44guiding light and the professor might be
25:46very involved for the first year with
25:47the idea and the fundraising and
25:48catalyzing the entire thing but where
25:50there there's this core of people who
25:52are the students who are really able to
25:54pick this up and carry for Miss Sarah
25:55for uh Sarah was a recent example of
25:57this where you know Scott you know Scott
25:59and Nick were very involved in the
26:01formation of the company but you know
26:03they found they found a good person to
26:05work with to be the CEO and then the the
26:07CTO was Martine who was the the top cred
26:09stood in the field who really cared the
26:11company forward and Martine now actually
26:13is a big executive at VMware and runs I
26:15don't know a thousand people and has the
26:16five hundred million dollar business
26:17working for him and so that's a
26:20particularly vivid kind of clear success
26:21story and I think that's that's the kind
26:23of precedent to look at
26:25so yeah I mean I would argue it's
26:26actually a special case of a bra of a
26:29broader misunderstanding of startups
26:31that sculpted that outside of academia
26:33as well and it comes from pop culture
26:36you know you see the Facebook movie and
26:37they like write an equation on the board
26:39and it's as if that equation were like
26:41the secret right when in fact it's 10
26:43years of tons of engineering and
26:46marketing and all sorts of other things
26:47right and network effects and the you
26:49know the million other things or you
26:51know the the idea that you there's I
26:54didn't see the movie but it's like the
26:55guy who invents the intermittent
26:56windshield wipers you know so you invent
26:58this idea you patent it you hand the
27:00patent over to the business guy who then
27:02goes the next in the next scene he's
27:04like living in a mansion right like it's
27:05like and so and it's sort of the pop
27:08culture view of it and when in fact what
27:10startups really are is a dynamic process
27:13that probably in you know 90% of it is
27:16frankly recruiting a great product and
27:19engineering team and what we find is
27:21only it just just empirically is it only
27:23great product engineering people are
27:26able to recruit those people and the
27:28product ends up changing a lot over time
27:30I almost never see I don't know I got an
27:34open eye Sarah for example but I've been
27:35involved with many startups where it
27:37came from academia but when you actually
27:39look at the product four years later it
27:40was very very different I mean some of
27:41the core ideas might have been there but
27:43there were a lot of changes and that
27:45were non-trivial that you needed you
27:47know sort of the core people to do I
27:48mean yeah that's your experience so we
27:50funded the service so we funded us here
27:51at Omega Kia's networking lab and it was
27:53the open flow open flow was that the
27:54sort of protocol was the approach and
27:56Martine had had been one of the main
27:58people doing that and so they starting
28:00to see her all excited cuz we're like
28:01it's great we've got open flow we've got
28:03the open flow technology we've got the
28:04open flow team like this is gonna be the
28:05open flow company and the first board
28:07meeting Martine sets this down he's like
28:09okay the first thing guys need to
28:10understand is we're throwing away all
28:11that open flow stuff and he said you
28:15know that was great for a research lab
28:17but like what for commercial
28:18applications we need to do another we
28:20need to do another version we need to
28:21take everything we learn from that
28:22research that we now need to build the
28:23actual commercial product and so even
28:26Google like PageRank people talk about
28:27PageRank I mean it was very quickly
28:29copied by the idea of using inbound
28:32the ranking very quickly copied by the
28:34competitors they actually did a lot of
28:35other thing it's one of like PageRank is
28:37like one of like 900 factors now that go
28:39into a Google search result the other
28:41899 happened after they left Stanford
28:43now by the way there are professors Jim
28:45Clark was my partner when I started
28:47Netscape he started sold in graphics he
28:48was actually a case of a professor who
28:50left he left academia and went and
28:52started a company and feel like the two
28:54exceptions are RSA and Qualcomm maybe
28:56where there actually is sort of a core I
28:59mean if you agree but like there's a few
29:02exceptions where it actually does happen
29:03it always in no like Instagram you know
29:05there's a couple he's a really
29:06high-profile case where the guy just
29:07does have an epiphany and it becomes
29:09this and it and it takes it what
29:11unfortunately it unwinds like five years
29:12of us trying to argue because they're
29:16very high-profile cases but usually
29:18happen but for the most part there is no
29:20substitute there is no such thing right
29:22this applies to everything we fund it's
29:23what we work on all the time there is no
29:25substitute to the first five years the
29:27people in the office 18 hours a day six
29:29raining away and it's grinding away I'm
29:32making a product work for customers and
29:33they're working with the customers like
29:34it's just there there's there's no way
29:35we we have not found any way to shortcut
29:37that yet and so however the process of
29:41starting a company forms it has to
29:42really focus in on that can you talk a
29:44bit about the leadership development
29:45piece so on a day-to-day basis as a
29:47professor I am mentoring and helping
29:49students to develop technically but also
29:51it's a core commitment of mine to help
29:53them to develop personally as well too
29:55can you talk a bit about how we in our
29:57mentoring rules can help to develop
29:58people who will eventually be strong
30:00leaders who work technically competent
30:02but in addition to that also like
30:04relationally competent in terms of being
30:06able to be that technical CEO found a
30:08person yes so I think there's two sides
30:11I'm glad you asked that so I think
30:12there's two sides I think there's the
30:13soft side there's sort of assays the
30:15informal side of leadership and then the
30:16formal side of leadership so the
30:18informal side is just being able to work
30:20with people and being able to lead
30:21people and I think a lot of that has to
30:22do with what then you know what that
30:23happens in the lab or in the in the
30:25department and so you know having the
30:28high potential people be able to be in
30:29some kind of leadership role you know on
30:31projects or on research programs you
30:34know relatively early you know is
30:36certainly gonna be a good thing easier
30:38said than done but you know giving
30:39people the opportunity to lead and then
30:41giving people the coaching along the way
30:43or being able to find mentors who can
30:44come in and work with
30:45developing they're sort of the sort of
30:47methods soft skills but like the
30:48informal skills the people skills you
30:50know that you're kind of getting at
30:51there's another side of it that is also
30:55I think a big opportunity which is the
30:57formal side which is business skills
30:59training for engineers and for computer
31:02scientists one of the things that is
31:05very exciting actually and this is I
31:08know there other schools you doing this
31:09but we see this most vividly in Stanford
31:11and at Berkeley Stanford and Berkeley
31:14now both have organized programs to
31:16teach and students in the engineering
31:19school business skills and this is this
31:22is the funniest the comical version of
31:24this is what happens at Stanford which
31:26is there's this sort of time-honored
31:27tradition at Stanford where the Stanford
31:29Business School students the MBAs you
31:31know the the the new MBAs kind of view
31:33themselves as future CEOs and so you
31:36know they'll come up with an idea and
31:37then they'll go try to basically you
31:38know they'll go to the computer lab at
31:40like midnight with a pizza and try to
31:41get like an engineer to help them on the
31:42idea and there was just a Dilbert
31:45cartoon and they kind of immortalized
31:46this where the pointy-haired boss you
31:48know goes up to Dilbert and says I've
31:50got a great idea for startup and now all
31:51I need is an engineer and some funding
31:53and Dilbert says you know the economic
31:55term for what you have is nothing so the
32:01the the Stanford version of that
32:02literally is the MBA trying to go get
32:03the engineer the much better model is
32:06for the engineer to have the business
32:08skills for the engineer to be a top
32:10flight engineer but also have the
32:11business skills to be able to start a
32:12company and then the engineer hires the
32:14MBA as you know the head of marketing or
32:16the head of sales or the head of finance
32:17right and so we would always prefer that
32:19the engineers have the business skills
32:21so that they can be in the leadership
32:22position you know in these companies
32:24even if even at the founding level you
32:27know some schools are very resistant to
32:29this idea you know that engineers should
32:30be trained in business because it you
32:32know again it seems like a corruption of
32:33the process of what of what of what
32:35they're trying to do in their in their
32:36field some schools have gotten very
32:38enlightened on this what's so striking
32:40about the Stanford and Berkeley programs
32:42is those programs are completely
32:43separate from the business school but we
32:45see this you know we'll go over and
32:46speak at Stanford and now you know
32:47you'll get you'll get an invitation to
32:48go speak in the business school
32:49and you'll get the invitation to speak
32:50at the business class for the
32:52engineering school and it's a completely
32:54different set of students they're all
32:55engineers don't want to learn
32:56business and so I think really good idea
33:00you know I don't mean it's like I'm not
33:01saying it's like half in half but I'm
33:02saying it's like you know I don't know
33:04four or five courses you know over the
33:05course of four years to be able to learn
33:08you know the fundamentals of business to
33:09be able to learn you know maybe a
33:11management course maybe a start-up
33:12course maybe a finance course and maybe
33:15one or two other classes like an econ
33:17course because then you can you with
33:20just a little bit of starting knowledge
33:21like that you can set the engineers up
33:22to be able to think about think of
33:24themselves as business people out of the
33:25gate I mean III was I I did it I guess I
33:28did it the hard way which is I didn't
33:29take any business you never even
33:30occurred to me to take a business course
33:31they never offered and so everything I
33:34learned about business happened after I
33:35graduated and I think in retrospect had
33:37I had any formal training in it I think
33:38I would have been better I mean the
33:40other the other good model I think is to
33:41go to startup a relatively small
33:43start-up for a couple of years or in
33:46turn I mean obviously I in turn
33:47internships are like waterloo's a great
33:49example of that were they I think they
33:51they have it's like it's like they have
33:53a crazy number like six or plus
33:55internships over there four years of
33:57undergrad the whole sort of it's deeply
34:01integrated sort of working in industry
34:02and working in in and their coursework
34:05and they come out they we see them
34:07coming out of college and it's just like
34:09they're just very sophisticated another
34:10good example Penn's M&T program so they
34:14combined right it's like some smaller I
34:17think it's mostly engineering but some
34:18portion of Wharton yeah no no that's I
34:21um on the internship side this may be
34:23obvious but it's very important for
34:25what's happening out here which is
34:25college recruiting for computer science
34:27students has gotten to be brutally
34:29difficult these because there's so many
34:30tech companies and there's just so few
34:32top-flight computer science departments
34:33and so you know it's it's World War
34:35three the way the companies are looking
34:36at it is they're just desperate to be
34:37able to hire your students when they
34:38graduate and what they've now learned is
34:41they can't wait until your students
34:42graduate they have to they have to get
34:43them they have to get them sooner or
34:44they'll never have a shot of getting
34:45them and so they have to get them at the
34:47intern have to get them at the intern
34:48level and so in the last five years
34:51alone in the valley the tech companies
34:53have hugely increased their focus on
34:54their internship programs basically the
34:56goal is to get all of your best students
34:58as interns and then be able to basically
34:59get them locked in at least a year in
35:01advance to be able to come on full-time
35:03the good news of that for you guys is
35:05like you know and I don't know the tech
35:06companies are being fully transparent
35:07with you on this but like their espera
35:09they're really desperate
35:11and so I think you guys have the
35:13opportunity to really set your students
35:14up in these in these best of class
35:16companies and and in fact not just one
35:18but like a whole series of them over the
35:20course of time that was one of things
35:22and then on the Waterloo plane that was
35:24what I think silent when I was in
35:24Eleanor's one of the things I give them
35:25a lot of credit for is they also had
35:27they did that they call it the co-op
35:28program but they would you know they
35:31would there was a they would support you
35:32and going out for think I worked at one
35:35of my stints was a full nine months
35:36summer and then the fall semester at IBM
35:39which was just a you know for me as a as
35:42a kid was just a huge you know be able
35:43to spend nine months in a company you
35:45know it was just I learned just an
35:46enormous amount back to the topic of
35:49patents and IP for a moment very
35:51specific question but probably a lot of
35:53us in here are building companies based
35:54on work that we did at the University
35:57that is perhaps patented at the
35:58University and the university owns so we
36:00have to license it from them
36:02I'm just curious on your thoughts about
36:04this is this sort of a red flag or
36:06problematic from an investor's point of
36:08view or not so much I'll give my answer
36:11and then to Chris thing so it depends on
36:13the University and it depends on the
36:14terms so there's there lot I mean we do
36:16fund companies all the time that have
36:17licenses to University patents I would
36:20say in some cases those are actually
36:21useful licenses like but I would argue
36:23is like the further down the stack the
36:24more useful the patents are like patents
36:26at the level of chips or at radios you
36:28know can be or you know I don't know
36:30memory or something like that can be
36:31incredibly valuable patents at the level
36:34of software or applications we generally
36:36don't do is very very valuable and I
36:37could go into you know more detail why
36:39but generally not very valuable you know
36:41to be able to get the right team out of
36:43the university into a start-up if
36:44there's a licensing agreement that is
36:46only on the margin for the economics you
36:48know then it's just a costume doing
36:50business and we'll kind of put up with
36:52it I have seen cases and I've lived
36:55through cases where the universities
36:56take a very draconian view on this and
36:58the university licensing office has this
36:59just highly inflated view of the value
37:01of what they're gonna be licensing and
37:03they they can fill companies in their
37:04cradle by taking that approach so I
37:06think a lot of it depends on the
37:07specifics of the university yeah I would
37:10just add I think that some of the
37:12universities are doing themselves we
37:14believe a disservice by by doing this
37:18counter-intuitively I mean
37:21to them it may be counterintuitive we
37:24think they're actually generating less
37:25revenue by asking for more but it's it's
37:29it's such a it's such a counterintuitive
37:30argument that it's very hard to make
37:32Nick from Nick McEwen from I Sarah
37:35actually he spoke I thought was really
37:36interesting that he was two years ago at
37:38our academic round table and he's a
37:41professor at Stanford he said that he I
37:44thought it was an amazing talk he said
37:45he literally with every new thing that
37:47his lab would invent they would go to
37:49the incumbent company like in his case
37:52Cisco like it was a new networking
37:53invention and literally just offer them
37:55for free all of the inventions and he
37:58said sometimes they would say yes in
37:59which case we found let's go to our next
38:00thing doesn't meant something else
38:02and when they say no he'd say ok well
38:04that might be interesting because it
38:05might be like too futuristic Francisco
38:07and therefore let's go do it and
38:09apparently he said Stanford just let him
38:11do it and didn't I don't think I think
38:13they're I don't know exactly how it
38:14works Stanford but it sounds from
38:15everything I hear I guess some Dan you
38:17might know but you are okay so so yeah
38:22so I think it's very liberal as my
38:23understanding and they they then in that
38:28case in my Sarah went and did this
38:29company and you know it's that there's a
38:31very different kind of philosophy and
38:32it's worked very well it's just very
38:35hard I know like I've tried to I went to
38:37Columbia undergrad and I've tried to
38:39argue this philosophy to the
38:40administrators there and they did is
38:42just very hard argument to make because
38:43it sounds very counterintuitive as
38:45people it's also outdated legislation
38:47there's the build the bayh-dole act
38:49which actually requires University
38:52exactly what it is but it's like
38:53university research is paid for by the
38:55public there has to be the the financial
38:58rights have to accrue back to the public
38:59and so a lot of universities are still
39:02grappling with that soon I think a
39:04really good reform would be to just
39:05eliminate that law and let universities
39:06make these decisions in their own but
39:08that's not happening anytime soon
39:12today about sort of a contrarian view
39:15investing and I'm sort of pointed out
39:17the investment in tachyon Nexus of you
39:20know many people saying oh that storage
39:23is not going to get as flat and when it
39:26does then tachyon is gonna be a good
39:27place so having a contrarian view for
39:30investing he was sort of espousing as an
39:32important sort of concept for the firm
39:34how is it that the firm and embodying
39:38you guys are sort of talking about your
39:40worldviews and your your lenses and yet
39:44you know fairly openly and yet you still
39:46have the ability to sort of find the
39:49things and identify the things that that
39:51others aren't so yeah talk about the
39:55your theory on this which think of the
39:59strong weak strong weak yeah this is
40:03you're asking like this is a whole book
40:05here material probably we can talk about
40:07but um well I say a couple of things
40:10like one is venture capital works very
40:12differently than the public stock
40:14markets for example where it's not here
40:16we just sort of decide ok let's invest
40:17in this company this company a lot of it
40:19is frankly the entrepreneurs drive the
40:20process and decide who do they want to
40:22work with so it's just that that's
40:24probably a common mistake Outsiders make
40:26about this industry and they think you
40:28know Chris had this great idea of like
40:30I'm gonna go invest in X or Y instead
40:32there's some of that like you have to be
40:34smart about what you want to invest it
40:36but frankly a lot of what we do is we
40:38try to build an institution that's that
40:40just is the most attractive place the
40:42daunt refers want to come and so a lot
40:43of the businesses frankly about that and
40:45and a lot of the investments we've made
40:48that have been successful our
40:49investments other VCS wanted to make and
40:51we just the entrepreneur chose to work
40:53with I mean it so that's it's it's kind
40:54of reversed from it's not how the public
40:57markets work you know the the farmer
40:59stock doesn't choose the hedge fund
41:00right like it's the opposite so so that
41:04that's a big part of it I would say I
41:05think the other thing is that you could
41:08talk about your broad ideas you know
41:10mobile phones will change you know are
41:13changing the way that you know services
41:16are provided or something it's also it's
41:17very different to find the right set of
41:19entrepreneurs working on a specific idea
41:21often Peter Thiel has talked about how a
41:25lot of these you know like Airbnb is a
41:26great example where very early on it
41:28looked like a really kind of ridiculous
41:29idea a lot of these things you know you
41:32can you can speak publicly broadly about
41:34that you have to be contrary but like
41:36actually in the moment figuring out kind
41:40the puzzle of a particular startup and
41:44whether it makes sense and the team
41:45makes sense and the timing makes sense
41:46ends up being very applied it's probably
41:49like in like more like almost like an
41:51engineering problem like you can talk
41:52about lots of broad principles and still
41:54going and actually building the program
41:56is a is a very different thing and so so
42:00Chris likes to say that basically like
42:01so maybe the advantage big companies
42:04have invented companies have they have
42:05all these resources that you know Cisco
42:07and Intel and all these companies does
42:08the engineers think all this money and
42:09they got all these people and it's just
42:11all these customers in this brand the
42:12Salesforce they have all these huge
42:13natural advantages and so if there are
42:15any good ideas that are floating around
42:17they're just gonna go do them and so for
42:20example Apple you know is doing great
42:21and smart phones and so a really good
42:23idea would be a smart phone that's like
42:24you know half is you know half as thick
42:26and has a battery life twice as long
42:27like that's a really good idea people
42:29really love that Apple fully understands
42:31that and they're spending billions of
42:32dollars trying to make that happen so we
42:34can't fund a start up against that good
42:36idea we basically have the disadvantage
42:41there's a consequence that we can't do
42:43though so the idea is that we can find
42:44are the ideas that look like that ideas
42:45right there the idea is that just look
42:47silly or stupid specifically to the big
42:49companies and so and then of course the
42:52the twist to it is out of the universe
42:55of ideas that look like bad ideas most
42:56of them actually are bad ideas which is
42:59something you can very easily learn the
43:00hard way and so you know and we see you
43:04know 2,000 startups a year and you know
43:06many many of them are good ideas that
43:07are just too obvious and then many of
43:09them are bad ideas that are just bad
43:10ideas and then every once in a while
43:12there's a good idea that looks like a
43:13bad idea and it's the thing that's just
43:15like counterintuitive and it's radical
43:18and you know and it's it's you know it
43:21by the way even there we're gonna have
43:24we assume we basically have a 50 percent
43:25failure rate so you know we assume that
43:27half the time you know we're gonna back
43:29the what we think is a good idea it
43:31looks like a bad idea and it's gonna
43:32turn out to actually been a bad idea the
43:34good news with the good ideas that look
43:36like bad ideas is when they work that's
43:38how you can build a major new company
43:40right because then that's the psychotic
43:42psychology under which you can get a
43:43jump on all the other big incumbents
43:44well interesting to watch when you're
43:47actually involved in it is they people
43:48continue to think it's a bad idea for a
43:50I'm like you know like Twitter got
43:52really big before people finally stopped
43:54saying it was about like tweeting what
43:56you're eating for lunch today like it's
43:57it's amazing how long people will go on
43:59saying it I mean like VR is a good
44:00example where I just it's very hard to
44:03go see a demo of the latest oculus stuff
44:05and not think it's amazing it's pretty
44:09much my experience of one-to-one
44:10correlation between having tried the
44:12greatest latest demos and being excited
44:13about the field and yet it's confidence
44:17highly controversial even within Silicon
44:18Valley you know so and you know so
44:21that's the other thing is you get it's
44:23it's kind of surprises you how long in
44:25my opinion how much Headroom you get on
44:28these good ideas they'll think that
44:30ideas like until us you know companies
44:32almost going public or generating
44:34millions in revenue like people still
44:36doubt it heavily and you're kind of in
44:38this weird state as as the startup as
44:40well as as the venture firm you're kind
44:41of in this weird state where you want
44:43the world to understand that it's a good
44:44idea because you want them to like buy
44:46it you want them to be everybody become
44:47a customer of it but you also kind of
44:48want them to continue to think it's a
44:49bad idea so that you don't face direct
44:51competition and so it's this really
44:53weird schizophrenic you know and we
44:55can't help ourselves like we're out
44:56evangelizing on behalf of our companies
44:58but ya know we have stuff that we're
45:00like we have stuff that's going
45:01incredibly well we're you know if I
45:02tweet about it I just get this immediate
45:03backlash well that's really you know
45:05that's so stupid like you guys are so
45:06stupid how can you possibly think that
45:07that's really I mean we've come any that
45:09are like you know 100 million in revenue
45:11and I'm like fanatic followings and you
45:14get you tweet about them and people are
45:15like well that'll no don't ever buy that
45:16you just okay and you can just it's
45:19amazing actually you can just like lay
45:20out the whole business plan you can
45:21explain how everything works and most
45:24people just think it's so ridiculous
45:25that it this day like they take Twitter
45:27as an example where it's it's for anyone
45:30the tech world and in the journalism
45:32world for example it's a critical piece
45:34it's a critical business tool right it's
45:36how we share links and how we share work
45:38information publicly in the tech world
45:40and like to this day I think that you
45:43know and it's a billion dollar revenue
45:44business built in less than 10 years and
45:46and to this day I most of the mainstream
45:48press coverage still kind of tweet
45:50treats it as you know like it's the
45:51silly thing for you know talking about
45:53what you had for lunch or some things
45:55you know and then it could go away at
45:56any moment so we can't complain about it
45:58too much because it is it is our little
46:00secret a little secret weapon yeah okay
46:03well thanks is it good good thanks