YT YouTube2Text

YouTube2Text → Transcripts → How Jensen Works

How Jensen Works — Transcript

by Founders Podcast · 10,369 words · 1,555 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00About a year ago, I read this book
0:01called The Nvidia Way by Tay Kim. The
0:03book is a company history and a partial
0:05biography of Jensen. The episode that I
0:07made on the book, the first time I read
0:09it covered a lot of Jensen's life and
0:11then how he built Nvidia. That was
0:13episode 376. For this episode, I reread
0:16the book and then I wanted to strip away
0:19everything that wasn't ideas for how
0:22Jensen runs his company. So, how Jensen
0:24works and nothing else. And so what I
0:27did is I made a list of about 20 of
0:29Jensen's ideas that I want to run
0:32through with you. So starting with idea
0:34number one, which I call Professor
0:37Jensen. So a common trait of history's
0:39greatest founders, they spend a lot of
0:42their time teaching their organization.
0:45The best description of this I ever
0:46heard actually comes from Jim Synagol,
0:47the founder of Costco. He says, "If
0:49you're not spending 90% of your time
0:51teaching, you're not doing your job."
0:54The very first sentence of this book is
0:57in another life, Jensen might have been
1:00a teacher. Colleagues call him Professor
1:02Jensen for his ability to explain
1:04complicated concepts on the whiteboard
1:06in a way that just about anyone can
1:09understand. There's another line in this
1:11book that comes later on where the
1:13author take him says, "I would become
1:14his student much like his employees do."
1:18Jensen's teaching is both persuasive and
1:21pervasive. They say that you can talk to
1:23two unrelated Nvidia employees who don't
1:26know each other and they still will say
1:28the same thing. Jensen has this Vulcan
1:32mindmeld with everybody inside of his
1:34company. He spends a great deal of his
1:36time communicating with his employees
1:38and ensures that everyone at the company
1:40knows the overall strategy and vision
1:42which leads us ties into idea number two
1:45which I've already mentioned and that is
1:47whiteboard. Jensen wants the whiteboard
1:49to be used as the primary form of
1:51communication in meetings. Why? Because
1:54it forces employees to demonstrate their
1:56thought process in real time in front of
1:59an audience and there's no hiding.
2:01Jensen wants drawing and talking at a
2:04whiteboard to be the center of every
2:05meeting. Whiteboarding forces people to
2:07be both rigorous and transparent. It
2:10requires them to start from scratch
2:11every time they step up to the board and
2:14therefore to lay out their thinking as
2:15thoroughly and clearly as possible. It
2:18becomes immediately apparent when
2:21someone hasn't thought something
2:22through. At the whiteboard, there is no
2:25place to hide. And I would say there's
2:27this beautiful underlying tie-in to his
2:29overall company building philosophy. And
2:31it's the fact that he just believes in
2:32constant reinvention. So it says the
2:34whiteboard represents both possibility
2:36and ephemererality. the belief that a
2:39successful idea, no matter how
2:41brilliant, must eventually be erased and
2:44a new one must take its place. So this
2:48deep belief in constant reinvention as
2:51mandatory, as not optional, something I
2:53think that Jensen has in common with
2:54Michael Dell, one of my favorite things
2:56that Michael Dell ever said was this. He
2:57said, "I stood up and told the company
2:59that 5 years from now, we will have a
3:01new competitor. And that new competitor
3:03is going to be in every business that we
3:05are in except they're going to be
3:07faster, more efficient, and more
3:09capable. And they're going to put us out
3:11of business. And the only way that we're
3:14going to prevent that is if we become
3:17that company. It is gut-wrenching stuff
3:20to reinvent and reimagine your business,
3:23but if you don't do it, you'll go out of
3:25business. Which leads to idea number
3:28three, which is complacency kills.
3:31Jensen repeats this belief over and over
3:33again throughout the book. Complacency
3:35kills. I would say that Jensen has this
3:37combination of extreme self-confidence
3:40and charisma matched with this inner
3:42voice that says that he sucks. An inner
3:44voice that says nothing he ever does is
3:47good enough. And I I truly believe for a
3:49lot of history's greatest founders that
3:50inner voice is impossible to satisfy.
3:53And there's a great line in the book on
3:55how this manifests. It says at Jensen's
3:57company, innovation is a necessity, not
4:00an option. I don't know what Jensen's
4:03answer would be if you ask him like what
4:04is his biggest fear? I would highly
4:06suspect that complacency taking root
4:09inside of his company is going to be
4:10towards the top of that list. And so
4:13throughout the entire book, he's just
4:14constantly hounding on this. Here's one
4:17example. Jensen resists overly positive
4:19accounts of Nvidia's startup period and
4:21his own missteps. When we were younger,
4:23we sucked at a lot of things.
4:26And video wasn't a great company on day
4:28one. We made it great over 31 years. And
4:31so in addition to rereading this book
4:33for the second time, I also watched
4:34Jensen's commencement speech that he
4:36gave at Caltech. And as with any
4:38commencement speech, they're going to
4:39give you a nice introduction. They're
4:40going to give you like a short
4:41biography. And after the nice
4:43introduction, he begins his speech by
4:46saying, "Thanks for the kind
4:47introduction. It really makes me cringe
4:50listening to all of that. I hate hearing
4:52about myself." In the book, he
4:54summarizes his first 15 years as CEO.
4:59This is what it says. He slipped into
5:01the third person. If Jensen wasn't even
5:04involved in the first 15 years of our
5:06company, I would really like that. He
5:09wasn't proud of how the company was
5:11managed then, or his own naive and lack
5:14of strategic thinking. Jensen runs the
5:16company the way he does because he
5:18believes that Nvidia's worst enemy is
5:20not the competition, but itself. The
5:23worst enemy is the complacency that
5:25grips any successful company. That is
5:28yet another thing that Jensen has in
5:30common with a lot of history's greatest
5:32founders. They all believe that
5:33complacency kills. In fact, Andy Grove,
5:35he had a mantra. His mantra was success
5:38breeds complacency. Complacency breeds
5:41failure. Only the paranoid survive. This
5:44idea of complacency killing your company
5:46just reappears over and over again.
5:47Here's another example. Herb Keller,
5:49founder of Southwest Airlines, he has a
5:50great way to describe this. A company is
5:52never more vulnerable to complacency
5:54than when it's at the height of its
5:55success. We must not let success breed
5:58complacency, cockiness, greediness,
6:01laziness, indifference, preoccupation
6:04with bureaucracy, hierarchy, or
6:07obliviousness to threats posed by the
6:10outside world. I read this great profile
6:13on the founder of LVMH, Bernard. It
6:16says, "He abhores complacency so much."
6:19Warren Buffett would repeat this over
6:21and over again in his shareholder
6:22letters. He says, "You need the ability
6:24to fight off the ABCs of business decay.
6:27Arrogance, bureaucracy, and complacency.
6:31When these corporate cancers
6:32metastasize, even the strongest of
6:35companies falter." And so, there's so
6:37many examples in the book of Jensen
6:39fighting back against complacency. One
6:41example early in their history says that
6:43each monthly company meeting, Jensen
6:45would say, "We're 30 days from going out
6:46of business." all the way back in 97. He
6:50talks about making sure that we're
6:51acting with a sense of urgency because
6:53one of the biggest companies in the
6:54world is trying to kill us. He says we
6:56need to kill Intel. Make no mistake,
6:58remember saying this in 97. Make no
7:00mistake, Intel is out to get us and put
7:03us out of business. They have told their
7:05employees and they have internalized
7:07this. They are going to put us out of
7:09business. Our job is to go kill them
7:12before they put us out of business. We
7:14need to kill Intel. I also wanted to
7:18include this in here because obviously
7:19he's a extremely unusual person and he's
7:22at the top right now, but he was like
7:24this for decades. At the time he just
7:27said this, he's telling his company,
7:29"We're going to go kill Intel." At that
7:31time, Intel was 860
7:35times larger than Nvidia in terms of
7:38revenue. Jensen abhores complacency.
7:42Only the paranoid survive. Again, that's
7:43a quote from Andy Grove. What's
7:45interesting about that is Andy Grove
7:47mentored Steve Jobs and Steve Jobs had a
7:49very clear philosophy about what to do
7:52after achieving something great and it's
7:54a way to again avoid complacency taking
7:56root in your company. This is what Steve
7:58said. I think if you do something and it
8:00turns out pretty good then you should
8:01just go do something else wonderful and
8:04not dwell on it for too long. Just
8:07figure out what's next. That is a
8:09mentality that history's greatest
8:11founders have. This is the way I
8:12summarize it. No rearview mirror. No
8:14resting on laurels or sleeping on wins.
8:16Make something great, then do it again.
8:19And one of the things I admire most
8:20about Jensen is this relentless
8:22dedication to improving his craft. In
8:25fact, craft is the word that he uses in
8:28that commencement address that he gave
8:29at Caltech. And Jensen's dedication to
8:32constantly improving his company reminds
8:33me of my friend Kareem, who's the
8:35co-founder and CTO of RAMP. Ramp is the
8:37presenting sponsor of this podcast. And
8:40Kareem is one of the greatest technical
8:41minds working in finance. I spend a lot
8:44of time talking to Kareem and every
8:46single conversation centers around his
8:48obsession with crafting a highquality
8:50product and using the latest technology
8:52to constantly create better experiences
8:54for his customers just like Jensen does.
8:57Kareem is running one of the most
8:58talented technical teams in finance and
9:00they use rapid relentless iteration to
9:02make their product better every day. So
9:05far this year, RAMP has shipped over 300
9:07new features. Ramp is completely
9:09committed to using AI to make a better
9:11experience for their customers and
9:13automate as much of your business's
9:15finances as possible. In fact, Kareem
9:17just wrote this. AI is all I think about
9:20these days. It is our duty to be first
9:22movers and push limits so we can make
9:24the greatest possible product experience
9:26for our customers. Many of the fastest
9:28growing and most innovative companies in
9:30the world are running their business on
9:31RAMP. Make sure you go to ramp.com to
9:34learn how they can help your business
9:36save time and money today. Let AI chase
9:38your receipts and close your books so
9:40you can use your time and energy
9:42building great things for your
9:43customers. Get started today by going to
9:45ramp.com. Number four, Jensen insists on
9:49a flat organization. Jensen has 60
9:53direct reports and he doesn't do
9:55one-on-one meetings. And so he talks
9:57about why he designed a flat
9:59organization. When we were a small
10:01company, we were plenty bureaucratic and
10:03plenty political. Jensen thought about
10:05how he could create an ideal
10:06organization from scratch. He realized
10:09he would choose a much flatter structure
10:11so that employees could act with more
10:13independence. He also saw a flat
10:15structure would weed out lower
10:16performers who were unaccustomed to
10:19thinking for themselves and to acting
10:21without being told what to do. This is
10:23what he said. I wanted to create a
10:25company that naturally attracts amazing
10:27people. A flat structure fights against
10:29the danger of slow decision-making. It
10:32is very interesting that Jensen picked
10:34up on how dangerous slow decision-m is.
10:37Jeff Bezos in one of his biographies
10:38said this. You can drive great people
10:40away by making the speed of decision-m
10:42really slow. Why would great people stay
10:44in an organization where they can't get
10:46things done? Jensen will describe his
10:48flat organization structure in the book
10:50and also a bunch of interviews. Here's
10:51an interview. In one of the interviews,
10:53he actually compared how he runs his
10:54company to what happens and what will
10:56happen when humanity is surrounded by
10:58all these AI agents that are smarter
11:00than they are. And he was talking about
11:02the fact that he gets that question like
11:03what's going to happen if you're
11:03surrounded by all these AIs are doing
11:05things so incredibly well and they're so
11:06much better than you. And he says when I
11:07reflect on that that's already my life.
11:10I have 60 direct reports. The reason why
11:12they're on my e staff is because they're
11:14world class at what they do and they do
11:17it better than I do. Much better than I
11:19do. And I have no trouble interacting
11:21with them. I have no trouble prompt
11:24engineering them. I have no trouble
11:26programming them. And what is
11:27fascinating is he refuses to change his
11:30management philosophy. He constantly
11:32insists on maintaining this flat
11:34structure. Says he did not want to spend
11:36time on career coaching because the
11:37majority of them had already reached the
11:39pinnacle of their careers. Instead, he
11:41focused on providing them collectively
11:42with information from across the
11:44organization as well as with his own
11:46strategic guidance. Jensen had
11:48steadfastly refused to change his
11:51management philosophy, even when new
11:53board members joined and recommended
11:55that he hired a chief operating officer
11:58to reduce his administrative burden.
12:00This part made me laugh. No thanks,
12:02Jensen replied. This is a great way to
12:05make sure everybody knows what's going
12:07on, referring to his direct
12:08communication with the rest of the
12:10company. And so when I'm reading to it,
12:11like I'm pulling all these these quotes
12:13and ideas are not in one section of the
12:15book. They are repeated over and over
12:16again throughout different sections of
12:17different time periods. That's how you
12:18know it's important important to him.
12:20And one of the ways I would summarize
12:22what he's about to say here is like, oh
12:23my god, Jensen made like an F1
12:25organization. Ultimately, my E staff is
12:28something that I have to know how to
12:30work with. The company's organization is
12:32like a race car. It has to be a machine
12:35that the CEO knows how to drive. Jensen
12:38created a company he could manage
12:40directly. That sentence is very
12:42important. Jensen created a company that
12:44he could manage directly. He created a
12:46company building philosophy that fit
12:48him. It doesn't matter what other people
12:50are doing. The root principle is to do
12:52things your way. And so he would repeat
12:54over and over again all the benefits of
12:55keeping his company flat. It turns out
12:57by having a lot of direct reports, not
12:59having one-on ones, we made the company
13:01flat and information travels quickly and
13:05therefore employees are empowered. That
13:07algorithm was wellconceived. You want a
13:09company that's as large as necessary to
13:12do the job, but to be as small as
13:14possible and not bogged down by
13:16overmanagement and process. Now, idea
13:19number five works very well with idea
13:21number four. Public criticism works well
13:23with a flat organization. Jensen does
13:25not believe. You hear this like a common
13:27trope that's repeated. And maybe it
13:29makes sense if your idea is to get along
13:31with people, which I think Jensen is
13:32like the the quality of the work
13:34obviously goes it's like top priority
13:36more than you know other people's
13:37feelings. Jensen doesn't believe in
13:38praise publicly criticized privately. He
13:41criticizes publicly so the entire
13:43organization can learn from a single
13:45person's mistake. Again, this is another
13:47reoccurring theme throughout the book.
13:49That's what he said. Over the years, I
13:50realized what was happening and how
13:52people protect their turf and they
13:53protect their ideas. I created a much
13:55flatter organization. Jensen's anecdote
13:57to the backstabbing, the gaming of
13:59metrics, and to political infighting is
14:02public accountability and if needed,
14:04public embarrassments.
14:06I just say it out loud. He said, I've
14:08got no trouble calling people out. And
14:11so, here's an example of Jensen's public
14:13criticism so the entire company can
14:15learn from the mistake. And I think it
14:17also ties together with that constant
14:19fight against complacency. Jensen was
14:21furious at the poor planning and
14:22execution of the chip. He called out its
14:25engineers at an all company meeting. Is
14:27this the piece of that you intended
14:29to build? He said, "The architects did a
14:31shitty job putting the product together.
14:34How could you not see the issue before
14:35it happened? Someone should have raised
14:37their hand and said, "Hey, we have a
14:39design issue here." Jensen's criticism
14:42didn't end after a single meeting. He
14:44invited an executive from Best Buy to
14:46speak to his employees. The executive
14:48spent much of the session talking about
14:51the NV30's poor performance and customer
14:54complaints. Jensen agreed. He's right.
14:58This is crap. Jensen reflected on the
15:00NV30s failure. Ultimately, it was his
15:03responsibility to ensure that Nvidia's
15:06teams collaborated effectively no matter
15:08how large the company got. Again, what I
15:11think is remarkable about reading these
15:12books about, you know, you obviously
15:15know this, like I'm obsessed with people
15:16that do things for a long time. Jensen
15:19is the longest running founder CEO of a
15:21tech company in the world. And the
15:23result of hopefully this podcast and
15:24obviously the book is we get to benefit
15:26from Jensen's pain and experience. All
15:30of these ideas that you and I are
15:31talking about today come from over three
15:33decades of him building his company into
15:35one of the best companies in the world.
15:37Now the way he thinks about this, it's
15:39very similar because I I do think this
15:40is one of the most unique ideas that you
15:42will hear and it reminded me of this
15:43conversation. I was watching this
15:45interview with uh Johnny IV one time and
15:46this is after Steve Jobs passed away. He
15:48talked about one of the most important
15:49lessons that Steve told him and it
15:51sounds exactly Jensen's not trying to
15:53hurt people's feelings just to hurt
15:54their feelings. He's trying to refine
15:56the character of his company. Listen to
15:59Johnny Iive talk about this. I remember
16:01asking Steve about this. I said it could
16:03be perceived that his critique of a
16:05piece of work and that in his critique
16:07of a piece of work he was a little harsh
16:09and we had been putting our heart and
16:10soul into this and so I asked him could
16:12we not moderate the things we said a
16:14little bit and he asked why and I said
16:17because I care about the team and then
16:20Steve said this brutally brilliant
16:23insightful thing. He said no Johnny
16:25you're just really vain. You just want
16:28people to like you. I'm surprised at
16:30you. I thought you held the work up as
16:33the most important, not how you believe
16:35you were perceived by other people. And
16:38Johnny replied, "And I was terribly
16:39cross because I knew he was right."
16:43Going back to why Jensen does this. He
16:44offers employees direct criticism in
16:46large meetings so more people can learn
16:48from a single mistake. I give feedback
16:50in front of everybody. Feedback is
16:53learning. For what reason are you the
16:55only person who should learn from this?
16:57We should all learn from that
16:59opportunity.
17:01I don't take people aside. We are not
17:03optimizing for not embarrassing
17:05somebody. We're optimizing for the
17:07company learning from our mistakes.
17:10Idea number six, tortured into
17:13greatness. A habit of self-criticism.
17:15Jensen is not just publicly criticizing
17:18others. He does this to himself. He
17:20actually has a great line about this. He
17:22calls it tortured into greatness. And so
17:23I mentioned earlier that Jensen has this
17:25combination of extreme self-confidence
17:26and charisma matched with this inner
17:28voice that says, you know, he sucks. An
17:30inner voice that says nothing ever that
17:32nothing he ever does is going to be good
17:34enough. And what you learn by reading
17:35this book is that he admittedly tortures
17:38his team into greatness. In fact, he has
17:40a line about this. He says, "I don't
17:41like giving up on people. I'd rather
17:43torture them into greatness." And it's
17:45clear from that that he sees this as a
17:47guard against complacency. He feels
17:49praise is a distraction and the
17:50deadliest sin is looking back on past
17:52accomplishments. But after you read this
17:55book, you realize that he has tortured
17:57himself into greatness as well. That it
17:59started with him. And so there's this
18:02great story in the book about this habit
18:04of self-criticism that he has. And it's
18:06told by an executive at the company. He
18:08says, "I'll never forget this. We had
18:09done a fantastic job. We just blew the
18:11doors off the quarter." And Jensen stood
18:13up in front of us and said,"I look in
18:16the mirror every morning and say, "You
18:18suck." The executive was struck by how
18:20someone so manifestally successful could
18:23still think in such terms. You and I
18:25have seen that idea over and over and
18:27over again. And I think it's tied again.
18:29He talks about one of my favorite maybe
18:31my all-time favorite maximum from the
18:32history of entrepreneurship is
18:34excellence is the capacity to take pain.
18:36I've never heard Jensen speak in that
18:38exact terms, but it's the same idea. I'm
18:40going to give you two examples. I'm
18:42going to pull two quotes that he said.
18:43One is from a speech at Caltech, which I
18:45referenced earlier, and another is a
18:47speech at Stanford. This is what he said
18:48at Caltech. This is my favorite thing he
18:50said in the entire commencement address.
18:52I hope you will see setbacks as new
18:54opportunities. Your pain and suffering
18:56will strengthen your character, your
18:58resilience, and agility, and they are
19:00the ultimate superpowers. This is my
19:03really my favorite part of my favorite
19:04part of all the things that I value most
19:07about my abilities, intelligence is not
19:09top of that list. My ability to endure
19:12pain and suffering. My ability to work
19:14on something for a very very long period
19:16of time. My ability to handle setbacks
19:19and see the opportunity just around the
19:21corner. I consider to be my superpowers
19:24and I hope they're yours. Says something
19:27very similar to that when he gave a talk
19:29at Stanford. I think one of the great
19:30advantages is I have very low
19:32expectations. Most of the Stanford
19:34graduates have very high expectations.
19:37People with high expectations have very
19:39low resilience. And resilience matters
19:41in success. I don't know how to teach it
19:44to you except that I hope suffering
19:46happens to you. I was fortunate that I
19:49had plenty of opportunities for setbacks
19:51and suffering. To this day, I use the
19:53phrase pain and suffering inside of our
19:56company with great glee. And I mean that
19:59in a happy way because you want to
20:01refine the character of your company.
20:03Now, let me pause there. I think that's
20:05exactly what a lot of Jensen's ideas do.
20:09They refine the character of your
20:11company. So let's go back to that. I
20:12mean that in a happy way because you
20:14want to refine the character of their
20:16company. You want greatness out of them.
20:19Greatness is not intelligence. Greatness
20:21comes from character. And character is
20:23not formed out of smart people.
20:25Character is formed out of people who
20:27suffered. I wish upon you ample doses of
20:31pain and suffering. Idea number seven,
20:34speed of light. Not only has he derived
20:37really great ideas from three, you know,
20:39over three decades of experience, but
20:40he's good at naming them and branding
20:42his ideas. So Jensen's speed of light
20:44idea is basically, hey, the only
20:46benchmark that we're going to judge how
20:47fast we move will be the absolute
20:49maximum speed possible. What does that
20:52mean? Jensen insists that all employees
20:54work at the speed of light. He wants
20:56their work to be contained only by the
20:58laws of physics. Sounds like Elon. Each
21:01project must be broken down into
21:03component tasks and each task must have
21:05a target time to completion that assumes
21:07no delays, no cues and no downtime. This
21:11sets the theoretical maximum, the speed
21:13of light that it is physically
21:15impossible to exceed. We will then judge
21:18ourselves against the speed of light,
21:21not what we used to do or what other
21:23companies are doing. As he saw it, he
21:26needed to prevent the kind of internal
21:27rot that he observed at other companies.
21:29Again, what's that reference to? That's
21:31a reference to idea number three.
21:32Complacency kills. Jensen would not let
21:35his employees think about what was
21:36likely to work or what they could
21:38reasonably achieve. He only cared about
21:40what would be possible with the maximum
21:42amount of effort and minimum amount of
21:45wasted time. Minimum amount of wasted
21:49time. Before we get into idea number
21:50eight, there is another tool and
21:52technology that I need to tell you about
21:53that does exactly that. And RAMP
21:56actually uses this tool. It is Vanta. I
21:58just told you earlier how RAMP has the
22:00best technical talent in their industry.
22:02And knowing that the team at RAMP trusts
22:03Vanta was a huge positive signal for me.
22:05And then when I talked to Vant's
22:07founder, Christina, it became obvious
22:09why Ramp uses Vanta because Vanta's
22:11value prop is very clear. Vanta helps
22:13your company prove that you're secure so
22:14more customers will use your product or
22:16service. Many companies won't sign
22:18contracts unless you're certified and
22:20this is causing you to lose out on
22:22sales. That is why the average Vanta
22:24customer reports a 526% return on
22:26investment after becoming a Vanta
22:28customer. Vanta helps your company
22:30automate compliance, security, and
22:32trust. Vanta helps your company pass
22:34audits without tons of manual work. So,
22:36not only do you make more money with
22:37Vanta, but you save more time. Manual
22:40compliance is slow and painful. The best
22:42companies and the best founders will not
22:44tolerate wasting valuable company time
22:46doing something with labor when
22:47technology can help you move faster,
22:49maybe even move at the speed of light
22:50even. And Vanta helps you do just that.
22:52Vanta will help you win trust, close
22:54deals, and stay secure faster and with
22:56less effort. Go to vanta.comfounders
23:00to learn more, and you'll get $1,000
23:02off. That is vanta.com/founders.
23:05Idea number eight. This is a combination
23:08of reoccurring themes as you see
23:09throughout the book. Jensen is
23:10unapologetically extreme. He is extreme
23:13in all things. He works all the hours,
23:15and he insists that no one will outwork
23:18him. Jensen is so relentless that he
23:20will grill you at the urinal. It's one
23:23of the the funniest uh stories in the
23:25book. No place in in company
23:27headquarters was safe from a driveby
23:30grilling from Jensen. Kenneth Hurley was
23:32at a urinal when Jensen walked up to the
23:34one next to him.
23:36I'm not the kind of guy who likes to
23:38talk in the bathroom, Hurley said.
23:39Jensen had other ideas. Hey, what's up?
23:41He asked. Hurley replied with not much,
23:44which earned him a sidelong glance from
23:46Jensen. Hurley panicked, thinking, "I'm
23:48going to get fired because he thinks I'm
23:49not doing anything." He proceeded to
23:51list 20 things he was working on.
23:53"Okay," Jensen replied, satisfied with
23:55the answer. So, Jensen is described as
23:57extremely hardworking by everybody that
23:59runs into him. And he also describes
24:01himself that way. He says, "There may be
24:03people that are smarter than me, but no
24:05one is ever going to work harder than
24:06me." This is very interesting. And I was
24:08reading the way that Jensen thinks about
24:10his own work ethic and he has a great a
24:12bunch of great lines that now me and my
24:13friends have also used from the first
24:14time uh the first episode the first time
24:15we read the book. But for some reason
24:17when I got this to this part the second
24:19time I was reading it it there's this
24:21idea that I thought was fascinating
24:22where like Larry Ellison likes having an
24:24enemy. He says that he would pick an
24:26enemy and then he would use that to push
24:29himself and his company. And one enemy
24:32that he picked was Bill Gates. in the
24:33whole thing that Larry Larry's point was
24:35like, you want a a foe worthy of your
24:37steel. You want a formidable enemy. And
24:40so in the middle of this war that's
24:42going on, this battle between Ellison
24:43and Gates, Ellison gives this interview
24:45and he talks about he understands Bill
24:47Gates very well. And this is how he
24:49described Bill Gates. He says there are
24:50a lot of people in the world smarter
24:51than Bill Gates. There are very few
24:53people in the world that have his focus
24:54and endurance. He is utterly relentless.
24:57He is indehaticable. He is absolutely
25:00focused and he wants it all. Barry
25:02Diller said that Bill is young and he's
25:05mean and he's not tired. That is a high
25:08compliment coming from Barry Diller.
25:11That's like Jensen. Very few people that
25:13I've ever come across have Jensen's
25:16focus and endurance. And he doesn't hide
25:18this. He is unapologetically extreme.
25:21Jensen said that he works all the hours
25:23and any hour that he is not working, he
25:25is thinking about work and he has a very
25:27low tolerance for anybody complaining
25:28about work. So there's a bunch of
25:30employees that were griping about long
25:31work hours and Jensen's response was
25:34typically direct. People who train for
25:35the Olympics grumble about training
25:37early in the morning, too. Jensen was
25:39sending a message. Long hours were a
25:41necessary prerequisite for excellence.
25:43To this day, he has not deviated from
25:45that view. He also was unapologetically
25:48extreme in his desire and his need, his
25:50deep need to be number one. The first
25:53time we came in second place, Jensen
25:55sternly told me, "Second place is the
25:58first loser." I realized I'm working for
26:00a boss who believes that we have to win
26:02at everything. Jensen's guidance was
26:04direct. Just go win. The idea was
26:08whoever could run faster would get all
26:10of the land. And Jensen is extreme in
26:12all things, including recruiting. He was
26:14trying to recruit the chief engineer
26:16from Silicon Graphics. So he invites the
26:18guy out to lunch and he goes for again
26:20his direct approach which I think is
26:22probably the only approach I think
26:23direct is the only way he approaches
26:25anything and he tells him he goes,
26:27"John, you should really think about
26:28coming to our company because ultimately
26:30I'm going to put SGI out of business."
26:33He is absolutely relentless about work.
26:35This is what he says. You have to allow
26:37yourself to be obsessed with your work.
26:39I work every day. There's not a day that
26:41goes by that I don't work. If I'm not
26:43working, I'm thinking about working.
26:46Working is relaxing for me. That line me
26:48and my friends repeat to each other.
26:50Working is relaxing for me. And then
26:53just one more line about him being
26:54unapologetically extreme. As Jensen put
26:56it when looking directly into my eyes,
26:58the secret to his company's success is
27:00nothing more than sheer will.
27:04Idea number nine, Jensen's top five
27:06email idea. It's a genius way to get
27:09unfiltered information from the entire
27:11company. As companies grow, as they
27:13become more successful, bad leaders wind
27:15up surrounding themselves with sickop
27:16fans or executives who are not
27:18incentivized to tell you the truth. In
27:20fact, I just started doing this new show
27:22where I actually sit down and talk to
27:23extreme winners in business. It's called
27:24David Sunra. And the very first episode
27:26was with my friend Daniel Ek, who is the
27:28founder of Spotify. And you got to
27:30listen to it if you haven't listened to
27:31it yet. He because he tells a hilarious
27:33story about this. And this story
27:34happened years ago, but one of Daniel's
27:36employees who now he was the head of
27:38product at Spotify. His name's Gustaf,
27:39but now he's one of the co-CEOs. After a
27:41product meeting, he pulled Daniel aside
27:43and he's like, "Hey, you really suck at
27:44this." And Daniel had a very human, he
27:46tells the story, he had a very human
27:48reaction where he goes home, he's like,
27:49"Oh my god, I have to fire this guy."
27:52And then he thought about it more and
27:53more and when his emotions calm down,
27:54he's like, "Oh my god, Gustaf is right."
27:57And so Daniel's point is that you want
27:59people to tell you the truth. Well,
28:00Jensen has tens of thousands employees.
28:03How is he going to make sure that he
28:04gets unfiltered true information from
28:06the entire company? He does this with a
28:09top five his top five email idea. So,
28:12Sister Jensen asked employees at every
28:13level of the organization to send an
28:15email that detailed the top five things
28:16they were working on and what they had
28:18recently observed in their markets,
28:20including customer pain points,
28:21competitor activities, technology
28:23developments, and the potential for
28:24project delays. The ideal top five email
28:28is five bullet points where the first
28:30word is an action word. It has to be
28:33something like finalize, build, or
28:35secure. To make it easier for himself to
28:37filter these emails, Jensen had each
28:40department tag them by topic in the
28:42subject line. That way, if he wanted to
28:44get all his recent emails on, say,
28:45hyperscaler accounts, he could easily
28:47find them through a keyword search. The
28:50top five emails became a crucial
28:52feedback channel for Jensen. He would
28:55tell his employees when asked why he
28:57liked the top five process. It's easy to
29:00pick up on the strong signals, but I
29:02want to intercept them when they are
29:04weak. Every day, he would read about a
29:06hundred top five emails to get a
29:08snapshot of what was happening within
29:09the company. There's another article uh
29:11written by the Wall Street Journal that
29:12I read about this. It says, "Employees
29:14have been sending notes known as T5Ts or
29:16top five things for decades, and Jensen
29:18has been reading them. The top five
29:19emails have become his preferred method
29:21of flattening hierarchy. Strategy isn't
29:23what I say, it's what they do," Jensen
29:25said. So, it's really important that I
29:28understand what everybody else is doing.
29:29And one way he does that is by reading
29:30these top five emails. He doesn't want
29:32information that's already made its way
29:33through layers of management. What he
29:35wants is information from the edge. That
29:37is a direct quote from him. He wants
29:38information from the edge. He's looking
29:40for the next 0 billion market, a
29:43frontier that hasn't been explored
29:45because it barely exists, but could one
29:47day be a thing. One of the weak signals
29:49that he intercepted years ago was a
29:51wonky but exciting development in
29:53machine learning that kept popping up in
29:55the emails. Jensen decided that Nvidia
29:57needed to invest more in tools for
29:59accelerating workloads on its GPUs.
30:01These days, those GPUs are the brains
30:04for artificial intelligence. Reading
30:06these emails is the thing that he did
30:08for fun. He says, "I drink a scotch and
30:10I do emails." Idea number 10, Jensen's
30:14communication style. It is blunt,
30:16concise, and direct. Jensen believes in
30:19direct, blunt communication. Jensen's
30:21emails are short and sweet like a ha
30:23coup. Napoleon was the exact same way.
30:26There's a great quote from this
30:27biography Napoleon I read. Long orders
30:29which require much time to prepare, to
30:32read and to understand are the enemies
30:34of speed. Napoleon could issue orders of
30:36a few sentences which clearly expressed
30:38his intentions and required little time
30:40to issue and to understand. This is yet
30:43another thing that Jensen had in common
30:45with Steve Jobs. There's this great book
30:46called Creative Selection which was
30:48written by a person that demoed directly
30:50to Steve. I covered the book all the way
30:52back on episode 281. Here's the quote.
30:55Steve was always easy to understand. He
30:58would either approve a demo or he would
31:00request to see something different next
31:01time. Whenever Steve reviewed a demo, he
31:04would say, often with highly detailed
31:06specificity, what he wanted to happen
31:08next. He was always trying to ensure the
31:11products were as intuitive and
31:12straightforward as possible. And he was
31:14willing to invest his own time, effort,
31:16and influence to see that they were. I
31:19chuckle to myself over how much time
31:21I've spent thinking about this iPad demo
31:24and how much Steve taught me in one
31:26meeting where he spoke just four
31:29sentences. Think about that. Get your
31:31point across and make it memorable in
31:34four sentences. That is worldclass
31:37communication. Idea number 11, Lua. When
31:40an employee starts rambling, Jensen will
31:43say Lua. Lua is a warning sign that
31:47Jensen's patience is growing thin. When
31:49he says it, Jensen wants the employee to
31:51stop and do three things. Number one,
31:54listen to the question. Number two,
31:56understand the question. Number three,
31:58answer the question. Everyone who works
32:01for Jensen has heard Lua.
32:05And so, just like Steve Jobs, Jensen is
32:07easy to understand. And when a person
32:09and their ideas are easy to understand,
32:11their ideas are easy to spread. The best
32:14storytellers know this and you'll find a
32:16lot of the best founders are also great
32:17storytellers. The best story wins and
32:20that is exactly what my partner
32:22collateral helps you do. Don Valentine
32:24who's the founder of Sequoia said one of
32:25my favorite things ever. He says the art
32:27of storytelling is critically important.
32:29Most of the entrepreneurs who come to
32:31talk to us can't tell a story. Learning
32:33to tell a story is incredibly important
32:35because that's how the money works. The
32:38money flows as a function of the
32:40stories. And that is exactly what
32:42Collateral does. Collateral can
32:44transform your complex ideas into
32:46compelling narratives. I need you to
32:48remember their website, which is easy to
32:50remember because it's collateral.com.
32:52Collateral crafts institutional-grade
32:54marketing collateral. And they do this
32:55for private equity, private credit, real
32:57estate, venture capital, family offices,
32:59hedge funds, oil and gas companies, all
33:01kinds of corporations. There are big
33:03companies now using collateral because
33:05they discovered it from this podcast. I
33:07have friends that have used collateral
33:09for their marketing collateral that have
33:10raised billions of dollars of capital
33:12and have made hundreds of millions of
33:14dollars. I will leave a link down below,
33:16but make sure you go to collateral.com
33:19and improve the way that your company
33:20tells its own story. Storytelling is one
33:22of the highest forms of leverage and you
33:24should invest heavily in it. And you can
33:26do that by going to collateral.com.
33:28Number 12, the mission is the boss and
33:31pilot in command. Jensen tells his
33:34employees that the ultimate boss is the
33:35mission itself. This is what he said.
33:37The concept of the mission is the boss
33:39makes a lot of sense because ultimately
33:41we're here to realize a particular
33:43mission, not in service of some
33:45organization.
33:46Under the mission is the boss
33:48philosophy, Jensen would start every new
33:50project by designating a leader, which
33:52he calls a pilot in command. That pilot
33:55in command will report directly to
33:57Jensen. We always have a pilot in
33:58command for every project. Whenever
34:00Jensen talks about any project, he
34:02always wants the name. Nobody can hide
34:05behind such and such team is working on
34:08that. Everything has to have a person's
34:10name attached. The pilot in command is
34:13accountable. I just did this episode,
34:15episode 399 on how Elon works. In that
34:18episode, it talks about Elon's algorithm
34:20that has five commandments that he uses
34:22and applies relentlessly and repeats it
34:24over and over again. In the first
34:25commandment, it says this, "Question
34:27every requirement. Each should come with
34:29the name of the person who made it. You
34:31should never accept that a requirement
34:32came from a department such as the legal
34:34department or the safety department. You
34:36need to know the name of the real person
34:38who made that requirement. In other
34:40words, find out who is accountable.
34:41Jensen organized his employees into
34:44groups centralized by function. So like
34:46sales, engineering, operations, and so
34:48on. They were treated as a general pool
34:50of talent and not divided by business
34:53units or divisions. We take the people
34:55that we have and we're able to redirect
34:57them into a new mission. That is mission
35:00is the boss and pilot in command. Number
35:0313. Strategy is not words. Strategy is
35:06action. This is what Jensen said.
35:08Strategy is not words. Strategy is
35:10actions. We don't do a periodic planning
35:13system. The reason for that is because
35:15the world is a living breathing thing.
35:19We just plan continuously. There's no
35:22five-year plan. So strategy is not
35:24words. Strategy is action. That's one of
35:26my favorite lines in the book. It
35:27reminds me very much of Henry Singleton.
35:29So Henry Singleton had this very
35:31distinct philosophy about planning. I
35:32discovered Henry Singleton because
35:34Charlie Mer said that Henry Singleton
35:36was the smartest person he ever met. And
35:38Warren Buffett said that it's a crime
35:39that business schools don't study
35:41Singleton. And Singleton famously said,
35:43"I know a lot of people have very
35:44definitive plans, but we're subject to a
35:47great number of outside influences on
35:48our businesses and most of them can't be
35:51predicted. So my plan is to stay
35:53flexible. My only plan is to keep coming
35:56to work every day. I like to steer the
35:58boat each day rather than plan way into
36:01the future. And so Jensen has that in
36:03common with Singleton. He also has
36:05something in common with Michael
36:06Bloomberg. That line where Jensen says,
36:08"There's no fiveyear plan. We just plan
36:11continuously. There's no five-year
36:13plan." That is exactly the same line of
36:16thinking that Michael Bloomberg used to
36:18build his company. In Bloomberg's
36:20autobiography, he wrote this. Life, I
36:22found, works the following way. Daily,
36:24you're presented with many small and
36:25surprising opportunities. To succeed,
36:27you must string together many small
36:29incremental advances rather than
36:31counting on hitting the lottery jackpot
36:33once. As a practical matter, constantly
36:36enhance your skills, put in as many
36:38hours as possible, and make tactical
36:40plans for the next few steps. Then,
36:42based on what actually occurs, look one
36:44more move ahead and adjust the plan.
36:47Don't devise a five-year plan or a great
36:49leap forward. Central planning didn't
36:51work for Stalin or Mao, and it won't
36:54work for an entrepreneur either. Number
36:5614, ship the whole cow. The innovator's
36:59dilemma is one of Jensen's favorite
37:01books. One of the things that he learned
37:02from the book is that the threat often
37:03comes from the low end of the market.
37:05So, this is what he said. We build
37:07Ferraris. All of our chips were designed
37:09for the high-end, the best performance.
37:11I don't want to let someone come in and
37:13be the price leader, lock me out of the
37:15bottom, and then climb their way to the
37:17top. And so, Jensen saw that they could
37:19stop throwing away parts that failed
37:21quality tests. These parts were not
37:23suitable for the company's Ferrari grade
37:24chips, but they were otherwise
37:26functional at lower speeds. Nvidia could
37:29repackage them into a less capable and
37:31cheaper version of the company's
37:33mainline products. This looked like a
37:36clear opportunity to make something out
37:38of nothing. rejected parts were
37:40generating no revenue. Nvidia could
37:43create a whole new derivative product
37:45line that could turn a profit without
37:48the expensive and time-consuming process
37:50of research and development. That line
37:52would serve as defense against
37:54competitors for whom the lowcost chip
37:56was the main product. Nvidia could
37:59easily afford to price its chips down so
38:01much that its competitors would be
38:04forced to sell at a loss.
38:07This strategy was dubbed ship the whole
38:09cow, a reference to how butchers find
38:12ways to use almost every part of the
38:14carcass, not just the prime cuts. Number
38:1915, go to school on everybody. There's a
38:21line in a biography of Jeff Bezos that
38:23says, "Jeff went to school on
38:24everybody." Jensen goes to school on
38:27everybody. There's a story in the book
38:28where he shows up at this conference.
38:30It's an academic conference where
38:31machine learning and neuroscience
38:32experts are presenting their latest
38:33findings. He runs into one of his
38:35employees who knew that Jensen wasn't
38:37scheduled to speak and then asked him,
38:39"What are you doing at the conference?"
38:40Jensen replied, "I'm here to learn."
38:42Jensen had not assigned someone else to
38:44attend and take notes on his behalf. He
38:47had shown up himself so he could absorb
38:49the recent developments in artificial
38:51intelligence. He wanted to be deeply
38:53involved. There's a great story. So
38:54Mitch Rails is the co-founder of
38:56Danahir. Danahir is this like $150
38:58billion company. And a friend of mine
39:00was at this like small intimate
39:01conference with Mitch recently and he
39:04was blown away by how focused Mitch was.
39:06He says that the entire time that Mitch
39:08Rails was there. He was listening to the
39:10speakers and he was taking pages of
39:12notes. And my friend had this thought
39:14which he relayed to me where he's like,
39:15"Mitch, you're more successful than
39:17anybody else here. Why are you taking
39:18notes?" And when my friend told me that
39:20story, I go, "No, no, no, wrong answer.
39:23There's a great line in Bill Walsh's
39:24book, The Score Takes Care of Itself,
39:26that I think is very important and
39:27related to this." And Bill Walsh wrote,
39:30"Champions behave like champions before
39:32they're champions." I guarantee you that
39:34Mitch didn't build $150 billion company
39:36and then start taking notes. I guarantee
39:38you that Mitch was doing this way before
39:41he was quote unquote the most successful
39:43person in the room. There's a million
39:45examples of Jensen just being an
39:47absolute learning machine in the book
39:49and consuming in enormous amounts of
39:51information that will help his company.
39:53You can distill down to that maximum
39:54that the good ones know more. And what I
39:57loved about reading this book is you'll
39:58realize that Jensen is in the details.
40:01He's paying attention to things that
40:03surprise everyone else around him. Let
40:05me give you an example from the book.
40:07Peter Young was first introduced to
40:09Jensen at a party for new hires. Jensen
40:11already knew who he was. You're Peter
40:14Young, Jensen said. You've been here for
40:15about a year before you were at Sony
40:17PlayStation and 3D FX prior to that. He
40:20had a similar recall of biographical
40:23details for all 50 attendees of the
40:26party. There is a great quote from Walt
40:29Disney himself that says, "If we lose
40:31the details, we lose it all." Jensen is
40:34in the details. Jensen manages his
40:36company directly involving himself
40:38deeply in product decisions, sales
40:40negotiations, investor relations, and
40:43more. Number 16, we need to create the
40:45market, not fight over share of an
40:48existing market. Jensen believed that
40:50Nvidia was nothing like any other
40:52company that had ever existed. Jensen
40:54doesn't believe in building commodity
40:56products because commodities are subject
40:58to downward pricing pressure as
41:00competition increases. Jensen has always
41:03said that we should be doing things that
41:05other people cannot. Jensen believes we
41:08need to bring unique values to the
41:09marketplace and he feels that by doing
41:11work that is cutting edge and
41:13revolutionary, it allows the company to
41:15attract the best people. We do not have
41:18a culture of just going after market
41:20share. We would rather create the
41:22market. And there's many examples of him
41:24doing this in the book. In that
41:25commencement address at Caltech, he
41:27mentioned their move into robotics and
41:28he explained the decision-m behind it.
41:30He says, "We decided to build something
41:32where we are sure there are no
41:33customers." And everybody laughed and he
41:36was serious. He's like, "This is why
41:37it's important. Where there are no
41:39customers, there are also no
41:41competitors. That is a 0 billion dollar
41:44market.
41:46something that barely exists, maybe
41:48doesn't exist at all, but one day will
41:50be giant. And I think Edwin Lan is a
41:54good comparison here. You know, I'm
41:56obsessed with Edwin Land. He's one of my
41:57personal heroes. He was Steve Jobs
41:59personal hero. And if you think about
42:00it, this is exactly the strategy Edwin
42:02Lan had. When Edwin Lan targeted the
42:05instant photography market, there were
42:07no customers and therefore no
42:08competitors because he invented the
42:10market. And by inventing the market,
42:12that allowed Edwin Lan to build a
42:14monopoly. And one advantage of Edmund
42:17Lan's approach and of Jensen's approach
42:19is pricing power. Here's an example from
42:21the book. Jensen understood that gamers
42:23who buy Nvidia's cards are willing to
42:25pay for performance. As long as they
42:27look at the screen and see something
42:28radically different, they're going to
42:30buy it. He made the same case to
42:32Nvidia's investors. He said that Nvidia
42:35would be a unique semiconductor company
42:37where products average selling prices
42:40rise. We are going to be the only guy
42:43where ASPs go up over time when everyone
42:46else's ASPs will go down. He said number
42:4917. I will choke you with gold. I found
42:53a story in one of the books I read on
42:54David Ogrevi where he talks about that
42:56the Medici family was trying to persuade
42:58this sculptor to move to Florence and to
43:01enter into their service. And one member
43:04of the Medici family is writing this
43:06sculptor a letter laying out why he
43:08should come to Florence. And he
43:10concludes this letter with this
43:12following line. Come, I will choke you
43:15with gold.
43:17Jensen wants to work with the best
43:19people. And to do so, Jensen is willing
43:21to choke you with gold. Jensen looks at
43:24his stock like his blood. He pours over
43:26the stock allocation reports. Managers
43:29can refer an employee for special
43:30consideration to senior executives.
43:33Jensen will then review this list of top
43:35contributors and give out special
43:37one-off grants that also vest over a
43:394-year period. When the grant is
43:41approved, the employee will receive an
43:43email. The subject line says, "Special
43:45grant authorizing the RSU grant in
43:48recognition of your extraordinary
43:49contributions with a clear description
43:51of the rationale behind the award."
43:54Jensen can also reach down into the
43:55organization at any time and award stock
43:57directly without waiting for an annual
43:59compensation review. This allows him to
44:03ensure that people who are doing great
44:05work feel appreciated in the moment. It
44:09is yet another sign of his interest in
44:10every aspect and level of his company.
44:14Number 18, work on your highest priority
44:15first. This is what Jensen says. I have
44:17a very clear priority list and I start
44:20from the highest priority work first.
44:22Before I even get to work, my day is
44:24already a success. I've already
44:26completed my most important work and can
44:29dedicate my day to helping others.
44:31Ruthlessly prioritize your time is what
44:33he's saying there. Focus on the most
44:35important activity at all times. Larry
44:37Ellison shares this belief that that you
44:39should only focus relentlessly on your
44:41most important goal. Here's a quote from
44:43a biography of Larry Ellison that I
44:44read. Ellison made no apologies for his
44:47quirks. If anything, he seemed amused at
44:49Jenny's concern about his time
44:51management. Jenny was his longtime
44:53assistant. Jenny and I approach things
44:55very differently, Ellison said. Jenny
44:57feels if there are a thousand things
44:58you're doing, you have to do all
45:00thousand of them. My view is different.
45:03My view is that there are only a handful
45:04of things that are really important and
45:06you should devote all your time to those
45:08and forget everything else. If you try
45:11to do all thousand things, answer all
45:13thousand phone calls, you will dilute
45:15your efforts in those areas that are
45:18really essential. Jensen would tell
45:21others to focus on the most important
45:23activity at all times and not be
45:25beholden to a schedule. And then related
45:27to 18 is finally number 19. Swarm your
45:30greatest opportunity. Obviously, a huge
45:32part of this book is about the fact that
45:34Jensen recognized the AI trend early,
45:36invested heavily, and maneuvered his
45:38company into the position to take
45:39advantage of it better than anyone else
45:41in the world. And there's probably 80
45:43pages on this swarming of his greatest
45:46opportunity, this march toward this
45:48unlimited AI opportunity. And so I just
45:51pulled out a few examples, a few
45:52decisions he made that put him in a
45:54position so he could swarm the greatest
45:57opportunity of his life. And I think one
45:59of the most important lessons is this
46:01plays out over two decades. And he's
46:03doing this when people tell him not to.
46:05He does this at a great expense. And he
46:07does this over a long period of time.
46:09So, the idea that turns Nvidia into a
46:11multi-trillion dollar company actually
46:13came from their customers coming up with
46:15unique ways to use their product. The
46:17first reference to using GPUs for
46:19non-graphical purposes was in a 2002
46:22research paper on using computers to
46:24simulate the movement of clouds. And he
46:27picks up on this really early. An
46:28increasing number of computer science
46:30researchers were using GPUs for
46:31non-graphics applications. They reported
46:34significant speed improvements over
46:36computers that relied on CPU power only.
46:40In other words, the researchers had
46:42hacked their GPUs. Jensen sees this as a
46:46way to expand the market for his
46:48products and he works on making it
46:51easier for this to happen. And so number
46:52one, he hires the guy who discovered
46:54this. And then number two, he makes it a
46:56priority to make it easier for people
46:58like him to find more non-graphical
47:01applications for GPUs. That is a
47:03principle and idea that you and I have
47:04seen over and over again in these books
47:06that anytime you make something easier
47:08for people to do, the market expands.
47:10Anytime you make something easier for
47:12people to do, the market expands. The
47:14guy he hires, his last name is Harris,
47:16also coined the phrase GPGPU, general
47:19purpose computing on GPUs to describe
47:21this phenomenon that he was witnessing.
47:24Jensen was quick to grasp that GPGPU had
47:28the potential to open up the market for
47:31GPUs far beyond mere computer graphics.
47:36And so in order to generate more demand,
47:38Jensen decides they have to make it so
47:41their cards are easier to program. This
47:43is when they create the new programming
47:45model called CUDA, which stands for
47:46compute unified device architecture.
47:49CUDA made it possible and easier for
47:51scientists and engineers to leverage the
47:53GPU's computing power. Why did Jensen
47:56want to do this? Jensen believed that it
47:58would expand Nvidia's reach into every
48:00corner of the tech industry. Remember
48:03what Jensen said earlier? all the things
48:05that I value most about my abilities.
48:06Intelligence is not at the top of the
48:08list. One of the things that he values
48:10most is the ability to work on something
48:11for a very very long period of time.
48:14That is evident in this book 2002 the
48:17first reference to using GPUs for
48:19non-graphical purposes. What I'm about
48:21to read to you is happening in 2007 and
48:242008. This is again about the importance
48:26of doing something for a long time. So
48:28much of entrepreneurial culture now is
48:30start, scale, sell. And if you want to
48:33be great, you build, retain, and own.
48:36You find your last company. The company
48:39you can work on for decades, just like
48:40Jensen did. Jensen's decision here to go
48:43all in on this will create trillions of
48:45value for the company in the coming
48:48decades. Imagine if he quit or sold
48:50here. So they call this the era of the
48:52GPU. The era of the GPU would create so
48:54many opportunities that that Jensen saw
48:56it as his mission to prepare Nvidia to
48:59take advantage of it. Even if no one
49:01could know exactly what those
49:02opportunities would be, everything else
49:05was secondary. He is swarming his
49:08greatest opportunity. And it is Jensen's
49:11way. He wants to push it everywhere. He
49:13wants to make it a foundational
49:14technology. And if Jensen has
49:16conviction, he will invest heavily over
49:18the long term. Jensen understood the
49:21importance of saturating the market with
49:22it. The more people who had CUDA in
49:24their hands, the faster technology would
49:25establish itself as a standard. we
49:27should push this everywhere and make it
49:29a foundational technology. This move was
49:32extremely expensive. Remember this
49:34happened in 2007208. The company
49:35invested so much in converting its GPUs
49:37for CUDA compatibility that its gross
49:39margin fell from 45 to 35%. At the same
49:42time it's increasing its spending on
49:44CUDA, the global financial crisis
49:46destroyed consumer demand and Nvidia's
49:49stock fell by more than 80%. Jensen
49:52believed so strongly in the market
49:53potential that he remained committed to
49:55the course he had chosen even as his
49:56investors demanded a strategic course
49:59correction. Swarm your greatest
50:01opportunity. I believe in CUDA Jensen
50:04said we are convinced that accelerated
50:05computing would solve problems that
50:07normal computers couldn't. We had to
50:08make that sacrifice. I had a deep belief
50:11in its potential. So what does he do?
50:14There's another idea embedded in this
50:15idea. Educate the market. This is also
50:18not a new idea. If developers didn't yet
50:21know what to do with CUDA, Nvidia would
50:23teach them. Nvidia's chief scientist
50:25began offering schools CUDA capable
50:27machines if they committed to teaching a
50:28class on the subject. He gave more than
50:30100 talks over the course of a year. He
50:32taught a class himself at the University
50:34of Illinois. There was no textbook, so
50:36he wrote one. The textbook sold tens of
50:38thousands of copies, which translated
50:40into several languages and was used by
50:42hundreds of schools. It was a major
50:43inflection point in attracting attention
50:45and talent to CUDA. What they understood
50:47is that we need to educate the market,
50:50not a new idea. Intel did the exact same
50:52thing when they invented the
50:53microprocessor. Intel discovered that
50:55the mark marketplace, this is from a
50:57company history of Intel that I read for
50:59the first time 9 years ago. Intel
51:01discovered that the marketplace wasn't
51:03just confused by the concept of the
51:04microprocessor. The market had to be
51:06educated. At one point, Intel was
51:08conducting more seminar seminars and
51:11workshops on how to use the
51:12microprocessor than the local junior
51:14college total catalog of courses. Bob
51:18noise, Gordon Moore, and Andy Grove
51:20became part of a traveling educational
51:23road show. Everyone who could walk and
51:26talk became educators. It worked. Jensen
51:30was focused on how much of an
51:32opportunity GPU powered deep learning
51:35would be for Nvidia. There was
51:36considerable debate within his executive
51:38team on the topic. Several of Jensen's
51:40key lieutenants were against investing
51:42more in deep learning in the belief that
51:44it was a passing fad. Jensen overruled
51:47them. Deep learning is going to be
51:49really big. We should go all in on it.
51:53Swarm your greatest opportunity.
51:55Everything that you and I have been
51:57talking about today and highlighting and
51:59studying has led up to this point right
52:01now.
52:03Jensen had built a company in the image
52:05of his own focused yet far-ranging mind.
52:08Now he would pull every lever at his
52:10disposal to navigate Invidia to the very
52:13center of the tech industry as the
52:15company whose hardware could bring about
52:17the AI powered future. It was definitely
52:19not a single day when the entire company
52:22changed forever. It was a period where
52:24Jensen was increasingly interested and
52:26started asking increasingly deep
52:28questions and then started encouraging
52:30the company to swarm to machine
52:33learning. Jensen announced the change in
52:35strategic focus in a company all hands
52:38meeting. We need to consider this work
52:40our highest priority. AI was going to be
52:43more important than anything else they
52:45could possibly be doing. Within a
52:48decade, Jensen was sure that AI would
52:50create quote the largest total
52:52adjustable market expansion of software
52:54and hardware that we've seen in decades.
52:58It is hard to describe Jensen's actions
52:59as anything other than the construction
53:01of a competitive moat. Interesting to
53:02know that Jensen doesn't like the term
53:04moat. He prefers strong self-reinforcing
53:07network. My guess is because believing
53:10you have a moat can lead to complacency
53:12and complacency kills. Nvidia made a
53:14generalpurpose GPU that represented the
53:16first major leap forward in
53:18computational acceleration since the
53:19invention of the CPU. The GPU's
53:21programmable layer, CUDA, was easy to
53:24use and opened up a wide range of
53:26functions across scientific, technical,
53:28and industrial sectors. As more people
53:30learn CUDA, the demand for GPUs
53:33increased.
53:35Jensen's strategic brilliance ensured
53:38that competitors would have difficulty
53:40breaking into a market that Nvidia had
53:43created. Swarm your greatest
53:46opportunity.
53:48And I started this episode with the
53:50first sentence in the book that in
53:51another life Jensen might have been a
53:53teacher. Here is the last paragraph in
53:56the book. There are no shortcuts. The
53:58best way to be successful is to take the
54:00more difficult route. And the best
54:02teacher of all is adversity. It is why
54:04Jensen still keeps going at a pace that
54:07would see most other people burn out. It
54:09is why Jensen still says to this day and
54:12without any trace of hesitation or irony
54:15or selfdoubt.
54:17I love Nvidia
54:20and hopefully some of these ideas help
54:22you build something you love too. I have
54:25made something new that I love and I'm
54:26already obsessed with. I started a new
54:28podcast called David Center where I have
54:29conversations with extreme winners in
54:32business. The first two episodes are
54:34already out. My conversation with Daniel
54:35Ek, the founder of Spotify, and my
54:37conversation with Michael Dell. Both
54:38Daniel and Michael are big fans of
54:40Founders. And so when I reached out to
54:41both of them, they immediately agreed
54:43when I asked them for help to launch the
54:45new show. And so I'm asking you for
54:46help. I need a favor from you. Make sure
54:48you go follow the new show. It is on a
54:50separate feed. So whatever app you're
54:52listening to this right now, Spotify,
54:53YouTube, whatever it is, search for
54:54David Center and follow the show there,
54:57please. Thank you very much and I'll
54:59talk to you again

About this transcript

This page contains the full transcript of How Jensen Works by Founders Podcast, generated from the public captions YouTube serves with the video. The transcript has 10,369 words across 1,555 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.