00:00hi and welcome to the a 16z podcast in
00:02another of our hallway conversation
00:04episodes Benedict Evans and Steven
00:06Sinofsky talk about Tesla and more
00:07broadly the nature of disruption overall
00:09how disruptive is Tesla really what
00:12exactly are they disrupting from the
00:13dashboard to car makers to vendors to
00:16energy source to autonomy overall and
00:18how much does each innovation matter
00:20looking back at other examples
00:21historically what kind of disruption
00:23matters most for market dominance good
00:27morning I'm Steven Sinofsky I'm
00:28pentateuch Evans what we thought we
00:30would talk about today is a little bit
00:32about the nature of disruption and in
00:34particular about about Tesla and the
00:37rise of electric vehicles and sort of
00:40how disruptive is Tesla you know we're
00:43in an interesting time like certainly
00:45economically like if you look at the you
00:48know the main US car companies you know
00:49Ford Fiat Chrysler GM all of them are
00:52worth less than Tesla and market cap
00:55individually and certainly if you look
00:57at the past year Tesla has been worth
00:59twice as much as them at some point so
01:02crazy interesting world but you know we
01:05use disruption a lot in Silicon Valley
01:07and and in technology in particular and
01:11what's interesting about the looking at
01:13Tesla is you know it it may or may not
01:16really be disruption in the way that the
01:18book is written which is sort of oh my
01:20god there's a whole new thing and then
01:22the old thing just goes away and it's
01:24all new players yeah I thought it was
01:27interesting to look at how people think
01:29about Tesla here because you sort of on
01:31the one hand you said you have the
01:33narrative oh my god they're doing this
01:35stuff that the car companies can't do on
01:37the other you have the narrative oh my
01:39god they're making the cars in a tent
01:41and bits fall off when you drive down
01:42the highway ha ha ha it'll never work
01:44and this reminded me a lot of sort of
01:47hearing similar conversations around for
01:49example the iPhone this is a terrible
01:51phone well I'd touch really easily no
01:54one will ever buy a phone for that much
01:56money sorry exactly there's on the one
01:59hand people dismiss people dismiss on
02:01but in both directions so they dismiss
02:04the new thing because he can't do what
02:05the old stuff does very well and don't
02:07realize that you might be able to learn
02:09that but on the other hand you also
02:10dismiss a difficulty of the old stuff
02:13dismiss what the barriers to entry might
02:15actually be and I saw I thought as I
02:17looked at Tesla I kind of wanted to pull
02:20apart well what are the different things
02:22that are happening here
02:23are they disruptive are they not are
02:25there barriers to entry which bits are
02:27there barriers to entry and barriers to
02:29entry to who was one of the kind of the
02:32historical comparisons I use webcourse
02:34Stephen has sort of scars on his back
02:36around this is that if you look at for
02:38example what Apple did in the PC
02:40industry Apple contributed to creating
02:41the PC we had one of the first popular
02:43PCs that Apple did not win pcs so you
02:47can all talk about whether Apple
02:48disrupted IBM but Apple didn't actually
02:50get the benefit from that not in an in
02:52fact PC companies didn't really get the
02:54benefit from that PC companies became
02:55low margin commodity companies and the
02:57people who got the benefit will
02:58Microsoft an Intel yeah well that's
03:00what's so interesting about that that
03:02disruption is sort of you know when can
03:05companies turn some technical innovation
03:07into there into a competitive advantage
03:09and when is the company's go-to market
03:11or technical innovation itself become
03:14like a hindrance to adoption and there
03:16our tech industry is littered with with
03:18examples of the innovator failing to
03:20become the Dominator so to speak my
03:22favorite one is just is replay TV you
03:25know gave us the DVR and then gave us
03:27TiVo and now we all just have DVRs
03:30everywhere certainly you know the
03:33mainframe and the mini computer led to
03:35the PC and the revenue numbers all came
03:38and dwarfed it and then you know what
03:40you know so well on the phone industry
03:41is very similar yeah so there is a
03:43question of is this new thing
03:44fundamentally difficult for the
03:46incumbents to do but also is it
03:49important so well it was sort of this
03:50sort of four things that I talked about
03:52in the in the blog post so the first is
03:54the Tesla kind of has to learn the old
03:56stuff Tesla halite has to learn how to
03:58make cars at scale and there was a
03:59period when people said oh my god
04:01they're reinventing manufacturing
04:02actually no they just bought a
04:03secondhand robot Factory yes well yes it
04:05has to work out how to do cars this is
04:06but we have to like we it's almost worth
04:08applause there just to remind people the
04:11other people's jobs are really much
04:14harder than you think they are and we
04:16tend to like even in the software
04:18hardware world i've yet to ever meet a
04:20company that makes hardware that think
04:22software is like really really hard and
04:26hardware company or a software company
04:28that thinks they can't just go to
04:29hardware and buy it in China and it this
04:32notion that like to innovate in cars you
04:36know you you need to understand like
04:38cars more and manufacturing and to
04:40innovate in the software in cars you
04:42need to understand software more if
04:44you're living in Detroit well this is
04:45the thing that people in software don't
04:47really understand enough about cars and
04:48people in cars don't really know enough
04:49about software but so does it to the so
04:51to the point on tests so it clearly
04:53there's this whole conversation now
04:54about the production hell and they're
04:56making cars intense and the panel gaps
04:58are terrible and you know they're havin
04:59fires in the paint paint plant the paint
05:01line and so on so Tesla has to learn how
05:04to do what Detroit already knows how to
05:05do and watch a pan already knows how to
05:07do which is to make cars reliably
05:08inefficient is Germany Japan okay well
05:11to make colors reliably well Detroit
05:13doesn't make ours anymore but make
05:14colors reliably and efficiently at scale
05:16that's just a condition of entry
05:18Tesla gets through production hell that
05:20doesn't get them victory that just gets
05:22them to continue to that just keeps him
05:24in the game well they also have to do is
05:26be doing something that the existing car
05:28industry's can't do or will struggle to
05:30do for kind of deep structural reasons
05:33that they won't just be able to hire
05:34engineers and just add that and that
05:37stuff has to be in some way
05:38fundamentally important it has to be
05:40like a profound reason why you would buy
05:41a car and it also has to be something
05:44that other tech companies will struggle
05:45to do which is kind of which is to the
05:48the Apple versus Microsoft point or the
05:51Apple versus Dell point so Apple did
05:53stuff that IBM for the sake of argument
05:55found it hard to do but Dell did it
05:57better in partnership with Microsoft and
05:59Intel equally HTC were the first people
06:02to make Android smartphones or the first
06:04people to make Android smartphones but
06:05HTC tend not to have the right
06:07positioning in the marketplace to take
06:09all the rewards from that and so you
06:10kind of have to kind of you can kind of
06:12look at like the beautiful product and
06:14you have to kind of unpick okay how are
06:17they going to make millions of them how
06:19what is it in that that is difficult as
06:21opposed to easy for other people to do
06:23which of those things are fundamentally
06:25important and which of those things also
06:27will like not just BMW you can't just
06:31say well BMW isn't going to make be able
06:32to make software you also have to say
06:34well BMW isn't going to be able to buy
06:35those from some combination of far away
06:39Google in order to get you to okay Tesla
06:43is going to be the new BMW or the new GM
06:44right that that's a sort of a very
06:46important point like in the sense of
06:49instead of looking at this as disruption
06:51another way to look at this is to use an
06:53old phrase that existed before a
06:55disruption and just refer to it as a
06:56secular shift and that this is a shift
06:58there's just a shift we're gonna all be
07:00in electric cars and electric vehicles
07:02and electric transportation at some
07:04point which is very different than it's
07:07disrupted because disruptive tends to
07:08focus on the micro like one company
07:10versus another company whereas if
07:12everything is going to move to this it's
07:14not clear that it just means that only
07:16the companies that are currently doing
07:18something are gonna benefit and you're
07:20going back to the DVR example it turns
07:22out DVRs are like a commodity now like
07:24everything that can receive video has
07:26the capability of just being a DVR so we
07:29should probably kind of dig through kind
07:31of what those separate components are
07:33yeah I mean the analogy that I use I
07:35mean I thought what I tried to do is to
07:37break it apart so there is the electric
07:40itself which is the battery and the
07:42motors and the power train and the
07:44controller software for that not exactly
07:46the most revolutionary technically yeah
07:48lithium-ion batteries are not something
07:49that got invented by by Tesla and
07:51Panasonic five years ago there's that
07:54then there is the sort of one level up
07:57all the integration of the control
07:58systems around the car and then there is
08:01a dashboard on the car and the
08:03experience the broader experience of
08:05buying at a car like do you go to a
08:07they're challenging stations everywhere
08:09do you have lots of fiddly little
08:10buttons or just one beautiful
08:11touchscreen and we kind of look at those
08:15and think well what are the how is how
08:18are the dynamics of each of those going
08:19to play out and how hard are they for
08:20new people to play in and if we kind of
08:22start with electric the analogy I
08:25thought was kind of interesting here was
08:27to look at multi-touch so Apple was
08:31actually not the first company to sell a
08:32mobile phone with a capacity of
08:34multi-touch screen I think there was an
08:35LG one a year earlier and maybe a couple
08:37of others but Apple was the company that
08:39said oh my god we can actually use this
08:40to totally change what it is to be a
08:41thing well it's actually an important
08:43point is that they they weren't first at
08:45the using technology but they were the
08:47first to integrate the technology and
08:48pull it all together
08:49yeah which which you know as we just
08:53as we go through and discuss each of
08:54these like it's worth saying that I'm
08:56fairly optimistic on on the prospects of
08:59being able to do solve this equation and
09:01others are gonna be fairly pessimistic
09:02and this is really about just analyzing
09:05that conversation not sort of debating
09:07the winner yeah exactly so you have this
09:11fundamentally you have this insight into
09:13a new piece of technology okay we could
09:14use this to make a phone
09:16lívia my own batteries are going to get
09:18cheap enough that you could use them to
09:19make a car this is like the foundation I
09:21live in inside of Tesla but if you go
09:24into a store today there are a thousand
09:26phones with capacitive multi-touch
09:27screens and so clearly just using a
09:30multi-touch screen of itself didn't get
09:32you anything because everyone could buy
09:33those black even black we were selling
09:35phones with capacitive multi-touch
09:37screens and so the within that will you
09:40split that out on the one hand the
09:42legacy companies so Nokia blackberry
09:44palm struggled to make a phone with a
09:46good capacity of multi-touch screen on
09:48the other hand in partnership with
09:50Google Samsung a lot of other people
09:52found it real easy to make friends with
09:53capacitive multi-touch screens and so
09:56today the entire industry makes these
09:57things this is also just like all of a
10:00sudden everybody adds a notch to their
10:01phone yeah like something that appears
10:03like whoa that's gonna be super tricky
10:05all of a sudden a supply chain appears
10:08other people with expertise appear and
10:10you have a lot of innovators sort of
10:13building the same yeah the same thing
10:14this is the thing if you look at what
10:16the P the way the PC industry works the
10:18way the mobile phone industry works
10:19indeed the way the car industry works
10:20it's not that there is one company that
10:23has to work out how to make this thing
10:25it's not that boss is going to have to
10:27learn electric it's so you have a whole
10:28ecosystem of hundreds of kind of
10:30different companies hundreds of very big
10:31companies full of good engineers who
10:33have to work out how to make this thing
10:34many of them have been making batteries
10:36in electric motors for a long time
10:38already just not quite the same kind and
10:40so as you look at electric it's seems
10:44pretty clear to me that only like a five
10:47or ten year view and bear in mind cars
10:49are on a five to ten year replacement
10:50cycle so it doesn't have to happen that
10:52quickly there will the entire car supply
10:54chain will have reoriented around
10:56electric and even more than that the
10:59entire electronics industry that already
11:01does electric staff will reorient around
11:03making components for electric cars
11:06so if you look at the kind of the tear
11:07downs of say Chevy Volt male wolf alot
11:11of the value in that comes from people
11:12that were not traditionally car
11:14manufacturers car component suppliers
11:15they're not the traditional tier one is
11:17all LG right but in fact what's super
11:19interesting about that too is that that
11:21the expertise at existing car companies
11:24is in acquiring those technologies
11:26building them out establishing those
11:28relationships negotiating the contracts
11:30and getting all of that to happen
11:32there's not like four even GM that makes
11:34the bolt on the Volt it's not like
11:36there's this massive lithium ion group
11:38at the company yeah and so what you get
11:41to there is you you sort of think okayed
11:44the car companies are going to be able
11:46to go out and buy these components just
11:48the way they buy their existing
11:49components and there's not some
11:51fundamental intellectual property here
11:53there's also no disruption story it's
11:55not like they're sitting thinking oh
11:56this is a terrible idea and we don't
11:58understand this or you know it's it's an
12:00integration into their existing
12:01manufacturing process on the other hand
12:04if you're a German company that makes
12:06gearboxes for the car industry you're
12:09not going to be able to switch to making
12:10this your modern batteries it's a
12:12totally different business and so your
12:14gearbox business is either going to
12:15disappear or you're going to shift to
12:17marine engines right and if you're an
12:18auto supply store on the corner you know
12:20and these cars don't need parts anymore
12:22like yeah that's like a thing to go
12:24short right now if you're in the
12:25business of speculating about timelines
12:27and things like yeah if you're in the
12:28business of making radiators for cars
12:30that business is going to go away and
12:32you're probably not going to replace
12:34that with a business making electric
12:36batteries or power control systems so
12:38that will go that's not even disruption
12:40that's just your whole industry just
12:41that's the secular shift exactly it's
12:43like all of a sudden horses are now
12:45centered around different set of
12:47technologies and different places that
12:48you use them exactly but that's a
12:50different layer in the stack to the
12:52common factors I mean the analogy I used
12:55in my blog post was that for the sake of
12:57argument the Internet was radically
12:59disruptive to travel age and it's been
13:00not disruptive at all to ellen companies
13:02airlines still sell tickets they sell
13:04them differently through different
13:05people but they still run planes and
13:06that IKEA hasn't changed their business
13:08for cars what an interesting view of
13:09this is the way that that that the
13:11change and the focus on safety permeated
13:13the car industry there was an era in the
13:1660s when nobody worked on safety and
13:19manufacture particularly like Volvo
13:21picked up on safety and then the Germans
13:22picked up on safety partially due to
13:24regulations in Europe and things like
13:25that moving faster and you know oh my
13:27god the American companies are not going
13:29to have anti-lock brakes they're not
13:30gonna be able to have airbags and I can
13:32have all these things and it turns out
13:33like now there's dozens of companies
13:35that contribute to that supply chain and
13:37it's just part of every car you can't
13:39even differentiate on safety anymore
13:42because they're able to build that up
13:43and that's a car version of multi-touch
13:45yeah I mean I think they say there's an
13:47interesting kind of question in this
13:50which is there is the new thing as
13:53opposed to all their sort of fundamental
13:55structural reasons while you're going to
13:56struggle to digit struggle to adjust to
13:58this and so again if you look at for
14:01example what happened with phones Nokia
14:05had a was totally oriented around what
14:10the handset what the mobile operators
14:12wanted they were totally oriented around
14:16optimization of component cost around
14:19having a huge supply of building blocks
14:22that they could use to make 50 or 60
14:23different phones every budget models
14:25like it's mind-boggling how many models
14:28they had exactly and so their whole
14:30structure was around was deeply
14:34challenged by what the iPhone proposed
14:36because the iPhone proposes they k1
14:38phone totally different components
14:41presume it lasts a day instead of two
14:43weeks presume it doesn't matter if you
14:45drop's instead of it if it drops when
14:46you break in presume it doesn't care at
14:48all about bandwidth consumption or
14:50memory presume you're basically
14:52indifferent at the component cost
14:53because you're selling of $600 instead
14:55one hundred and fifty dollars and there
14:56are people at Nokia said you will never
14:57be able to cellophane from all that
14:58happens if people Microsoft would say
15:00that yeah exactly and so this is as
15:02though this is more like the shift from
15:05ocean liners to aircraft you know it's
15:08not it's it's the difference would be on
15:11the one hand the shift from ocean liners
15:12to aircraft on the other hand the shift
15:14from propellers to jets
15:15and a shift from propellers to jets is
15:18basically all the same companies shift
15:20from ocean liners to aircraft it's not
15:21the same company yes although Hugh not
15:23actually bought an airliner in the 50
15:24ellen company in the 50s because they
15:26could thought maybe that was what was
15:27they could bit of course it didn't work
15:28well and also that's like for cars to
15:31jets and rolls-royce still being a
15:33manufacturer yeah exactly so you have
15:35that question of is this some
15:37fundamental thing that they don't know
15:39how to do or is it they just haven't
15:41done it yet and I think this is like a
15:43kind of a crucial misunderstanding
15:45people make which is they've done it
15:47first no one else is doing it today okay
15:51why is that is that because there's some
15:53fundamental structural reason they can't
15:55do it or is it because batteries are
15:57still $200 per kilowatt hour and that's
15:59not cost competitive with gasoline which
16:00incidentally is also big reason why
16:02Tesla obviously is still great money and
16:03they're waiting for the battery prices
16:05to come down and then you're gonna do it
16:06more slowly and and this is where it
16:08becomes very sort of in a sense
16:10emotional about disruption and like
16:12either you're just like a very big bear
16:14on how car companies behave and that
16:17there are entrenched bureaucracies and
16:18it is important to put car companies in
16:20some broader context like these are
16:22hundred-year-old companies that have
16:24survived many different waves of
16:27Technology and many different changes
16:29and in all fairness to them they
16:31invented modern management like
16:33everything that's interesting about
16:35management sort of came out of him out
16:37of TM j managed em I mean just as an ad
16:40like everybody listening to this please
16:41go read my years of GM by Alfred Sloan
16:44it's just an amazing amazing look
16:46because you're gonna recognize many
16:48things in that book that companies do
16:50today about how they manage brands how
16:52they deal with distributors distribution
16:54and networks and and manufacturing and
16:57all of that kind of stuff even if some
16:58of it appears dated to you because
17:00there's like labor unions and and things
17:01like that so the interesting kind of
17:04break point when one looks at the cause
17:06is you kind of go up a level from the
17:08electric and so I mean they're kind of a
17:11great kind of vignette of this is Tesla
17:14discovered the model so you had a
17:15problem with the brake they push down
17:17over-the-air firmware update that fixes
17:19the problem with the brakes and so if
17:21you look at like the way that like a
17:23conventional car you would buy today is
17:25put together there are dozens and dozens
17:27of separate subsystems in there all of
17:29which come through from from kind of
17:30separate vendors so the ABS is a system
17:33the backup camera is a system the
17:35airbags or a system and they will come
17:38from separate vendors they're all
17:39integrated as we were saying earlier by
17:42the comma fracture and they want all of
17:44those systems to be commodity so they
17:46on them and the only place if they if
17:48they have a user interface obviously
17:49selling them happening user interface if
17:50they have a user interface that
17:52manifests is a button on the dashboard
17:53so this old joke that you can see the
17:55org chart of a car company in the
17:56dashboard and you can like see that the
17:57HVAC people hate the steering wheel
17:58people or something yeah and so what you
18:00have is like an org structure that's set
18:02up to deal with these parts as
18:04components and not integrate them at all
18:05and you want them to be not integrated
18:07because then you can just swap out Bosh
18:09for Lucas and it doesn't matter and you
18:12then look at the way tester have built
18:14their car and it's one central computer
18:16running an operating system as opposed
18:18to a real-time operating system running
18:20on a real operating system on some Linux
18:22pork or something and the way that this
18:25has been described as you go from
18:26basically complex cars with very simple
18:27software to actually very simple
18:29hardware but with complex software so
18:31you have a computer controlling the car
18:32yeah which is also serve an analogy of
18:34what happens with with with feature
18:36phones because there's a camera and
18:37there's the phone app and there's the
18:40SMS and they're not integrated except on
18:42the on the screen and then you go to a
18:44smart phone where suddenly you've got a
18:45piece of software that's controlling all
18:47of these well and that's very similar
18:48that you know the PC industry actually
18:50had this exact the reason that there no
18:51none of the PC makers other than Apple
18:54in a sense are successful phone makers
18:56is because they were exactly like
18:58Detroit they they had a graphics group
19:01they had a peripheral group they had an
19:03i/o group they had a store used to
19:04describe Dell as being like a very being
19:07a very specialized version of FedEx yeah
19:09oh yeah that they buy the parts and kind
19:10of put them together on the way to
19:12getting them to you or a logistics
19:13business as much as they are a
19:14technology business there's lots of cool
19:17technology inside Dell as well but
19:18basically there were an assembly
19:20business and which is also what Detroit
19:21was and what Apple isn't and if you look
19:23at the literally look at the headcount
19:25of those companies like the number of
19:27Mechanical Engineers relative to the
19:29number of supply chain managers
19:31procurement people and you know QA
19:33people it's sort of out of whack like
19:35relative to what you would see at Apple
19:37so this get this gets you to kind of an
19:39interesting point like we've kind of
19:40we're kind of setting aside electric it
19:43seemed pretty clear electric is a
19:44commodity over time you go up one level
19:47this stuff is stuff so it's a bit kind
19:49of institutionally harder for car
19:51companies to think about because they've
19:53got a whole org chart that says well
19:54I've got an ABS man and I've got a
19:56backup camera person and I've got a
19:59and no Chester doesn't have any backup
20:01by person backup Tesla has a software
20:03team yeah well and it it's and now
20:06speaking as a manager like this is a
20:08very very real thing like your your
20:09building your new electric car a big
20:11exact existing car company and it's
20:13gonna have abs brakes in it so you're
20:16gonna go to your brake expert yes like
20:18you're not gonna go to the software team
20:21and say make me some brakes and yes it's
20:23not the software teams Joe there's not
20:25the software team make a device driver
20:26for the brakes it's really the brake
20:27people give you the brakes right and so
20:29the brake people are the brake people
20:31and they're gonna look at this problem
20:32and they're gonna go okay first job go
20:34to Bosch and go get the brakes that I'm
20:37gonna use for this and in fact some of
20:39this actually manifests itself in my
20:40Chevy Volt because like it's very clear
20:43that they went like they for climate for
20:45the heater and the and the a/c they went
20:47to the existing heater & a/c people and
20:49said I need a heater in an AC because
20:52one of the things that's super weird is
20:53like it's not really integrated with the
20:56battery powertrain that's in the car and
20:58all and it's a basically an old-school
21:01kind of heater yeah and the same with
21:03the dashboard it's like like the Chevy
21:05bolt dashboard looks like a GM dashboard
21:08from all the other cars yes and you and
21:10you see this through through the whole
21:12experience and so that's an interesting
21:14well it's an interesting locus for
21:16disruption because it's easier it's a
21:18lot easier to argue that this is
21:20difficult for car companies to adjust to
21:22than it is for electric per se I think
21:25and you see that manifest in things like
21:27you know the Tesla software update and
21:29in to some extent Tesla's ability to add
21:32new capabilities or new features to the
21:33car kind of over there never mind
21:35autonomy which we may come to yeah yeah
21:37I'm do you later but yeah you can do
21:39this you can do that it can they can do
21:41this thing all that there that cool
21:42thing and you can just decide to add it
21:44as opposed to you know your conventional
21:46car which as I've got a BMW 3-series it
21:48will get new features when I buy a new
21:50BMW I will not get new features device
21:51well you don't expect them you don't
21:53want them and some but but that really
21:55does speak to it it is quite conceivable
21:57that that there's a bunch of stuff that
22:00Tesla as a company is going to do that
22:03is in fact very very difficult for car
22:07companies to do the question I think is
22:09how this kind of gets to one of the kind
22:12of the full things we talked about
22:13how much does that change the
22:15competitive advantage of the car and I
22:16think we had a conversation about this a
22:18few weeks ago where you were comparing
22:19this with laptops and phones yeah that
22:22does this produce it does Tesla's
22:24approach produce a better cord is it
22:26produce a different car and and this
22:29this notion of better and different like
22:32this is why I think too it's such an
22:33emotional debate for people when they
22:35sit in my Tesla versus sitting in in a
22:39Chevy Volts or sitting in a in a gas
22:41combustion car is like it it feet like
22:44the Tesla experiences it's a very
22:46different kind of car and you also you
22:49have you have to image a lot when people
22:51talk about the Tesla experience a lot of
22:53what they're actually saying is stuff
22:54that's generic to electrics they say oh
22:55my god have you felt the acceleration
22:56yeah you do understand in 10 years the
22:59crappiest GM car you can buy will have
23:00the same accelerant electric like I came
23:03from a Prius to my bolt and like the
23:05thing that I I'm like oh my god this is
23:07the fastest car I've ever owned and this
23:09is very very similar the this is one of
23:12the things that happened with the
23:13original tablet PC that we made at
23:15Microsoft which was in around 2000 all
23:19of a sudden like we did these internal
23:21surveys do you love your new tablet PC
23:23we had like got in like a hundred units
23:25and deployed them for a test and
23:26everybody was like this is the greatest
23:28PC I ever owned and like all of a sudden
23:31were like oh my god we're on to
23:32something it's really big and then we
23:34dug into the researcher loved it and we
23:35realized that well the thing is that
23:37these new tablet PCs that we had just
23:39made like these one-offs were actually
23:42made to be super super good pcs they
23:45actually weighed like three pounds and
23:47they were super thin and they had really
23:50great screens on them because of the the
23:52the pen screen was made really well and
23:54so we realized nobody was actually using
23:56ink at all they just loved the fact that
23:59it was compared to their seven and a
24:01half pound think pad there's a sampling
24:03problem but it was the smallest lightest
24:05laptop they'd ever used but I think the
24:07thing about the Lapp the thing about the
24:09sort of the test integration is you know
24:11the first point is we sort of a big part
24:14of the experience is the acceleration
24:16all cars will have that that's not Tesla
24:17that's just electric and all cars will
24:19have no maintenance they will have oil
24:20changes and say whenever they will be
24:22they'll just this is just electric this
24:24is not Tesla and in all fairness I think
24:27like end up with a similar like miles
24:30per kilowatt-hour kind of range because
24:32the physics is sort of everybody shares
24:34these physics and there's not all this
24:36leakage that you might experience with
24:38choices you make and horsepower and gas
24:40combustion engines and stuff there'll be
24:42a variance of course yet even if you
24:44look today like they're really pretty
24:46clustered around the same sets of
24:47measurements exactly so the electric
24:49stuff is a commodity the anyway you get
24:52to kind of the integration of these
24:53components you can argue well it's going
24:55to be a lot more difficult for legacy
24:57car companies to do they're structurally
24:58like what's like the org chart yeah they
25:00actually have actually shipping or chart
25:01in the car you have actual reasons why
25:03it's difficult for them to do this
25:05what's not quite so clear to me is
25:07whether that translates into reason why
25:09you would or wouldn't buy the device by
25:10why you wouldn't or wouldn't buy the car
25:12and I think the analogy with you I think
25:14we were you were talking about a couple
25:15of weeks ago was sort of the difference
25:17between an Apple laptop where they
25:19either there's no choice of any of the
25:20components and it's super super
25:22optimized and the case is made out of
25:24machined aluminium to fit each component
25:25and so that the laptop is really really
25:28thin and has a really good battery life
25:30and so on as opposed to a Dell laptop
25:31where you've got a choice of 45
25:33different components and you can swap
25:34and you can you have this or you can
25:35have that and that means there's more
25:36empty space inside the laptop because
25:38they've got room for the bits you didn't
25:39choose and it's got a fit for different
25:41components right right and it might have
25:4210% worth battery life because it's not
25:44super optimized the difference is okay
25:46it's also you have the choice of all the
25:47components and I think that sort of the
25:50Tesla versus GM conversation that easy
25:52it's super super optimized and hyper
25:54designed around one specific
25:56configuration or is it you know okay
25:58we're running making five different cars
25:59on this line and well well mission will
26:01mix and match and we'll get this and
26:03we'll give that I mean example I saw
26:05that Mercedes have just announced in a
26:07an SUV electric SUV and they're putting
26:10the the electric motors in the front
26:13under the hood instead of kind of down
26:16on the chassis level next to the axles
26:17which is what what Tesla was doing and
26:19if you're only making electric cars it's
26:21better to do it the way Tesla is doing
26:23of course Mercedes is making on this on
26:25a line which is also making like these
26:26three series and the seer a ISM and
26:29therefore it's more efficient if they
26:31actually have sort of some overlap in
26:32the mechanical processes there and you
26:35can argue well maybe they'll lose 5%
26:36battery life by doing it like that all
26:38the weight distribution won't be quite
26:39so good on the other hand they might
26:40save 10% on the cost of the man making
26:42the thing which means it's $5,000 Chivo
26:44this is this thing this is it's so
26:46important to really hammer this point
26:48home because this is this isn't a sense
26:50disruption but it's disruption at a at a
26:53very micro level within an organization
26:55but when we were building surface and
26:57armed pcs at Microsoft like one of the
27:00things that happened as we showed up
27:01when we said look when you use ARM chips
27:03the graphics card is like right next to
27:05the CPU and they're all part of the same
27:07thing you can't buy an ARM chip from one
27:10vendor and a graphics chip from another
27:12and mix-and-match them and like most of
27:14the people who traditionally make pcs
27:16they I was looking at across the table
27:18from the graphics person and from the
27:20CPU person and they know which is
27:23getting fired one of they thought
27:24literally one of them was gonna not get
27:26to do their job and and then it went
27:29another level where they're like well we
27:30we actually need to leave room in order
27:34to be able to swap out like a new CPU
27:36because if we get a different one in the
27:38middle of the production run we want to
27:40save all of that upfront engineering
27:41cost on the chassis and on the assembly
27:44line and we're like well the thing is
27:45they all just come soldered to a board
27:47at manufacturer time so there's no you
27:50can't switch them and the dimensions the
27:52CPU could just move around a whole bunch
27:53like the it'll change and they literally
27:56couldn't they just weren't interested in
27:58making it because like it they didn't
28:00know what their job would be if they
28:03weren't optimizing that particular thing
28:05and and more importantly they didn't
28:07know what their job would be if they
28:09couldn't change around the parts because
28:11their whole economics of what they were
28:13building was based on optimizing the
28:17inbound supply chain for switching
28:19different things plus the tech
28:21enthusiasts side of it the purchasers
28:23who were like we need to have a bunch of
28:26graphics on this this device so we're
28:28gonna up the graphics level or we need
28:30this but device to have longer battery
28:32life so we're gonna lower that and the
28:33marketing people who want to have like
28:34good better best for every single PC
28:37like they couldn't imagine just having
28:39like good better best be defined like
28:42Apple does by amount of storage or a
28:43screen size like the the the Mac
28:46all-in-one desktop was always just small
28:49medium and large and and like they were
28:52basically the same except for the screen
28:54which turns out to be very consumer
28:56oriented way and what what's gonna
28:59happen with cars is gonna be varying
29:00because it's not just that offering it's
29:03the whole purchase process supply
29:05process advertising for every dealer
29:07compensation process all of these things
29:10there's a whole kind of interesting
29:11question around what electric does to
29:13the car industry which is that if you
29:14you know you can have you can imagine
29:15like a $15,000 car that does nought to
29:1860 in three seconds and so all of the
29:21questions and is super safe like by
29:23default and like maintenance that said
29:26of course a Porsche still dry well
29:28electric Porsche will drive an awful lot
29:29better than all that trick Tesla because
29:31just because it goes fast in a straight
29:32line there's more to being a good car
29:33right right right but it does remove
29:36like layers of like you buy the bigger
29:38engine I mean again I've got like a I
29:39have like a seven or eight year old BMW
29:41that I bought secondhand and it's I
29:42don't even know where the engine is but
29:44you look at the badges on the back and
29:45it says it 328 or the 330 or the 335 or
29:49the 335i and I there will be that will
29:51not mean anything there will be there
29:52will be one gearbox and it will be you
29:54know those those differentiations within
29:56the car will go away you know there will
29:58be not be different gearboxes so how do
30:00you so one last thing we have to talk
30:02about though is the the really the big
30:03one for in sense you know the sense of
30:05is software eating transportation is you
30:08know the very long-term vision of like
30:10where autonomy fits into all of this so
30:12this is a thing I mean if we kind of we
30:14kind of get back to our four layers so
30:15there was the electric there is the kind
30:17of integrations which we've just kind of
30:19been kind of musing about there is the
30:21driver experience and then there is the
30:24you know the driver and the dealers and
30:25they really are updates and the
30:27on-screen dashboard and then there is
30:28your time as part and if you kind of go
30:30through those the electrics the
30:31commodity the integration stuff is a
30:33bunch of interesting internal questions
30:35in the supply chain of the car
30:36manufacturers but it's not terribly
30:37clear that translates into a different
30:38car or car with strong competitive
30:41advantage there's a dashboard experience
30:43and then as the autonomous part and
30:46there's a dashboard and the dealers and
30:47everything else and then the I find
30:49experience the car experience so I kind
30:52of talk briefly on the dashboard then
30:54kind of talk about autonomy I think I
30:55think it's much the easiest place to
30:57locate or disruption is in the dashboard
30:59because all the things we've been
31:01talking about all about the org chart is
31:02really hard for a traditional car
31:04company to say we're not going to have
31:05any manual controls in the car except
31:08well you tried it with the 7-series yeah
31:10like 20 years ago we're only good but
31:12this is like literally not like there
31:13was they were the they were the sticks
31:15on the on the steering wheel and then
31:17there's a scream yep and I think there's
31:18a bunch of reasons why it's really hard
31:20for legacy car companies to do that the
31:23question is is this like iPhone hard or
31:27is this when you buy an iPhone it
31:29activates with AT&T over over the area
31:32right that's a channel hard or is is it
31:34like physics or yeah an easy does it
31:37make a fundamental difference to
31:38people's willingness to buy the car I
31:39mean I did a totally unscientific
31:41Twitter poll and my question was if BMW
31:44and Tesla and BMW sadie's tests are all
31:46selling a car with exactly the same
31:47track same drivetrain I the same
31:51acceleration the same electric
31:52everything is exactly the same it the
31:54only difference is that you have the
31:57big-screen dashboard as opposed to the
31:58Mercedes or BMW dashboard how excited
32:00would we be about Tesla as a company and
32:02it's like well would this really be a 50
32:04billion dollar company well two things
32:07on that one is that of course in a you
32:09you have to factor into that kind of
32:11choice all of the like that the negative
32:15selling that will happen from it from
32:17car companies without that they will
32:18talk about safety driver distraction
32:20they will literally go to the government
32:22and try to get dashboards like that
32:24banned like this is exactly what Detroit
32:26has been doing for decades over electric
32:29yep the phrase range anxiety was not
32:32dreamed up by the physicists at GM it
32:34was dreamed up by the marketing people
32:36selling against electric vehicles yeah
32:39and then the other half of that is just
32:41gonna be like the fans of gas combustion
32:46engines and the fans of existing
32:48companies like well if company XYZ that
32:50I love doesn't have an all-in-one - that
32:52means that they're bad yep and but it
32:55doesn't the thing about it is is that
32:56these lining up brands like this this is
32:58again go back and read my ears of GM
33:01because Detroit mastered the art of
33:04selling the same thing to different
33:07people with different prices with
33:08slightly different with slightly
33:09slightly different things I mean we had
33:11a Camaro type LT and I remember that
33:14when I was little because it had this LT
33:16right on the door where I would open it
33:18and I always ask my mom what is Lt mean
33:19my mom had no idea she just knew it was
33:22a Camaro hmm which was not the same as a
33:24Firebird or is a Trans Am even though
33:27they all looked like very sick like Burt
33:29Reynolds sort of drove the same car yeah
33:31as we had but it didn't have an eagle on
33:33it I was like oh the eagle is really
33:35expensive I'm like for the sticker
33:37they gave him one right right well
33:39that's different thing so okay but the
33:42autonomy is the software player so it's
33:45not the dashboard I think what we're
33:47getting at is the dashboard is fine this
33:49is not the wealth of nations this is not
33:51the hundred billion dollar change right
33:53right it's the autonomy because electric
33:55is a commodity the - the the integration
33:57is a commodity the dashboard is fine
33:59it's autonomy is it's a question and
34:02here we well there's an autonomy
34:05question and as a disruption computation
34:06and the disruption conversation in this
34:08is Tesla is not who are the people who
34:11are competing here it's not Silicon
34:13Valley software company versus dumb
34:14Detroit guys ray its Silicon Valley
34:16software company versus 20 other Silicon
34:18Valley software companies plus Silicon
34:19Valley software company and in China
34:21this plus China plus Silicon Valley
34:23software company that got bought by the
34:25dumb Detroit guys sorry Detroit but you
34:27know you know that's what they said this
34:29believe me that's what they said when
34:30that happen yeah who the Detroit guys
34:32all the Silicon Valley cars either way
34:33yeah yeah so the question here is
34:36there's clearly this is a fundamental
34:39profound new technology you can we can
34:41have a whole other conversation about
34:42how long away well we've done a lot of
34:44other podcasts about when will autonomy
34:46correct right will happen as a result
34:47and there and the cities and towns will
34:49change the lies will change and
34:50everything changes exactly everything
34:52changes but who is trying to build this
34:54right now Google all the big Chinese
34:58tech companies cruise several dozen
35:01smaller companies trying to build
35:03component parts of this all of which at
35:05some point will be available for sale to
35:07anybody who gets any of this working
35:08plus like people who make components all
35:11have projects going on exactly if you
35:14you you are affiliated with projects yes
35:17so there's large are companies as
35:18mapping companies there's you know all
35:20radar companies there's all sorts of
35:22people there's a simulation call the
35:23delivery companies all the trials is
35:25people building all the sort of
35:26disaggregated people building all the
35:28components so you've got way mo building
35:30building their own camera systems
35:31building their own databases building
35:33your own mapping systems and their own
35:34simulation tools on the other hand
35:36voyage will buy the simulation tool from
35:38Applied intuition and thereby the
35:40mapping from deep map space but the
35:41point is there is a whole ecosystem
35:42that's trying to create autonomy and so
35:45within that ecosystem Tesla is one
35:47company trying to build this as well now
35:51and there's a conversation about where
35:52we think Tesla is positioned within
35:53their kind of the battle to build
35:54autonomy form a disruption conversation
35:57it's not they're all disruptive or
35:59they're all innovative companies
36:00although all new companies well they're
36:01all participating in this giant
36:03transformation exec a whole new scenario
36:05all at the same time exactly so there is
36:07no sort of the new people are doing
36:09something that the old people will want
36:11to do it's there's 50 new people all
36:13competing against each other and and and
36:15also depending on where they are these
36:18people are all coming out of the same
36:19universities studying the same kinds of
36:22machine learning and then they're all
36:24ending up these companies and they're
36:25not and they're all spinning off from
36:27those companies and they're all changing
36:28jobs like there's this is a little
36:30community of knowledge that's being
36:32built at this at one time so there's a
36:34there's another point in here which is
36:35sort of been implicit in several of the
36:37sort of previous things we've talked
36:39about which is you have the sense that
36:40there is an entire ecosystem so there is
36:43a whole ecosystem making gasoline car
36:45engine components and supply chain those
36:47whole ecosystem making part of the
36:49reason that Apple ran into such
36:50difficulty in the 80s and 90s was Apple
36:52was trying to compete with the entire
36:53ecosystem they weren't just trying to
36:55compete with Microsoft they'll also try
36:57to compete with all the people who were
36:59selling component and they weren't
37:00trying to compete with Dell they were
37:02trying to compete with Microsoft and
37:03Dell and the 300 Cubs and Intel and
37:05Intel and the 300 people had Seagate and
37:08everybody and everybody who sold stuff
37:10to Dell they were trying to out-compete
37:11the entire ecosystem so yeah you have to
37:14really understand what was going on not
37:17that that wasn't but to what what this
37:18means is that there was Apple had to
37:21make one decision for every component
37:23that made a Mac and that was the only
37:26one they could make so when they picked
37:27like a hard drive they had to pick the
37:30brand of hard drive bright all the
37:32firmware integrated into the operating
37:33system and do all that and then they
37:34were done and if like the industry went
37:37a different direction they would just
37:39get left behind yeah and if they and if
37:42they'll decided hey we've got a
37:43turn on that driver sis dis so if they
37:45could just stop shipping see great in
37:46the start shipping max turn and it's
37:48super important because of the maturity
37:49of the industry you you actually needed
37:52that flexibility then because you just
37:54didn't know where things were gonna go
37:55like famously Apple like dragged out
37:58like their Apple talk for a very very
38:00long time even though networking it all
38:02moved on into tcp/ip and and then and
38:05then they won't scuzzy and then they
38:07were on scuzzy and then they were on
38:09firewire and they were all that whole
38:11era they seemed almost like a generation
38:13behind if they were even behind it all
38:15like sometimes they picked firewire and
38:17it just never made it to the PC
38:19ecosystem yeah and I think I I hear they
38:21think about this when I see Tesla making
38:24their own this or making their own lat
38:26and you think okay set-asides the fact
38:27that you have a major cash flow problem
38:29and why are you spending money to do
38:30this rather than just buy it don't worry
38:32buddy nevermind the question why are you
38:34competing why do you do you want to
38:36compete with the entire ecosystem or
38:38should you be riding on top of that
38:39ecosystem and finding the unique thing
38:41that you alone can do and should you bet
38:44that you will be the only person doing X
38:46when there's a whole ecosystem that's
38:47trying to do X and also this is where it
38:51has a lot of parallels to the early Mac
38:53because it's not just that they have to
38:54do all that which is almost
38:56insurmountable the way that at least we
38:57described it but they have to do it at a
38:59fairly low volume yeah like so and the
39:02low volume in many of these things is
39:04what sort of makes it really really
39:06difficult because then you can't even
39:08get the attention of manufacturers to
39:10help you even if they're like sort of
39:12white labeled parts of it yeah so this
39:13is I've heard sort of gossip about this
39:15with the in the car supply chain that of
39:17course tests a lot a lot of the bits of
39:19a inside a Tesla were bought from the
39:21car supply chain most of the bits inside
39:22a Tesla come from that buy they're not
39:24making it he did not do a Henry Ford and
39:25go donate don't make you go and buy
39:27tires starting from you they're not
39:29making their own glass for the windows
39:30right normally they're not making their
39:33own motors - why the seats backwards and
39:34forwards and of course the problem is
39:35the volume is so low that they can't get
39:37the best deals for the best
39:38manufacturers which is sort of it it's
39:41part of that whole ecosystem question
39:42and so we kind of come back to the
39:43autonomy question again like with the
39:46electric piece so we sort of think about
39:49another way of putting this so the kind
39:50of the ball case here would be Tesla is
39:52competing with car companies are doing
39:54software like they'll win
39:56they're in autonomy they're competing
39:58with software companies at doing cars
39:59they'll win the bear cases nananana
40:02Tesla is competing in cars Tesla is
40:05competing with car companies at doing
40:06cars and they're competing with software
40:08companies are doing software and so that
40:11sort of gets you back to the kind of the
40:13autonomy kind of is it disruption is
40:15only one of the strands through this but
40:17like where are you what's your
40:19competitive positioning is it that
40:21you've done something no one else can do
40:22is it that you've got or are you just
40:25trying to compete with a whole industry
40:26at doing something that industry knows
40:28how to do and where do you want to kind
40:29of put yourself within that I mean I
40:31think this is part of the kind of the
40:32genius of the turn of the Tim Cook here
40:34at Apple is no we're not gonna make all
40:35the phones ourselves we're gonna make
40:36our phones the phones themselves ya know
40:38we're not gonna make the chips ourselves
40:39no we're not going to design this I
40:40thought we will pick a certain number of
40:42kind of key points of leverage and make
40:44those ourselves but we're not designing
40:46the own gyroscope these incredible
40:49lessons from the Mac era yeah and and
40:53they're putting them to work and this
40:55notion of of a learning company is
40:57what's incredibly important and that's
40:59one of the things I would like to raise
41:01as this as a great car company example
41:03which is we talked about my years at GM
41:05and the early days well also another
41:07famous GM experiment was the GM Saturn
41:10brand which was this experiment than the
41:1180s that where GM was looking at Japan
41:14and they were losing everywhere they
41:16couldn't make small cars they couldn't
41:18make fuel efficiency labor costs were
41:20too high the dealer experience was hard
41:22but you know back when in the 80s when
41:23you wanted to buy a Toyota you would
41:25just go in and they would say do you
41:26want a red one a black one or a white
41:27one and and if you went to go buy an
41:30American car you it was that all the
41:32stereotypes of the worst they would
41:33sweat you in a small room forever you
41:35would have to like option packages
41:37number t-43 or qr7 and you would have to
41:40figure out and they overlapped and you
41:42it was a horrible experience
41:43so what GM did is they did clay
41:45Christensen before the book would
41:47existed which is they started a whole
41:48brand they hired all different people
41:50they relaxed every constraint imaginable
41:53and they said go do it except to your
41:56point what they were trying to do was
41:57have this one badge of GM compete with
41:59all of Japan and it turns out like it's
42:03very very hard to do that and it ended
42:05up costing billions of dollars and they
42:06shut the whole thing down is a failure
42:08and books have been written it's another
42:11great book is about the history of
42:12Saturn and so like did they really did
42:15they fail because they they couldn't
42:17make all the changes or they couldn't
42:19recognize the changes or they
42:21misunderstood what was really going on
42:23hmm yes I mean I suppose you could argue
42:26that the the kind of the Saturn thesis
42:27would have been that will shut down all
42:29of TM and Saturn will become GM and at
42:31that point maybe it would work but as
42:33long as you're going to kind of continue
42:34running it as a separate thing well what
42:35about the rest of GM and and that's
42:37where like so much of this becomes very
42:40very interesting because ultimately like
42:42your framework for thing about Tesla
42:44just it it really raises so many very
42:47interesting questions and I think it all
42:48comes back to what our founders are
42:50always needing to make sure that they
42:51think about which is it's never just the
42:53product it's never just the price it's
42:56never just the way you you promote it
42:58and and use channel management and it's
43:00never just about the the pricing
43:03structure yeah I mean it you have to
43:05really consider all of these I'll say
43:06the thing I was thinking about this
43:09recently was like one of the sort of if
43:11I'd like the accumulated learning over
43:12my career is I'm always a sucker for a
43:15beautiful product right and the thing
43:16that I've learned over time is okay yes
43:18but what's the route to market yes but
43:21what's your differentiation yes but how
43:22are you getting the components yes but
43:24what's your sales process yes but yes
43:27but and and really are our final thought
43:30on this as I think is is it the
43:32beautiful product can really get you in
43:34front of a customer but it takes a lot
43:37of things to get you in front of all of
43:38the customers or you can get in front of
43:40all of the customers but it's gonna take
43:42a lot more to fully meet their needs in
43:44a differentiated way and get the price
43:46and the margins that you need and all of
43:48those things are really coming together
43:49and I think where we're seeing things
43:51now is that you have to start to
43:53consider all of those and not just
43:55anyone and that's what's so interesting
43:57about this right thank you thank you