2.4s
the first of our Stanford AI Clubs 2026-2027 speaker series. It's a privilege to welcome Alfred Lin, a partner at Sequoia Capital. For over 15 years at Sequoia, Alfred has partnered with some of the most consequential companies of the past decade, including Airbnb, DoorDash, Citadel Securities, Cal Shi, Zipline, and etc. Before Sequoia, Alfred served as a
4.3s
2026-2027 speaker series. It's a
7.0s
privilege to welcome Alfred Lin, a
8.7s
partner at Sequoia Capital. For over 15
11.2s
years at Sequoia, Alfred has partnered
13.5s
with some of the most consequential
15.3s
companies of the past decade, including
17.8s
Airbnb, DoorDash, Citadel Securities,
20.8s
Cal Shi, Zipline, and etc.
23.3s
Before Sequoia, Alfred served as a
25.4s
chairman, COO, and CFO of Zappos, helping seal the company through its 1.2 billion acquisition to Amazon. He's also both a Harvard and Stanford alum, and we'll be having audience QA if you want to ask him which school is better want to ask him which school is better later. And he has a background in mathematics
28.8s
helping seal the company through its
30.3s
1.2 billion acquisition to Amazon. He's
33.7s
also both a Harvard and Stanford alum,
36.5s
and we'll be having audience QA if you
38.6s
want to ask him which school is better
40.0s
want to ask him which school is better later.
41.2s
And he has a background in mathematics
43.5s
and statistics. So everyone, please welcome me IN JOINING ALFRED. welcome me IN JOINING ALFRED. applause I PROMISE TO be here for as long as you guys have good interesting questions. But we're going to cut it off as as soon as you stop having interesting as you stop having interesting questions. The onus is on you guys. No pressure.
45.4s
welcome me IN JOINING ALFRED.
47.2s
welcome me IN JOINING ALFRED. applause
54.2s
I PROMISE TO be here for as long as you
56.0s
guys have good interesting questions.
58.5s
But we're going to cut it off as as soon
60.1s
as you stop having interesting
61.2s
as you stop having interesting questions.
62.1s
The onus is on you guys.
63.2s
No pressure.
64.0s
I think first question, a lot of the people in this room know you for your amazing work at Zappos and then all of the stuff you're doing at Sequoia, but I'd imagine that not a lot of people know about what came before that. So you were you were grew up in Taiwan
66.5s
people in this room know you for your
68.6s
amazing work at Zappos and then all of
70.2s
the stuff you're doing at Sequoia, but
72.3s
I'd imagine that not a lot of people
73.6s
know about what came before that.
75.8s
So you were you were grew up in Taiwan
79.2s
and you moved to New York as a kid and so we were wondering, you know, can you tell us a little bit about your upbringing and what that was like and how that shaped you today? Yeah, sure. You know, well I don't usually talk about this, but I grew up with two tiger parents. They were both commercial
80.6s
so we were wondering, you know, can you
82.4s
tell us a little bit about your
83.4s
upbringing and what that was like and
85.0s
how that shaped you today?
87.0s
Yeah, sure.
88.2s
You know, well I don't usually talk
89.3s
about this, but I grew up with two tiger
92.1s
parents. They were both commercial
93.7s
bankers. They valued education a lot. They had a hierarchy of the things that were acceptable for me to do. The highest was to be become a professor and contribute to the world's knowledge. If you couldn't do that, then be a doctor and cure disease. If you couldn't do that, be a lawyer and fight for justice. And
95.5s
They had a
96.8s
hierarchy of the things that were
98.2s
acceptable for me to do. The highest was
100.4s
to be become a professor and contribute
102.3s
to the world's knowledge. If you
103.6s
couldn't do that, then be a doctor and
106.0s
cure disease. If you couldn't do that,
107.9s
be a lawyer and fight for justice. And
109.6s
all the way at the bottom, because they lend to businesses, was to become a business person. And if you were in business, you were supposed to work with large companies, not small companies. And I basically chose to do the exact opposite of my parents' advice. Um because partly I saw them
112.3s
lend to businesses, was to become a
114.8s
business person. And if you were in
117.5s
business, you were supposed to work with
119.3s
large companies, not small companies.
121.1s
And I basically chose to do the exact
124.2s
opposite of my parents' advice.
126.4s
Um because partly I saw them
131.0s
basically built a life and a career in Taiwan, and they were fairly successful. And they decided to just risk it all, come to the US to give their kids a better life. And it all worked out, even though they they took all these risks. And my encouragement for generally people in this audience and for people in colleges,
134.2s
Taiwan, and they were fairly successful.
136.4s
And they decided to just risk it all,
138.6s
come to the US to give their kids a
140.6s
better life. And it all worked out, even
143.2s
though they they took all these risks.
145.4s
And my encouragement for generally
148.6s
people in this audience and for people
150.8s
in colleges,
152.2s
you're in this great situation, you're in a privileged situation to take risk. And especially in an environment where we have scaling laws, and the biggest mega trend that we've seen uh since the internet, since cloud, since broadband, you should take risks. And if it doesn't work out, you can always fall back on
153.8s
in a privileged situation to take risk.
157.0s
And especially in an environment where
160.2s
we have scaling laws, and the biggest
162.8s
mega trend that we've seen
164.8s
uh since the internet, since cloud,
167.6s
since broadband,
169.9s
you should take risks. And if it doesn't
171.8s
work out, you can always fall back on
173.6s
the fact that you have a freaking Stanford degree. Um I was expecting to get some laughs off of that. But wasn't so funny, huh? But wasn't so funny, huh? Um I was growing up, I always sort of felt I was growing up, I always sort of felt that there there are always like things that
174.7s
Stanford degree.
176.3s
Um I was expecting to get some laughs
178.9s
off of that.
180.9s
But wasn't so funny, huh?
183.2s
But wasn't so funny, huh? Um
184.0s
I was growing up, I always sort of felt
186.4s
I was growing up, I always sort of felt that
187.8s
there there are always like things that
189.4s
are just a little bit off about the world. And you should notice these things. If there are things that you notice that's a little bit off, it's usually a good thing that it's a start of a idea to solve a problem in the world. Every problem has a solution, every solution creates other problems,
190.8s
world. And you should notice these
192.5s
things. If there are things that you
194.0s
notice that's a little bit off, it's
195.5s
usually a good thing that it's a start
197.3s
of a idea to solve a problem in the
199.7s
world. Every problem has a solution,
201.9s
every solution creates other problems,
204.6s
and it just keeps going on and on and and it just keeps going on and on and on. I encourage you to find the problems that you want to keep solving, and the next set of problems that you want to keep solving, whatever that area is. Figure that out, because then you can
206.3s
and it just keeps going on and on and on.
207.3s
I encourage you to find the problems
208.9s
that you want to keep solving, and the
211.2s
next set of problems that you want to
212.4s
keep solving, whatever that area is.
215.0s
Figure that out, because then you can
217.0s
dedicate your life to being the best in the world at solving that set of the world at solving that set of problems, and then being the the the person that sort of becomes world class at it, be passionate about it, and then figure out the economic engine that will drive the solution for that um
219.0s
the world at solving that set of
220.2s
the world at solving that set of problems,
221.3s
and then being the the
223.2s
the person that sort of becomes world
224.6s
class at it, be passionate about it, and
226.9s
then figure out the economic engine that
228.9s
will drive the solution for that um
232.8s
for that um those problems. And every single company I work with, whether you know, when I dropped out of my PhD program at Stanford to join this company Link Exchange during the start of the internet. They solved these problems that other people thought were a little silly, a little crazy, and it became a
235.5s
single company I work with, whether you
237.8s
know, when I
239.0s
dropped out of my PhD program at
240.6s
Stanford to join this company Link
242.4s
Exchange during the start of the
244.1s
internet. They solved these problems
246.1s
that other people thought were a little
248.0s
silly, a little crazy, and it became a
250.2s
big problem. Link Exchange was the first company I joined. That company was sold to Microsoft for 265 million. At that time it was the third largest At that time it was the third largest acquisition by Microsoft. Today maybe 265 million if you're joining a new lab, they would raise that, you know, their first round of financing.
252.0s
company I joined. That company was sold
254.2s
to Microsoft for 265 million.
257.2s
At that time it was the third largest
259.0s
At that time it was the third largest acquisition
261.0s
by Microsoft. Today maybe 265 million if
264.0s
you're joining a new lab, they would
265.5s
raise that, you know, their first round
267.3s
of financing.
268.7s
of financing. Um but at that time it was it was a they solved this network problem. They're trying to link all these websites together. And nobody ever thought that that was a big problem to solve. It turned out to be a big problem. Next company I joined was Tellme Networks. They were automating 800 numbers.
270.0s
but at that time it was it was a they
272.3s
solved this network problem. They're
274.0s
trying to link all these websites
276.2s
together. And nobody ever thought that
278.2s
that was a big problem to solve. It
279.8s
turned out to be a big problem. Next
281.6s
company I joined was Tellme Networks.
283.4s
They were automating 800 numbers.
286.0s
Uh and that company was sold to Microsoft for 800 million and then it was also the third largest acquisition. And they were building service voice recognition servers by themselves because there wasn't a a concept of a cloud. And basically they needed these 800 numbers to be processed in the cloud. There's no concept of a
288.3s
Microsoft for 800 million and then it
290.2s
was also the third largest acquisition.
293.0s
And they were building service
295.4s
voice recognition servers
298.1s
by themselves because there wasn't a
300.2s
a concept of a cloud. And basically they
303.1s
needed these 800 numbers to be processed
305.6s
in the cloud. There's no concept of a
307.8s
cloud, so they built their own cloud. They figured out a solution for that. And then Zappos was selling shoes where people thought you can't sell shoes on the internet because you have to try them on. We solved that problem by them on. We solved that problem by sending the shoes to your home and then you can
309.6s
They figured out a solution for that.
311.8s
And then Zappos was selling shoes where
315.4s
people thought you can't sell shoes on
317.0s
the internet because you have to try
318.1s
them on. We solved that problem by
320.0s
them on. We solved that problem by sending
321.5s
the shoes to your home and then you can
324.2s
try it at home. And so every problem has a solution, every solution creates new problems, and that's the thing that sort of keeps the economic journey uh forward and and uh forward and and um a fun fun journey along the way. You have to enjoy you have to enjoy that
326.8s
a solution, every solution creates new
328.6s
problems, and that's the thing that sort
331.2s
of keeps the economic journey
333.8s
uh forward and and
336.5s
uh forward and and um
337.6s
a fun fun journey along the way. You
340.8s
have to enjoy you have to enjoy that
343.4s
have to enjoy you have to enjoy that process. I I to press down on one of the important decisions I think you made, right? So, you said you were went to Harvard, dropped out of Stanford PhD. At that point, like the world is kind of your oyster, right? Um and a lot of people here are in
344.8s
I I to press down on one of the
346.8s
important decisions I think you made,
348.4s
right? So, you said you were went to
350.1s
Harvard, dropped out of Stanford PhD. At
352.3s
that point, like the world is kind of
353.6s
your oyster, right?
355.0s
Um and a lot of people here are in
356.1s
obviously very similar situation. You chose to join an early-stage company, a startup. Um but you've also now worked with many generational founders who founded straight out of college or people who probably went on to have very successful careers, but maybe started their careers in big tech or at the frontier labs. Uh what is like the right
358.1s
chose to join an early-stage company, a
360.0s
startup. Um but you've also now worked
363.3s
with many generational founders who
365.2s
founded straight out of college or
367.2s
people who probably went on to have very
368.8s
successful careers, but maybe started
371.1s
their careers in big tech or at the
372.4s
frontier labs. Uh what is like the right
375.7s
thing for an ambitious ambitious student here to do? How do they know what's uh the right thing to optimize for, the right place to start their career? Yeah, I think the the real answer is you just don't know. Um and there's multiple paths to success. Uh I think you should optimize for maximum learning,
377.9s
here to do? How do they know what's uh
380.4s
the right thing to optimize for, the
381.5s
right place to start their career?
383.3s
Yeah, I think the the real answer is you
385.4s
just don't know. Um and there's multiple
387.7s
paths to success. Uh I think you should
390.0s
optimize for maximum learning,
393.0s
especially when you're young, and you should maximize for working with really good people. And that's never served me wrong uh at the beginning of my career and also the beginning of my career and also today. Like when we invest in founders, we want them to be people that we want to back
394.6s
should maximize for working with
397.1s
really good people.
399.0s
And that's never served me wrong uh at
401.5s
the beginning of my career and also
403.8s
the beginning of my career and also today.
405.0s
Like when we invest in founders, we want
408.1s
them to be people that we want to back
409.8s
for the for the long journey, because every single company that is a successful company is never an overnight success. It takes 10 years to build a really legendary company or longer. Look at how long Jensen Huang has been at building GPUs. The first 20 years does not look like the last 10 years. And we
412.3s
every single company that is a
413.7s
successful company is never an overnight
416.1s
success. It takes 10 years to build a
418.6s
really legendary company or longer. Look
421.8s
at how long Jensen Huang has been at
424.5s
building GPUs. The first 20 years does
428.0s
not look like the last 10 years. And we
430.8s
just love him because of the last 10 years. He's a founder of the largest smart cap company in the world. It's about to approach 6 trillion. You know, for the first 20 years, it was zero to a hundred billion dollars. It sounds like such a small number. It was a huge
432.8s
years. He's a founder of the largest
434.7s
smart cap company in the world. It's
436.2s
about to approach 6 trillion.
439.0s
You know, for the first 20 years, it was
441.1s
zero to
442.9s
a hundred billion dollars. It sounds
444.8s
like such a small number. It was a huge
446.6s
like such a small number. It was a huge number. But he kept going. He kept going for 30 But he kept going. He kept going for 30 years. Find something that you're passionate about for 30 years. about for 30 years. Um and just as an example of like there's multiple paths to success,
447.7s
But he kept going. He kept going for 30
450.2s
But he kept going. He kept going for 30 years.
451.2s
Find something that you're passionate
452.3s
about for 30 years.
453.8s
about for 30 years. Um
454.8s
and just as an example of like there's
457.2s
multiple paths to success,
459.7s
I literally told my parents I did not want to go to college and uh I told them they had a hard life coming to the to United States. They saved up a bunch of money to send me and my brother to college. And I told them please just give me the money that you're going to
462.1s
want to go to college and uh I told them
465.4s
they had a hard life coming to the to
467.2s
United States. They saved up a bunch of
468.8s
money to send me and my brother to
470.6s
college. And I told them please just
474.0s
give me the money that you're going to
475.2s
give to this like university um Harvard and let me start a company because I had been uh in high school building computers with my dad. One of the things that we did together was just build clone PCs. Back then, you know, you get the IBM PC that cost four or five thousand dollars. If you bought all
477.8s
um Harvard and let me start a company
480.5s
because I had been
483.0s
uh in high school
484.8s
building computers with my dad. One of
486.4s
the things that we did together was just
487.9s
build clone PCs. Back then, you know,
489.8s
you get the IBM PC that cost four or
491.6s
five thousand dollars. If you bought all
493.6s
the parts, you can build you could put it together and you could with the parts probably cost a thousand dollars. You could sell it for two thousand dollars. You can make a thousand dollars by putting these things together. You know what that company is today? Like that's Dell computers. So if my parents had just allowed me to
495.3s
it together and you could
497.6s
with the parts probably cost a thousand
499.6s
dollars. You could sell it for two
500.6s
thousand dollars.
501.9s
You can make a thousand dollars by
503.2s
putting these things together.
505.0s
You know what that company is today?
507.6s
Like that's Dell computers.
509.7s
So if my parents had just allowed me to
511.9s
start that company start that company laughter you know, I I might be as you know rich and powerful as Michael Dell, but no, I work at Sequoia Capital. they were like you have to go to college. I'm like, but where are we going to pay for four colleges? So I literally only applied to four colleges.
513.5s
start that company laughter
515.3s
you know, I
516.4s
I might be as you know rich and powerful
518.5s
as Michael Dell, but no, I work at
520.0s
Sequoia Capital.
524.8s
they were like you have to go to
525.9s
college. I'm like, but where are we
527.2s
going to pay for four colleges? So I
529.4s
literally only applied to four colleges.
531.2s
Now back then, you didn't have the common app and you can't apply to multiple colleges. I went to high school that limited you to apply to seven colleges. And I only applied to the four colleges that my parents would pay for expecting not to get in and they they would have to give me the money. That
532.7s
common app and you can't apply to
534.3s
multiple colleges. I went to high school
536.5s
that limited you to apply to seven
538.2s
colleges. And I only applied to the four
541.1s
colleges that my parents would pay for
544.7s
expecting not to get in and they they
546.8s
would have to give me the money. That
548.4s
plan didn't work. I had to go to college and then I had to come to grad school because I got an NSF scholarship. They're like, you got a scholarship to They're like, you got a scholarship to go get a PhD, you have to go. Uh and part of the reason I came was
550.2s
and then I had to come to grad school
551.7s
because I got an NSF scholarship.
553.4s
They're like, you got a scholarship to
555.0s
They're like, you got a scholarship to go
556.0s
get a PhD, you have to go.
558.4s
Uh and part of the reason I came was
561.1s
because I thought this side of the because I thought this side of the planet uh was where all the technology was happening and because I did, you know, sort of invest in Microsoft and Intel because I was putting these computers together, I knew those were like interesting companies and I wanted to
563.8s
because I thought this side of the planet
565.2s
uh was where all the technology was
567.8s
happening and because I did, you know,
570.7s
sort of invest in Microsoft and Intel
573.0s
because I was putting these computers
574.4s
together, I knew those were like
575.8s
interesting companies and I wanted to
577.7s
come out here to learn what was going come out here to learn what was going on. And I just fell in love with this place. And I just fell in love with this place. Um And so, the path to success is just very, very like interesting. Life is path-dependent, but I could have ended
579.5s
come out here to learn what was going on.
580.4s
And I just fell in love with this place.
582.3s
And I just fell in love with this place. Um
583.9s
And so, the path to success is just
587.6s
very, very like interesting. Life is
590.3s
path-dependent, but I could have ended
592.2s
up coming to Stanford, finishing my PhD. My original plan was to join a hedge My original plan was to join a hedge fund. fund. Um And uh I could have done that. Uh instead, if you ask my parents which one of us I have a younger brother um it's about 3 years apart. We're very
595.3s
My original plan was to join a hedge
598.4s
My original plan was to join a hedge fund.
599.4s
fund. Um
600.8s
And uh I could have done that. Uh
603.4s
instead, if you ask my parents which one
606.6s
of us I have a younger brother um it's
609.1s
about 3 years apart. We're very
610.3s
competitive. Which one of us would work in the entrepreneurial world, and which one would work for a hedge fund? They would absolutely think that I would work in a hedge fund, and my brother would work as an entrepreneur. And it completely flipped. My brother works at Citadel Secur- Citadel, and he runs global fixed income
612.6s
in the entrepreneurial
614.4s
world, and which one would work for a
615.7s
hedge fund? They would absolutely think
617.2s
that I would work in a hedge fund, and
618.6s
my brother would work as an
620.5s
entrepreneur. And it completely flipped.
623.0s
My brother works at Citadel Secur-
624.4s
Citadel, and he runs global fixed income
626.5s
for for Khan. Uh and I work with small companies. And hope- We like to back founders from idea to IPO and beyond. Uh we work with the smallest companies possible from from the idea with two or three people. My parents think that was insane. She He My parents literally thought I was stupid. Um And to the
630.5s
companies. And hope- We like to back
632.9s
founders from idea to IPO and beyond.
635.8s
Uh we work with the smallest companies
637.7s
possible from from the idea with two or
640.2s
three people. My parents think that was
642.5s
insane. She He My parents literally
645.1s
thought I was stupid. Um And to the
648.0s
point where I when I joined my friends at Link Exchange uh when it was sold to Microsoft, they said, Well, that's great. So, you can now go finish your PhD. finished their PhDs and joined Google and had great lives, and and you know, you can do things in multiple ways of
650.2s
at Link Exchange uh
653.0s
when it was sold to Microsoft, they
655.5s
said, Well, that's great. So, you can
657.0s
now go finish your PhD.
663.8s
finished their PhDs and joined Google
666.0s
and had great lives, and and you know,
668.4s
you can do things in multiple ways of
670.5s
you can do things in multiple ways of success. Don't feel like you're tied to whatever choices you make uh while you're young. Yeah, definitely. And kind of zoning in on what you were saying at the start the most important thing to maximize for in your career, especially early on, is your career, especially early on, is learning.
671.6s
Don't feel like you're tied to whatever
673.6s
choices you make uh while you're young.
677.0s
Yeah, definitely. And kind of zoning in
679.4s
on what you were saying at the start the
681.7s
most important thing to maximize for in
683.5s
your career, especially early on, is
685.1s
your career, especially early on, is learning.
686.2s
I think it goes without saying that we're in a very unique period of time with AI accelerating and rapidly sat- saturating skills like math and coding. So, we're curious in your view, how do you view the importance of mathematics and computer science and being technical in today's era?
687.3s
we're in a very unique period of time
689.9s
with AI accelerating and rapidly sat-
692.6s
saturating skills like math and coding.
695.9s
So, we're curious in your view, how do
698.6s
you view the importance of mathematics
701.0s
and computer science and being technical
703.3s
in today's era?
705.2s
I I I have this concept of and and um we we think about things as holding two opposite ideas in tension at Sequoia. Uh and this happens with many concepts. One of which is fire and ice, which I wrote about recently. One of which is like operating at the highest level of uh strategic
709.9s
we think about things as holding two
711.8s
opposite ideas in tension at Sequoia.
715.5s
Uh and this happens with
718.2s
many concepts. One of which is
720.5s
fire and ice, which I wrote about
721.8s
recently. One of which is like operating
724.4s
at the highest level of uh strategic
727.0s
level and also at the lowest level of detail, which is what DoorDash does and detail, which is what DoorDash does and Tony Xu embodies that. It has to do with the fact that things are getting commoditized at the same time they could still be valuable. And so, yes, math and coding is being
728.9s
detail, which is what DoorDash does and
730.6s
detail, which is what DoorDash does and Tony
731.6s
Xu embodies that. It has to do with the
734.3s
fact that things are getting
735.7s
commoditized at the same time they could
737.3s
still be valuable.
739.0s
And so, yes, math and coding is being
740.9s
And so, yes, math and coding is being commoditized. But the understanding of how it works has not has never been more valuable. And you can see that by the amount of And you can see that by the amount of money that people are willing to pay for people who really really understand how
742.5s
But the understanding of how it works
745.2s
has not has never been more valuable.
748.2s
And you can see that by the amount of
750.8s
And you can see that by the amount of money
752.0s
that people are willing to pay for
753.9s
people who really really understand how
755.8s
the model works, who can really really make magic out of these models. And the people who understand that really well are no longer coding. In the past, they were coding. But today, do you really really understand the models to make the best use out of it? And if you do, you will do extremely well.
757.8s
make magic out of these models.
761.0s
And the people who understand that
762.3s
really well are no longer coding. In the
765.6s
past, they were coding. But today, do
768.1s
you really really understand the models
769.8s
to make the best use out of it? And if
772.3s
you do, you will do extremely well.
775.4s
And the And And the thing that I would encourage people who are very very technical is to also understand humans. Humanities is actually very important. You know why? If you want to be a successful founder, we build for 8 billion people on the planet. And maybe I'll get to 10, maybe I'll get to six.
777.1s
encourage people who are very very
778.9s
technical is to also understand humans.
782.8s
Humanities is actually very important.
785.0s
You know why? If you want to be a
786.8s
successful founder, we build for 8
788.7s
billion people on the planet. And maybe
790.8s
I'll get to 10, maybe I'll get to six.
792.8s
Whatever the number is, there's a lot of people on this planet. And the people are the ones that have And the people are the ones that have money. money. Um it all flows through people. Whether it's an institution that invests your money, it's still people. And so, don't forget that you're building for human beings. And you're
794.8s
there's a lot of people on this planet.
797.2s
And the people are the ones that have
798.9s
And the people are the ones that have money.
800.3s
money. Um
801.4s
it all flows through people.
803.8s
Whether it's an institution that invests
805.8s
your money, it's still people.
808.3s
And so, don't forget that you're
809.8s
building for human beings. And you're
812.5s
building for the human race. Maybe going deeper on this point of founding and you work with so many exceptional founders. You said that great founders are obviously here. They're obsessed with this problem and they quite question basically every assumption that people take for granted. For a student who knows they want to like found down the line,
815.1s
Maybe going deeper on this point of
816.7s
founding and you work with so many
818.0s
exceptional founders. You said that
820.1s
great founders are obviously here.
821.5s
They're obsessed with this problem
823.5s
and they quite question basically every
825.3s
assumption that people take for granted.
827.8s
For a student who knows they want to
829.1s
like found down the line,
831.1s
but they haven't found that meaningful problem yet. What would you tell them to do? What advice would you give? I think it's some of this is a journey. You may not be a founder straight out of college. You don't need to be. There are plenty of people who have founded companies in
833.0s
problem yet. What would you tell them to
835.4s
do? What advice would you give?
837.6s
I think it's
838.8s
some of this is a journey. You may not
840.4s
be a founder straight out of college.
842.6s
You don't need to be. There are plenty
843.8s
of people who have founded companies in
845.3s
their 20s, their 30s, their 40s, their their 20s, their 30s, their 40s, their 50s. 50s. Um I think the one thing I would continue to ask to think about is like what are interesting problems for you? And why do you want to solve those problems? And how valuable are those problems? One of those sort of concepts
847.1s
their 20s, their 30s, their 40s, their 50s.
848.3s
50s. Um
849.4s
I think the one thing I would continue
851.2s
to ask to think about is like what are
853.4s
interesting problems for you?
855.4s
And why do you want to solve those
856.7s
problems? And how valuable are those
858.7s
problems? One of those sort of concepts
860.2s
problems? One of those sort of concepts that we sort of think about is how are you going to create value? How then do you deliver value? How then do you like uh capture value? Those are the three essential stages of company building. And you can learn that by joining a friend and building a company together.
861.6s
we sort of think about is how are you
862.8s
going to create value? How then do you
864.9s
deliver value? How then do you like
868.1s
uh capture value? Those are the three
871.4s
essential stages of company building.
875.1s
And you can learn that by joining a
877.2s
friend and building a company together.
879.5s
In the first three companies I worked at, I basically found people that I really enjoyed working with and learning from. And uh you're here. You're in this privileged position uh to find people like that here and beyond. And some maybe you don't start a company from here today, but you
881.4s
at, I basically
883.2s
found people that I really enjoyed
885.2s
working with and learning from.
887.0s
And uh you're here. You're in this
890.2s
privileged position
892.0s
uh to find people like that here
894.9s
and beyond. And some maybe you don't
897.8s
start a company from here today, but you
900.2s
will find a set of people here that you can connect with, you can build with on the side, you can build with at another company, and eventually maybe start a company together. I I think the sort of the essence of something that started here uh is DoorDash. There are four co-founders that came together. They
902.5s
can connect with, you can build with on
904.4s
the side, you can build with at another
906.3s
company, and eventually maybe start a
908.7s
company together.
910.5s
I I think the sort of the essence of
912.4s
something that started here
914.4s
uh is DoorDash. There are four
917.9s
co-founders that came together. They
919.8s
co-founders that came together. They they two from the business school, from the engineering school, two graduate students, two undergrads. They came together. They just iterated. If you actually heard them like go through, it wasn't the actual DoorDash idea that they first worked on. They worked on a few things together. And then eventually they came to the
920.9s
two from the business school, from the
922.7s
engineering school, two graduate
924.8s
students, two undergrads. They came
926.7s
together. They just iterated. If you
929.3s
actually heard them like go through, it
931.6s
wasn't the actual
934.1s
DoorDash idea that they first worked on.
936.2s
They worked on a few things together.
938.2s
And then eventually they came to the
939.5s
problem that they were all passionate about and they wanted to go solve. And I think that's really, really powerful. And you know, I could, you know, say that Stanford is above to build while you're you're a student because at Harvard a similar a student because at Harvard a similar company was supposed to start
940.7s
about and they wanted to go solve.
943.1s
And I think that's really, really
944.7s
powerful. And
946.5s
you know, I could, you know, say that
948.2s
Stanford is above
954.4s
to build while you're you're
956.6s
a student because at Harvard a similar
959.0s
a student because at Harvard a similar company
960.3s
was supposed to start
962.6s
and the administration or Harvard Student Agency shut it down because you have to go through Harvard Student Agency. You have this privilege here of being able to build because the administration here gives you a lot of latitude and you should appreciate that. Talking more on on building, there's probably been the most amount of
964.4s
Student Agency shut it down because
966.8s
you have to go through Harvard Student
968.3s
Agency. You have this privilege here of
970.1s
being able to build because the
971.3s
administration here gives you a lot of
973.3s
latitude and you should appreciate that.
976.0s
Talking more on on building,
978.9s
there's probably been the most amount of
982.1s
AI demos that we've ever seen coming out every single day. Um and for both people looking to start their companies here and also people coming out of college who want to join a coming out of college who want to join a startup, startup, um what kind of signals do you look for in
984.7s
every single day.
986.0s
Um and for both people looking to start
988.4s
their companies here and also people
990.7s
coming out of college who want to join a
992.2s
coming out of college who want to join a startup,
993.6s
startup, um
994.6s
what kind of signals do you look for in
997.4s
startups or ideas or demos that give you a hunch into distinguishing whether this is just going to be like another feature which gets, you know, killed by the next soda model release or something that can actually endure into a into a prolonged and successful business? I think it actually starts with the
1000.6s
a hunch into
1002.5s
distinguishing whether this is just
1004.0s
going to be like another feature which
1005.5s
gets, you know, killed by the next soda
1007.8s
model release or something that can
1010.0s
actually endure into a into a prolonged
1012.7s
and successful business?
1014.8s
I think it actually starts with the
1016.0s
founders and their ability to tell you why this is not this is not going to be a feature. I think the If you sort of bring it all back down to two people and an idea, there's nothing defensible about that. If a large company wants to kill you, they will kill you. But you know what? They're distracted.
1018.2s
why this is not this is not going to be
1020.0s
a feature. I think the
1021.9s
If you sort of
1022.9s
bring it all back down to two people and
1024.9s
an idea, there's nothing defensible
1026.8s
about that.
1028.0s
If a large company wants to kill you,
1029.6s
they will kill you.
1031.3s
But you know what? They're distracted.
1033.2s
They got bigger problems. They're trying to spend as much of their free cash flow to build data centers right now. They're not actually building the feature that you're building. So, don't worry about the fact that it has no defensibility. But, then okay, you built this feature, you get some people interested. Uh
1034.6s
to spend as much of their free cash flow
1036.6s
to build data centers right now. They're
1038.6s
not actually building the feature that
1040.2s
you're building. So, don't worry about
1042.9s
the fact that it has no defensibility.
1045.2s
But, then okay, you built this feature,
1047.2s
you get some people interested. Uh
1049.4s
what's the next feature? How do you make that into a product suite? I think one of the things that we find with technical people, they're very very good at telling the technical story, but they can't tell you the vision. they can't tell you the vision. And that's a bit of a problem because you
1050.8s
that into a product suite?
1052.8s
I think one of the things that we find
1054.1s
with technical people, they're very very
1055.5s
good at telling the technical story, but
1057.6s
they can't tell you the vision.
1059.7s
they can't tell you the vision. And
1060.9s
that's a bit of a problem because you
1063.2s
need to be able to inspire. If you're going to be a founder, you need to inspire other people to join you. You need to inspire them to join you for the long inspire them to join you for the long run. And you have people like Brian Chesky that can sell ice to
1066.1s
If you're going to be a founder, you
1066.9s
need to inspire
1068.8s
other people to join you. You need to
1071.2s
inspire them to join you for the long
1073.4s
inspire them to join you for the long run.
1074.4s
And you have people like
1076.2s
Brian Chesky that can sell ice to
1079.0s
Eskimos, but he just has a vision of what travel should look like that is different from the world. And I also different from the world. And I also just you know, sort of think about like how fragile that Airbnb sort of idea is. You can stay with strangers, it's a listing service, they didn't complete the
1081.2s
what travel should look like that is
1083.1s
different from the world. And I also
1085.2s
different from the world. And I also just
1086.2s
you know, sort of think about like how
1088.1s
fragile that Airbnb sort of idea is. You
1091.5s
can stay with strangers, it's a listing
1093.9s
service, they didn't complete the
1095.2s
transactions then. You need to email someone, check them out, and then 48 hours later they respond to you whether they'll rent you a room in their whether they'll rent you a room in their house. You we forget how fragile these things start out. It's not the empire that it it looks clears throat like today.
1097.5s
email someone, check them out, and then
1101.0s
48 hours later they respond to you
1103.1s
whether they'll rent you a room in their
1104.9s
whether they'll rent you a room in their house.
1106.4s
You we forget how fragile these things
1109.1s
start out. It's not the empire that it
1111.9s
it looks clears throat like today.
1113.1s
They're all very very fragile. And there are going to be lots of people who say no to that idea including someone like my father who when he was helping uh his bank expand helping uh his bank expand internationally, he would travel uh and stay at the same hotel chain everywhere he went in the
1115.2s
And there are going to be lots of people
1117.5s
who say no to that idea including
1120.4s
someone like my father who when he was
1123.7s
helping uh his bank expand
1126.6s
helping uh his bank expand internationally,
1128.5s
he would travel uh and stay at the same
1131.6s
hotel chain everywhere he went in the
1133.8s
hotel chain everywhere he went in the world. And order the same exact room. The room with the same hair dryer down to the where it's placed in which drawer. He knew exactly where the hair dryer was that he never used. travel can be hectic and he wanted no variance when he wants to sleep. He
1134.8s
And order the same exact room. The room
1139.8s
with the same hair dryer down to the
1142.2s
where it's placed in which drawer. He
1145.2s
knew exactly where the hair dryer was
1148.0s
that he never used.
1153.5s
travel can be hectic and he wanted no
1155.8s
variance when he wants to sleep. He
1158.4s
wanted zero variance in that room. For Brian Chesky, that's a bug because that's not living locally. That's not belonging anywhere. That is a flaw in the system. And back to like, you have to observe these to like, you have to observe these things. Someone's feature is another person's bug and vice versa.
1161.0s
For Brian Chesky, that's a bug because
1163.7s
that's not living locally. That's not
1166.1s
belonging anywhere.
1167.7s
That is a flaw in the system. And back
1170.6s
to like, you have to observe these
1172.5s
to like, you have to observe these things.
1173.6s
Someone's feature is another person's
1175.3s
bug and vice versa.
1177.1s
And so when he started, it's like, yeah, it's a stranger. There are locals that can show you around. There are locals that can tell you all the great things that are happening in this city. The feature is not feature is not uh the conference room downstairs. The feature is the kitchen, the living room where you
1179.1s
it's a stranger. There are locals that
1180.9s
can show you around. There are locals
1182.6s
that can tell you all the great things
1184.0s
that are happening in this city. The
1186.3s
feature is not
1188.4s
feature is not uh
1189.6s
the conference room
1191.3s
downstairs. The feature is
1193.6s
the kitchen, the living room where you
1195.4s
can hang out as a family. This is a different feature than what my father was looking for. These are human things that you want to observe, especially in this world where everything, quote unquote, is getting commoditized with AI. The thing that doesn't go away is human connection. So, you lived through the internet era,
1197.3s
different feature than what my father
1199.2s
was looking for.
1200.8s
These are human things that you want to
1203.1s
observe, especially in this world where
1205.4s
everything, quote unquote, is getting
1207.1s
commoditized with AI. The thing that
1210.3s
doesn't go away is human connection.
1213.3s
So, you lived through the internet era,
1215.4s
the dot com boom, the mobile era, cloud, and now AI. Uh you've called AI the biggest mega trend, per se, uh of your career. And what about this transition, I guess, feels fundamentally different that makes this the biggest mega trend compared to everything else? And is it analogous to like these previous bubbles bubbles that
1217.8s
and now AI.
1219.2s
Uh you've called AI the biggest mega
1221.5s
trend, per se, uh of your career.
1224.2s
And what about this transition, I guess,
1226.6s
feels fundamentally different that makes
1228.6s
this the biggest mega trend compared to
1230.2s
everything else? And is it analogous to
1232.6s
like these previous bubbles bubbles that
1234.6s
we've had, or do you even think that this is a bubble? That's a compounded question. Sorry. Do I have to answer all of them, or do Do I have to answer all of them, or do you Uh whatever you think is best. to unpack. Every mega trend has been the biggest mega trend of the time.
1236.2s
this is a bubble?
1238.6s
That's a compounded question. Sorry.
1242.5s
Do I have to answer all of them, or do
1243.8s
Do I have to answer all of them, or do you
1244.0s
Uh whatever you think is best.
1245.5s
to unpack. Every mega trend has been the
1249.4s
biggest mega trend of the time.
1253.4s
So, I lived my I didn't live through the semiconductor start, but started with PCs, I talked about I was building, you know, sort of clone computers. That was the biggest mega trend at that time. The internet was the biggest mega trend at that time. Cloud Broadband was the biggest megatrend of that time. Cloud was the biggest
1256.2s
semiconductor start, but
1259.2s
started with PCs, I talked about I was
1260.8s
building, you know, sort of clone
1262.2s
computers. That was the biggest mega
1264.0s
trend at that time. The internet was the
1266.4s
biggest mega trend at that time. Cloud
1268.9s
Broadband was the biggest megatrend of
1270.4s
that time. Cloud was the biggest
1272.2s
megatrend of that time. Mobile was the biggest megatrend of that time. AI is the biggest megatrend. So, that is just in the world of accelerating change. And unfortunately, you guys are young and you think, Oh my gosh, this is so overwhelming. Every generation had to face that uh as well. And it's the price of
1274.2s
biggest megatrend of that time. AI is
1276.3s
the biggest megatrend. So, that is just
1278.8s
in the world of accelerating change.
1282.0s
And unfortunately, you guys are young
1284.8s
and you think, Oh my gosh, this is so
1286.7s
overwhelming. Every generation had to
1288.5s
face that
1290.1s
uh as well. And it's the price of
1292.6s
progress. Every generation has new sets of tools that they can play with. And you're expected to do more than the previous generation. So, you have that previous generation. So, you have that burden. burden. Um and this is the largest one partly and this is the largest one partly because
1295.8s
of tools that they can play with. And
1298.0s
you're expected to do more than the
1299.8s
previous generation. So, you have that
1302.3s
previous generation. So, you have that burden.
1303.4s
burden. Um
1304.4s
and this is the largest one partly
1306.0s
and this is the largest one partly because
1307.6s
we we got we got machines to be able to reason and think. reason and think. Uh Previous generations of technology revolutions were the industrial revolution. We sort of we created things that uh were were basically going to like take the muscle power away from humans and animals and put them in
1310.6s
reason and think.
1312.2s
reason and think. Uh
1313.1s
Previous generations of technology
1315.3s
revolutions were the industrial
1317.6s
revolution. We sort of we created things
1320.6s
that uh were were basically going to
1324.1s
like take the muscle power away from
1326.5s
humans and animals and put them in
1328.7s
humans and animals and put them in machines. And then the other in sort of uh megatrends that we talked about were communications megatrends. As our partner Konstantin talks about As our partner Konstantin talks about this, you know, most of this megatrends in the past in the past few megatrends, whether there's the internet or mobile, were
1329.9s
And then the other in sort of
1332.8s
uh megatrends that we talked about were
1334.5s
communications megatrends.
1336.7s
As our partner Konstantin talks about
1338.7s
As our partner Konstantin talks about this,
1339.6s
you know, most of this megatrends in the
1342.1s
past in the past few megatrends, whether
1344.3s
there's the internet or mobile, were
1346.4s
communications megatrends. So, we've connected everybody in the world. Now, this is a compute megatrend. This is a computational megatrend. And the reason why it's so big is this computational megatrend is connected to 8 billion people across the world. And has the capacity to be connected to 8 billion people around the world. And
1349.0s
connected everybody in the world.
1351.4s
Now, this is a compute megatrend. This
1353.2s
is a computational megatrend. And the
1355.4s
reason why it's so big is this
1356.8s
computational megatrend is connected to
1359.0s
8 billion people across the world.
1361.5s
And has the capacity to be connected to
1363.4s
8 billion people around the world. And
1365.4s
that's why it's moving so much faster. In the past, everybody wasn't connected. Um and so, how do I think about what you need to do? One of the thing biggest things that is this misconception about technology is that just because it's It's megatrend, it's going to go very very fast. It's going faster than it was before,
1367.7s
In the past, everybody wasn't connected.
1370.6s
Um and so, how do I think about what you
1373.4s
need to do? One of the thing biggest
1375.7s
things that is this misconception about
1378.1s
technology is that just because it's
1380.0s
It's megatrend, it's going to go very
1382.0s
very fast.
1383.9s
It's going faster than it was before,
1385.9s
It's going faster than it was before, but it still takes a long time. One of the reasons I dropped out of my PhD program was the internet was happening. And some of you may have heard this, but I made this projection that Amazon was going to kill Walmart. Well, 30 years later, Walmart is 30 times larger
1387.0s
it still takes a long time. One of the
1389.5s
reasons I dropped out of my PhD program
1391.8s
was the internet was happening.
1394.0s
And some of you may have heard this, but
1396.6s
I made this projection that
1398.6s
Amazon was going to kill Walmart.
1401.9s
Well, 30 years later,
1404.6s
Walmart is 30 times larger
1407.4s
than it was in 1997.
1415.4s
last year was the first year that they had more was the first year that they had more revenue than than Walmart. Took a long time to for the internet Everybody could Everybody back then could see the internet was going to could see the internet was going to happen. Um but it took a long time for it to
1417.5s
was the first year that they had more
1419.2s
was the first year that they had more revenue
1420.4s
than than Walmart.
1422.7s
Took a long time to for the internet
1424.7s
Everybody could Everybody back then
1426.4s
could see the internet was going to
1427.7s
could see the internet was going to happen.
1429.0s
Um but it took a long time for it to
1431.3s
play out. Part of that is the diff- the gap between what is possible with the technology and the diffusion to regular human beings. You know what's the slowest thing to move in technology? The humans that work with it. Because, you know, we haven't really changed all of that much. Evolution moves very very slowly.
1433.7s
gap between what is possible with the
1436.2s
technology and the diffusion
1439.4s
to regular human beings. You know what's
1442.4s
the slowest thing to move in technology?
1446.2s
The humans that work with it.
1448.4s
Because, you know,
1450.2s
we haven't really changed all of that
1452.3s
much. Evolution moves very very slowly.
1455.6s
And so, one of the things that you can help solve is the gap between what the models are capable of doing and what people use it capable of doing and what people use it for. That gap is going to always be a large gap. And if you think about it, technology has always made higher and
1457.4s
help solve
1459.7s
is the gap between what the models are
1461.8s
capable of doing and what people use it
1464.5s
capable of doing and what people use it for.
1465.6s
That gap is going to always be a large
1469.0s
gap. And if you think about it,
1471.8s
technology has always made higher and
1473.7s
higher abstractions away from the machine itself. When the mainframes come out, people had to program in machine code. My father had to program in machine code. I've never programmed machine code except for a class for in assembly. And that's not even really machine code. And none of you probably really want to work at machine code level, either.
1475.8s
machine itself.
1477.8s
When the mainframes come out, people had
1479.8s
to program in machine code.
1482.4s
My father had to program in machine
1483.8s
code. I've never programmed machine code
1486.1s
except for a class for in assembly. And
1488.6s
that's not even really machine code.
1491.2s
And none of you probably really want to
1493.2s
work at machine code level, either.
1495.7s
We just keep building more and more abstractions because that's what it takes to diffuse technology throughout takes to diffuse technology throughout society. And one of the most underestimated things I think is what you need to do as a technologist to make it diffuse through the population. Really like your point about like the
1497.2s
abstractions because that's what it
1498.8s
takes to diffuse technology throughout
1501.2s
takes to diffuse technology throughout society.
1502.5s
And one of the most underestimated
1504.6s
things I think is
1506.2s
what you need to do as a technologist to
1509.1s
make it diffuse through the population.
1512.9s
Really like your point about like the
1514.9s
inertia with which the economy moves and how slowly it takes for technology just throughout history to fully integrate into society. So, I guess like looking at current trends within AI systems, they're saturating benchmarks every single day and they're showing, you know, s- superintelligence level capabilities in certain fields. So, in your experience working with founders, working with
1519.0s
how slowly it takes for
1521.7s
technology just throughout history to
1523.0s
fully integrate into society. So, I
1525.2s
guess like looking at current trends
1527.7s
within AI systems, they're saturating
1530.6s
benchmarks every single day and they're
1532.4s
showing, you know, s-
1534.4s
superintelligence level capabilities in
1536.4s
certain fields. So, in your experience
1539.6s
working with founders, working with
1540.7s
companies, and seeing their struggles, where in your mind are the biggest gaps between what AI models um can do and what people are actually doing which models can't solve right now? I think there is a number of ways you can take this. The the models are clearly more powerful than we are using
1542.8s
where in your mind are the biggest gaps
1544.6s
between what AI models um can do and
1547.7s
what people are actually doing which
1550.5s
models can't solve right now?
1552.9s
I think there is a number of ways you
1554.3s
can take this. The the models are
1556.1s
clearly more powerful than we are using
1559.0s
it for. And so, there are many places where you can apply the model and get economic results. Um whether it's in bio, in law. My partner Pack Brady loves to say there is value in the application layer. All this whole like notion that software is dead is kind of BS because whether
1561.2s
where you can apply the model and get
1564.0s
economic results. Um
1566.5s
whether it's in bio, in law. My partner
1569.0s
Pack Brady loves to say there is value
1570.9s
in the application layer.
1572.9s
All this whole like notion that software
1574.6s
is dead is kind of BS because whether
1577.3s
you call it software, you call it application, you call it the diffusion layer, that is the gap between what the models are capable of and human beings using it. And so, there's a ton of value using it. And so, there's a ton of value there. Uh and we're investors in uh in Harvey,
1579.2s
application, you call it the diffusion
1581.0s
layer, that is the gap between what the
1583.5s
models are capable of and human beings
1585.7s
using it. And so, there's a ton of value
1588.3s
using it. And so, there's a ton of value there.
1589.2s
Uh and we're investors in uh in Harvey,
1592.4s
in Open Evidence, in Profound, in in all these places where you're taking something that the model's capable of, but the people on the other side, whether it's a lawyer or a doctor or a marketing person or a finance person like Inigo or Rosebud uh examples, they are not going to go all the way
1595.9s
these places where
1597.7s
you're taking something that the model's
1599.3s
capable of, but the people on the other
1601.6s
side, whether it's a lawyer or a doctor
1603.4s
or a marketing person or a finance
1605.3s
person like Inigo or
1607.1s
Rosebud uh examples,
1609.6s
they are not going to go all the way
1611.6s
down and understand how the model works to go program it. Uh they're are specifically taking the models down to areas where you want to own your own intelligence. How do you sort of own your own intelligence for pharmaceutical company? You're going to take open weight models and program it on your own data. And how
1613.6s
to go program it.
1615.4s
Uh they're are specifically taking the
1617.5s
models down to areas where
1620.1s
you want to own your own intelligence.
1622.3s
How do you sort of own your own
1624.0s
intelligence for pharmaceutical company?
1626.5s
You're going to take open weight models
1628.3s
and program it on your own data. And how
1631.0s
do you do that? And how do you go about doing that? These are going to be very important things that we're not uh just going to let two or three models win. That's not how economies work. We don't like that. We don't like that concentration of power. We like the fact that they keep pushing the the frontier,
1632.5s
doing that? These are going to be very
1634.5s
important things that we're not uh
1638.2s
just going to let two or three models
1641.2s
win. That's not how economies work. We
1644.7s
don't like that. We don't like that
1646.4s
concentration of power. We like the fact
1648.9s
that they keep pushing the the frontier,
1651.4s
but then we don't want them to gain too much power and we don't want to put all our data there. So then we're going to take some of the data out and we're going to we're going to we're going to train it ourselves. And so how you do how you going to build
1653.3s
much power and we don't want to put all
1654.6s
our data there. So then we're going to
1656.0s
take some of the data out and we're
1657.6s
going to we're going to
1659.0s
we're going to train it ourselves. And
1660.7s
so how you do how you going to build
1662.3s
tools for developers to to do that such tools for developers to to do that such as you know, so how do you build a a model for drug discovery like child discovery is doing or Isomorphic is doing? How do you build tools for developers like LangChain is doing? How do you build open weight
1665.2s
tools for developers to to do that such as
1665.9s
you know, so
1667.0s
how do you build a a model for drug
1670.1s
discovery like child discovery is doing
1672.2s
or Isomorphic is doing? How do you build
1674.2s
tools for developers like LangChain is
1676.2s
doing? How do you build open weight
1677.8s
models like Reflection is doing? There there's so many things that you can do because the diffusion gap is only getting bigger and bigger and bigger. And if you want if you fill that diffusion gap, there's just a lot of work to do. Again, every problem has a solution, every solution has creates other problems. We didn't have alignment
1679.4s
there's so many things that you can do
1681.9s
because the diffusion gap is only
1684.4s
getting bigger and bigger and bigger.
1686.9s
And if you want if you fill that
1688.8s
diffusion gap, there's just a lot of
1691.0s
work to do. Again, every problem has a
1693.7s
solution, every solution has creates
1695.7s
other problems. We didn't have alignment
1697.4s
problem until we told the models to go figure out some cybersecurity like risks. Well, if you figure out the cybersecurity risk, I guess it's going to figure out how to get out of the sandbox. Now you have to go control the fact that it can get out of the sandbox. It's so fun because we're always going
1699.5s
the models to go figure out some
1701.7s
cybersecurity like risks. Well, if you
1704.3s
figure out the cybersecurity risk, I
1706.0s
guess it's going to figure out how to
1707.3s
get out of the sandbox. Now you have to
1709.2s
go control the fact that it can get out
1710.9s
of the sandbox.
1712.4s
It's so fun because we're always going
1715.7s
to have more work to do. Which is why not a deemster, deemler because there's just so much work to do and we can keep improving things over and over again. and over again. clears throat I think an interesting point you noted there was like value accruing accruing in the application layer. But I wonder,
1717.9s
Which is why not a deemster, deemler
1720.4s
because there's just so much work to do
1723.5s
and we can keep improving things over
1725.8s
and over again.
1727.3s
and over again. clears throat
1727.6s
I think an interesting point you noted
1728.8s
there was like value accruing accruing
1730.8s
in the application layer. But I wonder,
1733.7s
do you think that there will be like someone that goes maybe one step deeper or maybe this is like a logical next step for these application companies where you want to sort of take over the services layer where it's like maybe a trivial example is like a Harvey sells like a service right now to
1735.2s
someone that goes maybe one step deeper
1737.0s
or maybe this is like a logical next
1739.0s
step for these application companies
1741.0s
where you want to sort of
1742.8s
take over the services layer where it's
1744.5s
like maybe a trivial example is like a
1746.4s
Harvey sells like a service right now to
1748.0s
a lawyer, but what if Harvey can become the lawyer using their turn tools? I think all of those things are I think all of those things are possible. Remember Harvey started building out on on frontier models and they're training their own models because they wanted to own their own intelligence. They work primarily with um
1750.5s
the lawyer using their turn tools?
1753.2s
I think all of those things are
1754.4s
I think all of those things are possible.
1755.9s
Remember Harvey started building out on
1758.5s
on frontier models and they're training
1760.8s
their own models because they wanted to
1762.3s
own their own intelligence.
1764.3s
They work primarily with um
1767.2s
large law firms, the top 10 100 law firms. Are they going to work just with those law firms? Are they going to work with corporates? They're starting to work corporates. These things are all possible. And because of the power of these models, I would encourage everybody to just think much more ambitiously than than before.
1771.0s
firms. Are they going to work just with
1773.7s
those law firms? Are they going to work
1775.6s
with corporates? They're starting to
1776.8s
work corporates. These things are all
1779.2s
possible. And because of the power of
1782.6s
these models, I would encourage
1784.3s
everybody to just think much more
1785.8s
ambitiously than than before.
1788.6s
Uh when I was growing up and I had to learn a whole programming language to go build a game. I see the delight of kids being able to just write code a game and how quickly they just enjoy playing that game. Well, what are they going to do next? They're going to create more games and they're
1791.3s
learn a whole programming language to go
1793.1s
build a game.
1794.4s
I see the delight of kids being able to
1796.2s
just write code a game and how quickly
1799.9s
they just enjoy playing that game. Well,
1801.8s
what are they going to do next? They're
1803.0s
going to create more games and they're
1804.2s
going to do more things with it. And I think there's going to just be a ton of things that is capable of being advanced in human potential because of advanced in human potential because of it. In a recent Twitter post, uh I think it's pinned on your Twitter last time I checked, you wrote about the
1806.2s
And I think there's going to just be a
1808.3s
ton of things that is capable
1811.4s
of being
1813.0s
advanced in human potential because of
1815.2s
advanced in human potential because of it.
1816.6s
In a recent Twitter post, uh I think
1818.4s
it's pinned on your Twitter
1820.2s
last time I checked, you wrote about the
1821.8s
importance of the duality of fire and ice in a company and you mentioned this earlier. Where fire represents the passion and the vision and the burning desire to solve a critical problem. But ice represents the composure and the clear-minded strategy and the execution. Um I was wondering why this idea of fire
1824.6s
ice in a company and you mentioned this
1826.4s
earlier. Where fire represents the
1829.1s
passion and the vision and the burning
1830.8s
desire to solve a critical problem. But
1833.7s
ice represents the composure and the
1836.2s
clear-minded strategy and the execution.
1838.8s
Um I was wondering why this idea of fire
1842.2s
and ice stood out to you in particular. And maybe we were wondering like could you give some examples in practice of where you've seen this be the difference maker in what makes a company great? I think when when you start a company there's a lot of passion. It starts with a vision. It starts with
1845.8s
And maybe
1847.0s
we were wondering like could you give
1848.2s
some examples in practice of where
1850.3s
you've seen this be the difference maker
1853.1s
in what makes a company great?
1855.6s
I think when when you start a company
1857.5s
there's a lot of passion.
1859.4s
It starts with a vision. It starts with
1861.5s
creativity. It starts with something that nobody's ever seen before. And then you quickly understand that you need to take that idea and go execute it. You need to build a product. You need to sort of figure out how to get the product into the hands of people. You need to figure out how to
1864.2s
that nobody's ever seen before.
1866.4s
And then you quickly understand that you
1868.2s
need to take that idea and go execute
1870.8s
it. You need to build a product. You
1872.2s
need to sort of figure out how to get
1873.7s
the product into the hands of people.
1876.5s
You need to figure out how to
1878.4s
get it bought and people paying for it. And all of that sort of that journey is the entrepreneurial journey. If you only build a as I said like if you only create value and it never gets into the hands of other people you've not delivered. And so many companies start out with a lot of
1882.8s
And all of that sort of that journey is
1884.8s
the entrepreneurial journey. If you only
1886.9s
build a
1888.1s
as I said like if you only create value
1890.6s
and it never gets into the hands of
1892.6s
other people you've not delivered.
1895.3s
And so
1896.8s
many companies start out with a lot of
1898.9s
fire. Uh and over time they need the just the cold hard discipline of of management and discipline. And that discipline allows you to create to have more fire cuz then you can take on more ideas. And then again you you need the discipline of it. My lived experience with that was at
1901.8s
just the cold hard discipline of
1904.6s
of management and discipline.
1907.6s
And that discipline allows you to create
1909.6s
to have more fire cuz then you can take
1911.8s
on more ideas. And then again you you
1915.1s
need the discipline of it.
1916.9s
My lived experience with that was at
1919.2s
Zappos. We started with shoes. Then we expanded to to handbags. Then we expanded to clothes. And our journey was relatively short as opposed to uh snorts Amazon. When we got acquired by Amazon we saw very similar sort of approach where they started with books. They got the discipline of books being profitable. Then they went to music.
1921.4s
expanded to to handbags. Then we
1924.0s
expanded to clothes.
1925.9s
And our journey was relatively short as
1928.5s
opposed to
1930.1s
uh snorts
1931.1s
Amazon. When we got acquired by Amazon
1933.0s
we saw very similar sort of approach
1935.7s
where they started with books.
1937.5s
They got the discipline of books being
1939.0s
profitable. Then they went to music.
1940.7s
They got the discipline of music being profitable. That allowed them to fuel the next category and the next category. Eventually they became the everything Eventually they became the everything store. These ideas are the are just like this tension between the two. Sometimes you can find it in one person and and I've like talked about
1942.5s
profitable. That allowed them to fuel
1944.7s
the next category and the next category.
1946.6s
Eventually they became the everything
1948.0s
Eventually they became the everything store.
1949.4s
These ideas are the
1952.1s
are just like this tension between the
1954.5s
two. Sometimes you can find it in one
1956.5s
person and and I've like talked about
1959.4s
this where Tony Hsieh is like that one person where he has a lot of creativity yet he works at the lowest level of detail. And then there are certain situations where uh there's one person who has a lot of fire and another person who's who has a lot of ice on the founding team which
1961.4s
person where he has a lot of creativity
1963.8s
yet he works at the lowest level of
1965.4s
detail. And then there are certain
1967.1s
situations where
1968.8s
uh there's one person who has a lot of
1971.1s
fire and another person who's who has a
1973.4s
lot of ice on the founding team which
1975.9s
was what I I was the ice and Tony Hsieh uh different Tony at Zappos was the fire. Um and fire. Um and you you should know yourself very clearly on which side you land and what you need to complement yourself with whether it's on a management team or uh or being a founding team.
1979.8s
uh different Tony at Zappos was the
1982.6s
fire. Um and
1985.1s
fire. Um and you
1986.2s
you should know yourself very clearly on
1988.7s
which side you land and what you need to
1990.9s
complement yourself with whether it's on
1993.3s
a management team or uh
1995.8s
or being a founding team.
1998.2s
I think one interesting thing that I saw quite a few questions about um of people who signed up was there's this notion right now that money and AI is moving very circularly, right? Um and it's like oh an Nvidia invest in this company they have money from Nvidia and they go buy Nvidia GPUs
2000.4s
quite a few questions about um of people
2002.5s
who signed up was
2004.0s
there's this notion right now that money
2005.8s
and AI is moving very circularly, right?
2008.8s
Um and it's like
2010.8s
oh an Nvidia invest in this company
2014.0s
they have money
2015.7s
from Nvidia and they go buy Nvidia GPUs
2017.5s
with that money, right? Obviously, you move billions of dollars in AI. So, snorts we're kind of wondering what is your take? small amounts of money. But I feel like you you'd know a lot about this and what what's your kind of take on this? Is this actually like something to be concerned about or is
2019.8s
move billions of dollars in AI. So,
2022.0s
snorts we're kind of wondering what is
2023.5s
your take?
2024.2s
small amounts of money.
2027.1s
But I feel like you you'd know a lot
2029.0s
about this and what what's your kind of
2030.2s
take on this? Is this actually like
2031.4s
something to be concerned about or is
2033.0s
this like this is normal or are people sort of misconstruing or misunderstanding this? I think it's a good kind of concern to address. good kind of concern to address. Um I think if if money is only circular it would be a problem. But don't forget money has always been circular.
2036.4s
sort of misconstruing or
2037.5s
misunderstanding this? I think it's a
2039.2s
good kind of concern to address.
2042.0s
good kind of concern to address. Um
2043.3s
I think if
2045.4s
if money is only circular
2048.6s
it would be a problem. But don't forget
2050.8s
money has always been circular.
2052.8s
And so maybe you know one of the things that I would encourage people to sort of study is just take an economics class. They can understand how the banking system works. Uh you have 100 you deposit in the bank. What does the bank actually do? The bank takes they they a reserve
2054.9s
that I would encourage people to sort of
2056.7s
study is just take an economics class.
2059.6s
They can understand how the banking
2061.2s
system works.
2062.9s
Uh you have 100 you deposit in the
2065.9s
bank. What does the bank actually do?
2068.6s
The bank takes they they a reserve
2070.6s
requirement of 20. They lend the 80 to other people. Those people take that money, invest it somewhere it goes into someone else's bank account. That goes in the bank, whatever they have a 20 reserve. So, there is velocity in money. And so, I don't think of this as like something that you should be 100 concerned about
2074.4s
to other people. Those people take that
2077.5s
money, invest it somewhere it goes into
2079.2s
someone else's bank account.
2081.6s
That goes in the bank, whatever they
2083.6s
have a 20 reserve. So, there is
2086.2s
velocity in money.
2088.1s
And so, I don't think of this as like
2091.8s
something that you should be 100
2094.0s
concerned about
2095.6s
unless there's no reserve requirement. Meaning all the money that comes in goes exactly out. And then, I would be a little bit concerned. But, that's not the case. And of course, Nvidia wants to invest because they want you to use Nvidia GPUs as opposed to some other GPUs. And other companies are doing the same
2098.6s
Meaning all the money that comes in goes
2100.6s
exactly out.
2102.4s
And then, I would be a little bit
2104.1s
concerned. But,
2105.9s
that's not the case. And of course,
2108.2s
Nvidia wants to invest because they want
2110.7s
you to use Nvidia GPUs as opposed to
2113.1s
some other GPUs.
2114.8s
And other companies are doing the same
2116.2s
And other companies are doing the same thing. But, that But, Jensen's reason for doing that is because he wants to empower the that is because he wants to empower the ecosystem. And that's a good thing. I think it's good to point that out. I think it's good to be cautious. I think
2117.2s
But, that But, Jensen's reason for doing
2120.4s
that is because he wants to empower the
2122.4s
that is because he wants to empower the ecosystem.
2124.0s
And that's a good thing.
2126.4s
I think it's good to point that out. I
2128.1s
think it's good to be cautious. I think
2130.2s
the bigger concern actually is just the amount of debt companies are taking on. We have companies that used to be nothing but cash flow positive, and now they're cash flow negative, and they're taking on debt to build more data taking on debt to build more data centers. And as all of you know, if you took economics, leverage
2132.7s
amount of debt
2134.8s
companies are taking on.
2137.1s
We have companies that used to be
2139.6s
nothing but cash flow positive, and now
2142.0s
they're cash flow negative, and they're
2143.8s
taking on debt to build more data
2145.3s
taking on debt to build more data centers.
2146.5s
And as all of you know, if you took
2149.2s
economics, leverage
2151.7s
uh is very powerful. It can make you or break you. And uh And so, I think the amount of leverage in this system that is actually the issue, not necessarily this phenomenon. All right. Alfred, thank you so much. As promised, we have 20 minutes for audience QA. Nowadays, we see founders raising uh
2155.1s
or break you. And uh
2157.4s
And so, I think the amount of leverage
2159.7s
in this system that is actually the
2162.0s
issue, not necessarily this phenomenon.
2165.3s
All right. Alfred, thank you so much. As
2167.7s
promised, we have 20 minutes for
2170.0s
audience QA.
2171.7s
Nowadays, we see founders raising uh
2173.8s
tens of millions of dollars before even scratching the surface of product market fit. Um therefore, what unconventional or underrated characteristics do you invest in in founders where you're taking a large risk? Well, I think the one one thing I would say that's uncharacteristic is just generally the just understanding the founders journey and what got them here.
2176.2s
scratching the surface of product market
2178.4s
fit. Um therefore, what unconventional
2181.7s
or underrated characteristics do you
2184.4s
invest in in founders where you're
2185.6s
taking a large risk?
2187.7s
Well, I think the one
2189.6s
one thing I would say that's
2190.9s
uncharacteristic is just generally the
2193.0s
just understanding the founders journey
2195.6s
and what got them here.
2198.2s
Uh we we use the term we like to back people who are outliers and just they don't fit any pattern. Uh in some sense they're an Uh in some sense they're an anti-pattern. Uh and they got to where they are because they took risk throughout their because they took risk throughout their life. If you're a
2201.2s
people who are outliers and just they
2203.6s
don't fit any pattern.
2205.8s
Uh in some sense they're an
2207.7s
Uh in some sense they're an anti-pattern.
2209.2s
Uh and they got to where they are
2210.8s
because they took risk throughout their
2212.1s
because they took risk throughout their life.
2213.8s
If you're a
2215.6s
If you're going to be a founder, you need to be able to take risk and you need to be able to take it with a lot of like conviction. And that's what we look for with founders. Like how much conviction do you have on your idea? And conviction to us is not passion. It's not just because
2217.0s
need to be able to take risk and you
2218.5s
need to be able to take it
2220.6s
with a lot of like conviction.
2222.9s
And that's what we look for with
2224.3s
founders. Like how much conviction do
2225.9s
you have on your idea? And conviction to
2227.6s
us is not passion. It's not just because
2230.2s
you can speak about it, but you actually have investigated the problem and really understand it. That's why you would come up with a novel and compelling idea. And I think one example I would always use is Tony Xu So Tony Xu wanted to start DoorDash because he knew the pain of merchants.
2232.5s
have investigated the problem and really
2235.3s
understand it. That's why you would come
2237.3s
up with a novel and compelling idea.
2239.8s
And I think one example I would always
2242.0s
use is Tony Xu
2244.4s
So Tony Xu wanted to start DoorDash
2247.3s
because he knew the pain of merchants.
2251.2s
Uh and because he he lived it. He lived it with his his uh mom standing next to her washing dishes in restaurants. And the issue that he noticed was that the restaurant capacity, the kitchen capacity was always greater than the capacity was always greater than the front of uh restaurant capacity. That insight didn't come from someone
2253.6s
it with his his uh mom standing next to
2257.0s
her washing dishes in restaurants. And
2260.1s
the issue that he noticed was that
2263.6s
the restaurant capacity, the kitchen
2265.2s
capacity was always greater than the
2267.0s
capacity was always greater than the front
2268.2s
of uh restaurant capacity.
2271.1s
That insight didn't come from someone
2272.7s
who just theoretically talked to them and wanted to clears throat sort of create a food delivery service. It was real insight from real insight from uh understanding the customer. Um I work in investment banking this summer and I know Sequoia has a lot of investors from very different backgrounds. Some came from like finance
2275.1s
and wanted to clears throat sort of
2275.8s
create a food delivery service. It was
2277.9s
real insight from
2279.6s
real insight from uh
2280.4s
understanding the customer.
2282.2s
Um I work in investment banking this
2283.8s
summer and I know Sequoia has a lot of
2285.8s
investors from very different
2287.3s
backgrounds. Some came from like finance
2288.9s
or private equity or some had like founding or operating roles. So in your opinion, how do these different backgrounds shape how someone evaluates a company or works with the founder? And with each of these different paths, what are some like strengths or blind spots that investors may have? One of my partners would say there's
2290.7s
founding or operating roles. So in your
2292.6s
opinion, how do these different
2294.2s
backgrounds shape how someone evaluates
2296.4s
a company or works with the founder? And
2298.5s
with each of these different paths, what
2300.3s
are some like strengths or blind spots
2302.6s
that investors may have?
2304.7s
One of my partners would say there's
2307.0s
many paths to Nirvana. Um and so I would say the if you had to break up early stage investing and growth investing, the difference is early stage investing is about founder quality and growth stage investing is about business quality. And with the tools that you get from being in investment banking or private equity
2308.8s
Um and so
2311.8s
I would say the if you had to break up
2314.8s
early stage investing and growth
2316.4s
investing, the difference is
2318.3s
early stage investing is about founder
2320.0s
quality and growth stage investing is
2322.5s
about business quality. And
2325.0s
with the tools that you get from being
2327.4s
in investment banking or private equity
2330.1s
you learn a lot about business quality. And what's missing may be founder quality, founder empathy. And if you work at a company and you work with founder, you know that because you that's a lived experience. And so I think advice is free for a reason. It only kind of works for the person that's
2332.8s
And what's missing may be founder
2335.5s
quality, founder empathy. And if you
2338.1s
work at a company and you work with
2339.5s
founder, you know that because you
2341.4s
that's a lived experience. And so
2344.7s
I think
2345.9s
advice is free for a reason. It only
2348.2s
kind of works for the person that's
2349.5s
giving it. So even here I kind of worked first in companies and so I know a set of experiences that is helpful to founders because I lived the problems that they lived and I can provide that founder empathy. Is who's the best mentor that you had and what is the best advice that they gave to you?
2351.8s
I kind of worked first in companies and
2355.4s
so I know
2357.0s
a set of experiences that is helpful to
2360.3s
founders because I lived the problems
2362.0s
that they lived and I can provide that
2364.2s
founder empathy.
2365.5s
Is who's the best mentor that you had
2368.6s
and what is the best advice that they
2370.0s
gave to you?
2372.2s
I would encourage you to not have the one best mentor and have everybody be a mentor in your life. Um my best mentor, if you wanted to sort of nail that down, was my father who told me there's something that you can learn from everyone. That was the best advice
2374.4s
one best mentor and have everybody be a
2376.6s
mentor in your life. Um my best mentor,
2379.7s
if you wanted to sort of
2381.4s
nail that down, was my father who told
2383.3s
me there's something that you can learn
2385.1s
from everyone. That was the best advice
2387.1s
he ever gave me. And uh that's kind of how I view the world. Like if I have a math problem, I'll go find the right math person for that. Um if I have a business problem, I'll go find the right business person, the right business. And I think it's really, really
2390.8s
how I view the world. Like if I have a
2393.8s
math problem, I'll go find the right
2395.8s
math person for that. Um if I have a
2398.5s
business problem, I'll go find the right
2399.9s
business person, the right business.
2402.0s
And I think it's really, really
2403.9s
important to just view the world as your oyster and being able to sort of learn from anyone. And you mentioned that great companies are generation centric and that they follow the trends of their time and they're aligned with them. I wonder what you think our generations priors that we're inheriting are. What are the
2406.4s
oyster and being able to sort of learn
2408.4s
from anyone.
2409.0s
And you mentioned that great companies
2411.5s
are generation centric and that they
2413.5s
follow the trends of their time and
2415.0s
they're aligned with them. I wonder what
2416.9s
you think our generations priors that
2419.2s
we're inheriting are. What are the
2420.8s
priors that are flawed that we have to recalibrate that may hinder great recalibrate that may hinder great founders? That I have not thought about very carefully, but I think the set of priors that you're likely to inherit because you are taught by people from the previous generation is they built companies that required a
2422.3s
recalibrate that may hinder great
2424.2s
recalibrate that may hinder great founders?
2425.7s
That I have not thought about very
2427.8s
carefully, but I think the set of priors
2430.2s
that you're likely to inherit because
2432.8s
you are taught by people from the
2434.8s
previous generation
2436.7s
is they built companies that required a
2440.7s
lot of people to write software. And so, the way they organize the companies are probably organized in a companies are probably organized in a way that required a lot of people to write software to control it, etc. And it might be very, very different in the future. the future. Um I also just think that
2443.8s
And so, the way they organize the
2446.1s
companies are probably organized in a
2448.1s
companies are probably organized in a way
2449.6s
that required a lot of people to write
2451.7s
software to control it, etc.
2454.5s
And it might be very, very different in
2456.4s
the future.
2458.7s
the future. Um
2460.4s
I also just think that
2462.3s
every generation always in the previous generation you required more
2465.4s
generation you required more
2473.3s
people to do certain jobs than the next generation. But, it's not like the companies got But, it's not like the companies got smaller. The the jobs shifted. And uh Like I think if you were building a company in automotive, clears throat there's a lot of command and control. You listen to Jensen and there's very
2475.0s
But, it's not like the companies got
2477.1s
But, it's not like the companies got smaller.
2478.7s
The the jobs shifted.
2481.4s
And uh
2483.3s
Like I think if you were
2485.2s
building a company
2487.4s
in automotive, clears throat
2488.3s
there's a lot of command and control.
2490.2s
You listen to Jensen and there's very
2492.0s
little command and control. Management systems need to change for the new technologies. And I don't know what the answer is for AI. Because certainly it's going to be much more decentralized than it was much more decentralized than it was before. This notion that one person can build and code with an idea can
2493.8s
Management systems need to change for
2496.5s
the new technologies.
2498.5s
And I don't know what the answer is for
2501.8s
AI. Because certainly it's going to be
2504.4s
much more decentralized than it was
2506.2s
much more decentralized than it was before.
2507.3s
This notion that one person can build
2509.3s
and code with an idea can
2511.8s
build a product from beginning to end and launch it and generate revenue we've never seen before. And so, if you want to build a big company with a big surface area with lots of people who are doing that, what does that organization look like? Would be the the question. I think also some frameworks are just wrong because
2513.9s
and launch it
2515.3s
and generate revenue we've never seen
2517.4s
before. And so, if you want to build a
2519.5s
big company with a big surface area with
2521.8s
lots of people who are doing that, what
2523.7s
does that organization look like?
2526.1s
Would be the the question. I think also
2528.5s
some frameworks are just wrong because
2530.6s
they were never right. As lean startup was developed here. As lean startup was developed here. Every every year the amount of capital that seed companies raise since I've been at Sequoia has only gone up, not down. And so I don't even understand whether lean startup ever work. So, you should be you should be wary of
2533.1s
As lean startup was developed here.
2536.1s
As lean startup was developed here. Every
2537.4s
every year the amount of capital that
2540.2s
seed companies raise since I've been at
2542.8s
Sequoia has only gone up, not down.
2547.0s
And so
2548.4s
I don't even understand whether lean
2550.1s
startup ever work.
2552.7s
So, you should be you should be wary of
2555.8s
frameworks that gain popularity because it tells a good story. Um one of the sort of frameworks I have is beware of simple narratives. And comps people hate this, right? Because comps people will tell you have a simple narrative. Why? Because simple narratives are powerful. We love simple narratives. You love simple answers. You want
2559.3s
it tells a good story.
2561.5s
Um one of the sort of
2563.5s
frameworks I have is beware of simple
2566.0s
narratives. And comps people hate this,
2568.4s
right? Because comps people will tell
2570.0s
you have a simple narrative. Why?
2571.8s
Because simple narratives are powerful.
2574.2s
We love simple narratives. You love
2576.0s
simple answers. You want
2578.0s
And this it works it works to a point. But again, the reason why we're in whatever troubles that we are is because we don't understand the nuances of that simple narrative and when it breaks. But to your point, I would encourage all of you to think about your frameworks because all of us have frameworks. We
2582.1s
But again, the reason why we're in
2585.2s
whatever troubles that we are is because
2587.2s
we don't understand the nuances of that
2589.4s
simple narrative and when it breaks.
2592.0s
But to your point, I would encourage all
2593.8s
of you to think about
2595.9s
your frameworks
2598.0s
because all of us have frameworks. We
2600.4s
don't always think from first don't always think from first principles. And know the edges of where that breaks. One example for me over the 16 years I've been at Sequoia was like quite refreshing in one sense once I understood it. This whole idea every every year some somehow the check sizes keep going up. It's like how can
2601.7s
don't always think from first principles.
2602.8s
And know the edges of where that breaks.
2605.5s
One example for me
2607.9s
over the 16 years I've been at Sequoia
2609.8s
was like quite refreshing in one sense
2612.3s
once I understood it. This whole idea
2615.0s
every every year some somehow the check
2617.7s
sizes keep going up. It's like how can
2619.7s
this be? Must be returns must be getting worse because we invest more dollars at higher valuations. So we're Well, returns have not gotten worse because the outcomes have gotten bigger. So, don't cap your like thinking that the outcomes are like static. And the example I would give is when I joined Sequoia the largest market cap
2622.2s
worse because
2624.3s
we invest more dollars at higher
2626.2s
valuations. So we're
2627.7s
Well, returns have not gotten worse
2629.6s
because the outcomes have gotten bigger.
2632.6s
So, don't cap your like thinking that
2635.5s
the outcomes are like static.
2637.7s
And the example I would give is when I
2640.1s
joined Sequoia the largest market cap
2641.8s
company was 400 billion dollars. Today we have Nvidia that is few days ago was 5 trillion something today might be 6 trillion but it has 15 x in 15 years and so one framework is you need more capital in this world to compete. The other framework is you also if you're going to
2644.1s
billion dollars.
2645.6s
Today we have Nvidia that is
2649.0s
few days ago was 5 trillion something
2651.2s
today might be 6 trillion but it has 15
2654.6s
x in 15 years
2656.6s
and so
2659.4s
one framework is you need more capital
2661.4s
in this world to compete. The other
2663.1s
framework is you also if you're going to
2665.1s
found a company and you want to build a legendary company today the distance you have to travel is 15 x. Now unfortunately our timelines have not increased our lifespans have not increased. You just have to do a lot more work and accomplish a lot more in the same time frame because yes your life expectancy you can see
2667.1s
legendary company today the distance you
2669.7s
have to travel is 15 x.
2672.5s
Now unfortunately our timelines
2675.6s
have not increased our lifespans have
2677.2s
not increased. You just have to do a lot
2679.7s
more work and accomplish a lot more in
2681.8s
the same time frame because
2684.1s
yes your life expectancy you can see
2685.9s
will be slightly higher than mine but will be slightly higher than mine but not unless we figure out how to join the singularity and merge with machines I think you know your lifetimes will probably still be limited by 120 years or 150 years. I might self cancer who knows I might
2688.3s
will be slightly higher than mine but not
2689.6s
unless we figure out how to
2691.7s
join the singularity and merge with
2693.3s
machines I think you know
2695.6s
your lifetimes will probably still be
2696.9s
limited by 120 years or 150 years.
2700.8s
I might self cancer who knows I might
2709.6s
incredible inspirations for Taiwanese Americans like us and I'm wondering as a founder or an investor you're always laughter going to have to get some kind of buy-in from other people either more C-suite than other people either more C-suite than you
2712.0s
Americans like us
2714.4s
and I'm wondering as a founder or an
2717.3s
investor you're always laughter going
2719.1s
to have to get some kind of buy-in from
2721.9s
other people either more C-suite than
2724.1s
other people either more C-suite than you
2735.5s
Getting buy-in? Yes like how do you persuade other Yes like how do you persuade other people? I mean I think the the sort of simple answer is like have such a compelling vision that everybody wants to join you that makes it simple um and the other way I say it is if they
2737.0s
Yes like how do you persuade other
2739.5s
Yes like how do you persuade other people?
2741.1s
I mean I think the the sort of simple
2743.3s
answer is like have such a compelling
2746.8s
vision that everybody wants to join you
2749.4s
that makes it simple um
2752.3s
and the other way I say it is if they
2754.5s
don't want to join they're not the right people and so find your community find the right people that want to join you. And so you know you obviously want to create a vision that everybody wants to join, but if you can't do If you're You're probably not going to do that at the
2755.8s
people and so find your community find
2758.6s
the right people that want to join you.
2760.9s
And so
2762.1s
you know you obviously want to create a
2763.6s
vision that everybody wants to join, but
2765.4s
if you can't do If you're You're
2767.4s
probably not going to do that at the
2768.4s
beginning, but when you sort of know what your vision is, then find the people who want to join that that vision. If you were to start a AI-native hedge fund today, what do you think would be the most um durable edge that AI can do? durable edge that AI can do? Uh
2771.8s
know what your vision is, then find the
2773.4s
people who want to join that
2775.5s
that vision.
2776.6s
If you were to start a AI-native hedge
2779.7s
fund today,
2781.4s
what do you think would be the most um
2784.6s
durable edge that AI can do?
2788.5s
durable edge that AI can do? Uh
2789.8s
I think the one thing I would sort of think through is like what areas that regular like the sort of the There's a bunch of hedge funds that are using machine learning to price things. And so, your edge has to be that they are pricing something not as well as your AI model.
2791.9s
think through is like what areas that
2793.9s
regular like the sort of the
2796.4s
There's a bunch of hedge funds that are
2797.9s
using machine learning to
2800.8s
price things.
2802.2s
And so, your edge has to be that they
2804.7s
are pricing something
2806.4s
not as well as your
2808.6s
AI model.
2810.2s
And that is a hard task. So, I would start pricing things that they're not focused on. And they may not be focused on calibrations, or they might not be focused on crypto, or they're not focused on perpetuals, or some other focused on perpetuals, or some other thing. And that's probably the edge that you want to start with.
2813.1s
So, I would start pricing things that
2814.6s
they're not focused on.
2816.5s
And they may not be focused on
2818.6s
calibrations, or they might not be
2820.2s
focused on crypto, or they're not
2822.3s
focused on perpetuals, or some other
2824.7s
focused on perpetuals, or some other thing.
2825.8s
And that's probably the edge that you
2827.4s
want to start with.
2828.9s
And so, this is more about the market the market clears throat and focusing on the market where like I think sometimes like the market is more valuable than the technology, and just focus on where like they're not focused. I'm a doctor by training, and I'd love to get your idea of when you think about
2831.3s
the market
2832.9s
the market clears throat
2833.3s
and focusing on the market where like I
2835.7s
think sometimes like the market is more
2837.8s
valuable than the technology, and just
2840.0s
focus on where like they're not focused.
2842.2s
I'm a doctor by training, and I'd love
2844.3s
to get your idea of when you think about
2846.1s
investing and building in regulated markets, what frameworks and thoughts you have. And I'm thinking in the back of my mind about things like open evidence where without the support of JAMA, New England Journal of Medicine, these types of tools which are normally behind a paywall, that kind of value proposition isn't
2848.7s
markets, what frameworks and thoughts
2850.9s
you have. And I'm thinking in the back
2852.4s
of my mind about things like open
2853.6s
evidence where without the support of
2856.1s
JAMA, New England Journal of Medicine,
2858.2s
these types of tools which are normally
2859.9s
behind a paywall,
2861.5s
that kind of value proposition isn't
2863.4s
quite as strong. And I'm sure there are plenty of other examples of other regulated markets where the existing data brokers or whichever institutions have a lot of power, how you think about breaking it and and working that space. Yeah, I I think as a early stage startup, you shouldn't try to break
2864.7s
plenty of other examples of other
2866.5s
regulated markets where the existing
2868.7s
data brokers or whichever institutions
2871.6s
have a lot of power, how you think about
2873.5s
breaking it and and working that space.
2876.9s
Yeah, I I think as a early stage
2879.7s
startup, you shouldn't try to break
2886.1s
period of time. You should figure out how to work around them. Um And so, Open Evidence didn't break anything. They just figured out how to build a product that doctors love. Doctors don't want to use Chat GPT even if they gave the same answers. What do they care about? They care about
2887.3s
how to work around them. Um
2889.9s
And so, Open Evidence didn't break
2892.5s
anything. They just figured out how to
2895.0s
build a product that
2897.0s
doctors love. Doctors don't want to use
2900.1s
Chat GPT
2901.6s
even if they gave the same answers. What
2903.6s
do they care about? They care about
2904.8s
do they care about? They care about evidence-based evidence-based uh conversation and And so, instead of just asking Chat GPT, they wanted something that is evidence-based and it was trained on a corpus of body of work that trained on a corpus of body of work that is published. And so, whether it's regulated or not is not the reason
2906.8s
evidence-based uh
2908.4s
conversation and And so, instead of just
2911.5s
asking Chat GPT, they wanted something
2913.2s
that is evidence-based and it was
2914.5s
trained on a corpus of body of work that
2917.8s
trained on a corpus of body of work that is
2919.9s
published. And so,
2923.2s
whether it's regulated or not is not
2926.3s
the reason
2928.0s
the reason um you wouldn't want to break a oligopoly. The reason not to break an oligopoly is they're powerful. And the way you sort of go around them is like you sometimes partner with them in the the sense of like Open Evidence partner with getting all the corpus of articles that have
2929.0s
you wouldn't want to break a oligopoly.
2931.1s
The reason not to break an oligopoly is
2933.5s
they're powerful. And the way you sort
2935.6s
of go around them is like you sometimes
2937.8s
partner with them in the the sense of
2940.9s
like Open Evidence partner with getting
2943.2s
all the corpus of articles that have
2945.7s
ever been published in in or sometimes you work uh around them or with them in a different way. Uh Jordan Ashton work with large enterprise companies for a long period of time. They work with the small medium-size restaurants, not with the the large McDonald's or Starbucks until they had some heft. they had some heft. So, So, the
2948.6s
you work uh
2949.8s
around them or with them in a different
2951.9s
way. Uh Jordan Ashton work with
2956.8s
large enterprise companies for a long
2959.0s
period of time. They work with the small
2960.4s
medium-size restaurants, not with the
2964.1s
the large McDonald's or Starbucks until
2966.7s
they had some heft.
2969.2s
they had some heft. So,
2971.4s
So, the
2972.8s
the essence of the idea is to to work in areas that large companies do not care about. Which is basically That's kind of the reason why I have a job, right? Like if big companies can do everything, I'd be out of a job, but they can't. They're you work in areas where they just don't
2975.1s
idea is to
2977.0s
to work in areas that large companies do
2980.6s
not care about.
2982.4s
Which is basically That's kind of the
2984.4s
reason why I have a job, right? Like if
2987.0s
big companies can do everything, I'd be
2989.6s
out of a job, but they can't. They're
2991.7s
you work in areas where they just don't
2993.5s
you work in areas where they just don't care. Um if you were if you were a student founder in our shoes, how would you spend the first 90 days working on a project or startup idea? One of my takeaways from your talk on this diffusion gap between technology and adoption is that arguably more important
2994.5s
Um if you were if you were a student
2996.1s
founder in our shoes, how would you
2997.5s
spend the first 90 days working on a
2999.3s
project or startup idea? One of my
3001.8s
takeaways from your talk on this
3003.3s
diffusion gap between technology and
3004.8s
adoption is that arguably more important
3007.4s
than building is learning about your customers' behavior and human habits. So, how would you do that? Uh it's specific for the person. I would start with like finding a group of people that you want to work with, that you want to go to battle with. And you know, the example I gave was Tony
3009.2s
customers' behavior and human habits.
3011.9s
So, how would you do that?
3013.8s
Uh it's specific for the person. I would
3016.2s
start with like finding a group of
3018.1s
people that you want to work with,
3020.0s
that you want to go to battle with. And
3021.6s
you know, the example I gave was Tony
3023.7s
Shu Tony Shu and Evan Moore and Stanley Tang and Andy like came together and just started working together. I I would find a set of people you think you want to go to battle with for the next 10 years. And you work together to find out whether you're compatible. You have like
3028.9s
Tang and Andy like came together
3032.2s
and just started working together.
3034.2s
I I would find a set of people you think
3036.2s
you want to go to
3038.4s
battle with for the next 10 years.
3041.9s
And you work together to find out
3043.8s
whether you're compatible. You have like
3046.1s
different skill sets. They happen to be a business person, a marketer, uh product person, and an engineer. Like that's like a clears throat good team. That's a really, really good solid team they came together. Brian Chesky, uh Nate and Joe, they they came together. It was Joe's idea, Brian's leadership, and Nate actually built the first like system.
3048.1s
a business person, a marketer, uh
3051.3s
product person, and an engineer. Like
3053.9s
that's like a clears throat good team.
3055.8s
That's a really, really good solid team
3057.9s
they came together.
3059.2s
Brian Chesky,
3061.0s
uh Nate and Joe, they they came
3063.2s
together. It was Joe's idea,
3065.8s
Brian's leadership, and Nate actually
3068.6s
built the first like system.
3071.2s
These people all work together and for a period of time to find out that they can work together. And then the problem kind of just falls out of that. What are each person passionate about? Why do they care? passionate about? Why do they care? Um is this problem worth working on, etc.
3073.7s
period of time to find out that they can
3075.9s
work together.
3077.2s
And then the problem kind of just falls
3079.3s
out of that. What are each person
3081.0s
passionate about? Why do they care?
3083.6s
passionate about? Why do they care? Um
3085.2s
is this problem worth working on, etc.
3088.6s
I think sometimes people over focus on the problem and not focus on whether you have the and not focus on whether you have the right group of people coming together. And you have lots of people you can work with here. One one one of the biggest pieces of advice my father gave me when
3091.0s
the problem
3092.9s
and not focus on whether you have the
3094.9s
and not focus on whether you have the right
3095.8s
group of people coming together.
3098.2s
And you have lots of people you can work
3099.6s
with here. One one one of the biggest
3101.6s
pieces of advice my father gave me when
3103.5s
I was leaving college, now we're graduating, uh you know, he's meeting my friends, we're having dinner, like, what do you think of my friends? And he's like, Your friends are great. They're really smart. I was like, That's not sounding very positive. What's What's wrong? He's like, They have too many options.
3104.8s
graduating, uh you know, he's meeting my
3107.0s
friends, we're having dinner, like, what
3108.4s
do you think of my friends? And he's
3109.6s
like, Your friends are great. They're
3111.4s
really smart.
3113.0s
I was like, That's
3114.9s
not sounding very positive. What's
3116.7s
What's wrong? He's like, They have too
3118.3s
many options.
3120.5s
And I think all of you have a lot of And I think all of you have a lot of options. And your ability to sort of think about all the different options and optimizing them will prevent you from focusing on the one or two things that can help you build a company.
3122.4s
And I think all of you have a lot of options.
3124.1s
And your ability to sort of think about
3127.9s
all the different options and optimizing
3130.1s
them will prevent you from focusing on
3132.7s
the one or two things that can help you
3134.4s
build a company.
3136.2s
Because at the end of the day, a company has one core idea. One core vision. One core vision. And uh that's very, very hard to say, I'm going to dedicate the next 10 years for that one thing. And it's scary. But you know what? If it doesn't work, you still have your Stanford degree to fall back on.
3138.0s
a company has one core idea.
3141.2s
One core vision.
3143.6s
One core vision. And
3145.3s
uh that's very, very hard to say, I'm
3148.5s
going to dedicate
3150.0s
the next 10 years for that one thing.
3152.8s
And it's scary.
3154.5s
But you know what? If it doesn't work,
3156.8s
you still have your Stanford degree to
3158.1s
fall back on.
3160.2s
All right, I got it. Slightly more laughs the second time around. Curious. I feel like I and the general public love stories about the founders, great founders. I mean, what makes them unique? You mention Brian, you know, Demis the visionary, and Sheryl Sandberg's the business genius. I feel like we hear lots of those stories. And I'm curious
3161.7s
laughs the second time around.
3163.2s
Curious. I feel like
3165.7s
I and the general public love stories
3167.7s
about the founders, great founders. I
3169.8s
mean, what makes them unique?
3171.4s
You mention Brian, you know, Demis the
3174.2s
visionary, and Sheryl Sandberg's the
3176.0s
business genius. I feel like we hear
3178.2s
lots of those stories. And I'm curious
3180.4s
for you, what makes you exceptional at what you do, and what is your unique edge? How do you think about putting your product out? your product out? clears throat clears throat Uh I'm not that good. I mean, if you look at the number of companies we invest in and how many times we're wrong, we're not that good.
3183.1s
what you do, and what is your unique
3185.6s
edge? How do you think about putting
3187.1s
your product out?
3189.2s
your product out? clears throat
3189.8s
clears throat Uh
3191.2s
I'm not that good.
3193.6s
I mean, if you look at the number of
3195.1s
companies we invest in and how many
3196.6s
times we're wrong, we're not that good.
3199.8s
times we're wrong, we're not that good. Um And so, I think the one thing I would say that I'm a little I'm like impervious to pain. I uh don't feel risk the same way as most people feel risk. I think the challenge with hiring people from Stanford at Sequoia has been that
3201.5s
And so, I think the one thing I would
3204.0s
say that I'm a little
3206.4s
I'm like impervious to pain.
3209.4s
I uh don't feel risk the same way as
3212.2s
most people
3213.7s
feel risk.
3215.4s
I think the challenge with hiring people
3217.1s
from Stanford at Sequoia has been that
3219.9s
people don't know how to take risk and with failure. My parents told me I was a failure. I'm never going to amount to anything. I I'm not a good student. And so I never cared what other people And so I never cared what other people think. The best thing The best thing that can happen to you is
3221.8s
with failure.
3223.7s
My parents told me I was a failure. I'm
3226.2s
never going to amount to anything. I I'm
3228.4s
not a good student.
3231.4s
And so I never cared what other people
3233.2s
And so I never cared what other people think.
3234.2s
The best thing The best thing that can
3236.0s
happen to you is
3238.4s
figure out who you are and be your own figure out who you are and be your own person. And I figured that out relatively early in life. The funny thing is like I got good grades because I know how to test well, but I was not a good student. And I never I never fit in anywhere.
3240.2s
figure out who you are and be your own person.
3241.7s
And I figured that out
3243.5s
relatively early in life.
3245.8s
The funny thing is like I got good
3247.1s
grades because I know how to test well,
3249.1s
but I was not a good student. And I
3250.8s
never I never fit in anywhere.
3253.8s
And if you want to go on an entrepreneurial journey, you have to be okay with being an outsider for a long period times. That probably allows me to be a better operator, investor than most because I will take all the signals from the world and process it and make up my own
3256.4s
entrepreneurial journey, you have to be
3259.1s
okay with being an outsider
3262.5s
for a long period times.
3264.7s
That probably allows me to be a better
3268.6s
operator, investor
3271.1s
than most because
3274.0s
I will take all the signals from the
3275.6s
world and process it and make up my own
3278.0s
world and process it and make up my own mind. And I will always sort of try to learn from my failures and not see it as a from my failures and not see it as a failure. Um one way to think about it is there's no such thing as failure. There There's
3279.4s
And I will always sort of try to learn
3282.6s
from my failures and not see it as a
3284.8s
from my failures and not see it as a failure.
3286.6s
Um one way to think about it is there's
3288.9s
no such thing as failure. There There's
3291.6s
There's mistakes that you learn from and there's mistakes you don't learn from. Don't Don't do the latter. You make mistakes, learn from them. There was There was an expensive tuition, but learn from it. What was your first job? Like every single kid at a paper route as a first job. Paper route. Paper route. clears throat
3293.6s
there's mistakes you don't learn from.
3296.0s
Don't Don't do the latter.
3298.6s
You make mistakes, learn from them.
3300.9s
There was There was an expensive
3303.4s
tuition, but learn from it.
3305.3s
What was your first job?
3307.3s
Like every single kid at a paper route
3310.8s
as a first job.
3311.6s
Paper route.
3313.0s
Paper route. clears throat
3313.2s
What was the closest near-death experience you faced during your company's launch? I mean I had a real near-death experience. Isn't that more interesting? Closest near-death. I I was uh I clears throat was driving to Tellme Networks and um I was driving pretty fast. Uh it was during traffic and it was the
3315.8s
experience you faced during your
3317.7s
company's launch?
3319.4s
I mean I had a real near-death
3321.6s
experience. Isn't that more interesting?
3322.9s
Closest near-death.
3324.5s
I I was uh
3326.7s
I clears throat was driving to Tellme
3328.6s
Networks and um
3331.7s
I was
3333.3s
driving pretty fast.
3336.0s
Uh it was during traffic and it was the
3339.1s
first day of rain on 280 and someone cut me off and I was going to hit them and rear end them. So, instead of that, I turned to the right and then my car spun 180. I was in the opposite direction on 280. What about your data here from a What about your data here from a company?
3342.4s
me off and I was going to
3344.6s
hit them
3346.2s
and rear end them. So, instead of that,
3347.9s
I turned
3349.7s
to the right and then my car spun 180.
3353.6s
I was in the opposite direction on 280.
3357.4s
What about your data here from a
3358.7s
What about your data here from a company?
3359.2s
I think I think laughter If somebody would have died, Okay. So, what's the most interesting place or story for working with a founder and finally like signing a term sheet with finally like signing a term sheet with them? I mean, you sign a term sheets. term sheets. I sign a term sheet at a bar.
3360.8s
I think laughter
3361.5s
If somebody would have died,
3367.5s
Okay. So,
3369.0s
what's the most interesting place or
3371.4s
story for working with a founder and
3373.5s
finally like signing a term sheet with
3375.0s
finally like signing a term sheet with them?
3376.2s
I mean, you sign a
3378.5s
term sheets.
3381.7s
term sheets. I
3383.0s
sign a term sheet at a bar.
3389.7s
How many drinks did you have? laughter The founder had three drinks and I had The founder had three drinks and I had none. none. laughter All right. We had a good We had a good signing. Thank you, everyone. Thank you, everyone. applause
3389.9s
The founder had three drinks and I had
3392.1s
The founder had three drinks and I had none.
3394.2s
none. laughter
3398.4s
All right.
3398.8s
We had a good We had a good signing.
3401.1s
Thank you, everyone.
3402.8s
Thank you, everyone. applause
