00:00welcome to the a 16z podcast I'm Michael
00:02Copeland and this is sort of our holiday
00:05special we're going to talk about food
00:06and technology and to help us do this we
00:09have three people who are perfectly
00:11suited on the line are on skype I should
00:14say is UC Santa Barbara professor
00:16Chandra Krantz and with us also here at
00:19granular in San Francisco is the CEO of
00:22granular Sid Gorham and Tino Scola she
00:25was the co-founder and CTO for Wishon
00:30Chandra welcome thank you thanks for
00:33having us so we have a thesis at the
00:36firm that software is eating the world
00:38and this is where puns could start
00:40coming fast and thick which people try
00:42not to make but in agriculture I want to
00:44talk about how technology is making
00:46inroads and and each of you are working
00:49toward that in in sort of overlapping
00:52but different ways i'm shawna we can
00:54start with you you've got this concept
00:55of smart farm but what where has a
01:00Gregorio been or where is it at present
01:02in terms of you know it's adoption of
01:04technology what kinds of things and then
01:06what are you trying to get moving faster
01:10than it might otherwise sure that's
01:13great great question so we believe that
01:16agriculture is tremendously underserved
01:18by technology it's been left behind and
01:22both software and hardware although
01:24hardware has done a little bit better
01:25than software so we think it's a prime
01:28area that we can contribute to both on
01:31the research side and working with
01:33startups and industry on the industry
01:36side we see that they're being
01:40increasingly open minds to trying
01:43technologies in order to make farms and
01:46the farm processes more efficient and so
01:51there's been a lot of movement in a
01:53device realm and the hardware realm
01:55where sensors very low cost sensors have
01:58been developed they are connected via
02:01radio and cellular networks on the farms
02:04to get to collect information to collect
02:06data turns out that these sensors
02:08however are proprietary and they all do
02:12a little bit something
02:13different and have different interfaces
02:14and so it's really hard to go between
02:18fenders if you wanted to try a lots of
02:20different sensors on your farm whether
02:21ones that run on your tractor or that go
02:23in the ground and tell you how much how
02:25much water your soil needs or how your
02:28plants are doing and so part of our
02:31initiative at UCSB from a research
02:35perspective and an open-source community
02:37buildings perspective is to try to
02:39design systems develop systems that are
02:42freely available and open source that
02:44start to bring data together from
02:46different vendors aggregate it put it
02:48into a form that growers can control and
02:52access it and that interoperates with
02:54all the great technologies that are
02:56coming out of the public cloud in a more
02:59of a hybrid cloud fashion so you can do
03:01some work on the farm some work in the
03:03cloud and get the benefits of both local
03:06and more cloud-based technologies Tim oh
03:09you you guys have a lot of sensing
03:11equipment both hardware but then you
03:13know more importantly you your extract
03:14all this data from vineyards how does
03:17your technology in your hardware and
03:21software sort of fit into this picture
03:22that Chandra is describing so there is
03:25an opportunity from running analytics on
03:27whatever information is on the cloud the
03:31issue is what parameters is the most
03:35relevant to assess the specific plant
03:37needs and plant needs up front for
03:40different crops so there is a lot of
03:42work that then needs to be done once you
03:45have a uniform eyes access to all those
03:49crop parameters or land parameters or
03:52other elemental parameters to run the
03:54analysis and that's where you need to
03:56enter with plant intelligence to then
03:59help the farmer who has so many other
04:02things to deal with getting fast to the
04:05best best decision Shauna you were
04:08talking about how how agriculture hasn't
04:11seen kind of the benefit of some
04:12technology that we've seen in other
04:13industries let's say finance for example
04:15granulars is is kind of a workflow for
04:19farmers and you know we all think of
04:21farms and we think of people on tractors
04:24walking down rows it's very kind of you
04:26know pristine and pastoral but can it be
04:30both that and also sort of up to date
04:33and and utilizing the best of technology
04:35yeah but I think what's interesting in
04:38this transformation is that a lot of
04:39people assume that farmers are just kind
04:42of waking up to the modern world and
04:44saying huh maybe I can use technology
04:46and I actually think they're they're
04:48early adopters of technologies and many
04:50ways have kind of a hacker mindset
04:52they're very remote businesses
04:54self-reliant businesses they're juggling
04:57a lot of different you know complexities
04:59and I think really what's happened is
05:01that cloud and mobile have have made it
05:04possible to use technology on the farm
05:06and the farmers are willing and ready to
05:08pick up whatever tool will help their
05:10business right so it isn't so much a
05:12story of sleepy businesses waking up to
05:17technology as in my mind technology or
05:19the right technology arriving at the
05:22farm right well and so it's got to work
05:24and in the kind of evidence is fairly
05:28clear right yep yep I mean I think many
05:32different things have worked both on the
05:34hardware side and on the analytics and
05:36software side and I think granulars
05:38particular focus is on making farms
05:40better run businesses we think it's
05:43critical that farms understand where
05:45their profit comes from what resources
05:48generate a return for them so that they
05:50can be more careful about using those
05:53resources and make sure one of the
05:55phrases we like to use is we want them
05:57focused on their dollar yield not just
05:59on their bushel or you know pound or ton
06:02yield because really you want farms to
06:05be very very careful about how they use
06:08resources and make sure they're only
06:09using resources that that generate
06:11profit for the business right so all the
06:14inputs and the outputs at the end in
06:15terms of what whatever profit that's
06:17right - so you guys have all I think
06:20sort of touched on the resources tchibo
06:23and Chandra how do sensors and this data
06:28help utilize resources better - but you
06:31were talking about the right plant for
06:32the right job sort of yeah but how does
06:35that get that kind of precision and
06:38insite helped change what farms do and
06:42how they act and at the end of the day
06:43what we eat or drink so it is true that
06:46there is an explosion of new sensing
06:48technologies and some are providing data
06:51that are just nice to have but not
06:54necessarily leading to actions some are
06:57mission critical so if I take the
07:00example of the water flow inside a plant
07:02it's a little bit like the blood flow
07:04inside a human body by diagnosing the
07:07the water flow you can sense when the
07:10plant is actually in need of water and
07:13if the need gets to a severely limiting
07:17the plant will underperform but you can
07:20also see if the plant is not in need of
07:22water but still underperforms you can
07:25then point your finger but other
07:26resources that are limiting and that in
07:29the past were misdiagnosed for water
07:32deficit so it exacerbates a little bit
07:36your ability to tease apart what the
07:39plant really needs and what the farmer
07:40really need to adjust in order to
07:43maximize the the crop performance so
07:46that's just one example by looking at
07:48the potential benefits from looking at a
07:52vital index such as the the water flow
07:55genre what are you saying then along
07:58yes we see a lot of opportunities and
08:02excitement around using the analytics
08:06like tebow has been talking about to
08:09actually control a lot of the operations
08:12on the farm like the irrigation
08:14scheduling can be done automatically
08:16based on historical and current inputs
08:20people referred to them as parameters
08:22it's exactly right we see a precision
08:25application of fertilizers and
08:27pesticides you can even do this as with
08:30robots so completely eliminate the need
08:32for manual intervention there and that
08:36allows you to direct the pesticides or
08:39the fertilizers which are harsh on the
08:40environment to exactly where the problem
08:43is instead of more globally across a
08:46field and that prevents you know runoff
08:48and all the other side effects that we
08:50have downstream from these technologies
08:55it saves farmers money if you're not
08:57they're not going to apply tons of
08:59pesticides and fertilizers and I also
09:01see it a lot and harvest if you can
09:04control and guide irrigation and
09:05fertilization very precisely you can
09:08insure and control and guide when the
09:12plants are producing when they and make
09:15it more consistent if they all produce
09:17at one time say for almonds you only
09:20need to do one shake of the trees to
09:22harvest them that lowers farm costs that
09:26Sid was talking about optimizing the
09:28business side so the it's really an
09:30end-to-end solution that can be applied
09:33in lots of different ways no and so just
09:36just to be clear because I've never been
09:38to an almond farm but I like the one
09:39shake is because like it's that precise
09:41moment where all the almonds are ready
09:44to go and ready to drop and so you just
09:48you part of the farm has less water than
09:52another part certain trees will be ready
09:55for harvest before the others and that
09:57caught that's costly because you'd have
09:59to perform two harvests what you want to
10:02do is manipulate both the fertilization
10:05and the watering process so that all the
10:08trees are ready for harvest at the same
10:10time so you put out one team you get it
10:12done in a day and you don't have to go
10:15back and do it again said from from your
10:19perspective win-win farms whether it's
10:22small-scale large-scale sort of adopt
10:24these kinds of technologies to run more
10:26efficiently as businesses how do they
10:29change well I think one of the
10:31interesting things is most farms in
10:33America are family-run businesses even
10:35the very biggest farms are typically
10:37family-run businesses and as such and be
10:41and because nobody as often people spend
10:44their entire careers in farming so they
10:46don't know how other businesses run they
10:49they run with a very informal management
10:51style and so I think one of the sort of
10:54side effects of using products like
10:55granular or other software packages on
10:58the production side or on the business
11:00side it sort of pushes them to be more
11:04more organized and more a little bit
11:06more regimented in how they think about
11:09the business I think that's true in
11:10every industry with often what happens
11:12when you adopt software is it formalizes
11:15your process and makes here a little bit
11:17you know better run business you know
11:19above and beyond the impact of the
11:21software and we see that in farming for
11:24Tibo you guys started in the wine
11:26industry is that because it's just such
11:28a high value like why were they willing
11:30to do it and it you know said you said
11:32that and I agree that farmers are these
11:34hackers you can fix a tractor and fix it
11:36you know this and they run off the grid
11:38if they have to etc but why the wine
11:41industry and how because there's this
11:43weird kind of binary maybe weird it's
11:46nothing right worried about like part of
11:47the wine industry is like plant the
11:49skull of the cow under a full moon and
11:53your wine will grapes will be great and
11:56the other part maybe it's just much more
11:57you know technology and data-driven
12:00so first is because clearly the wine is
12:03a high cash crop right value and then
12:06there is another thing is that
12:08historically vineyards were always left
12:12to grow in areas where we couldn't grow
12:14anything else so the plant itself is a
12:17program to perform in otherwise very
12:22challenging and tough conditions so it
12:25prefigures a little bit what could
12:27become the future of agriculture as land
12:30aridity and global warming keeps making
12:33the land a bit more tough on the plant
12:36so for those two reasons high cash crop
12:39and also because the nature of the plant
12:41itself makes it very suitable to to push
12:44a little bit the boundaries of where we
12:47can still grow it's a very good
12:49beachhead to start developing this
12:51approach that the plant is very good at
12:54resisting drought the plant is very the
12:56divine plant is very good at performing
13:00even when there is not so much
13:02fertilizer or or nutrient availability
13:05so it helps you understand a little bit
13:08to which extent we can still make it
13:11profitable to farm a passel of land
13:15while still and hansung the production
13:18of a very profitable crop and so the
13:22market is good to test new technologies
13:25for that since you're saying that you
13:26know farmers are actually pretty eager
13:29to go on this but the wine industry
13:32seems fairly old-school in some ways I
13:36mean did they embrace this immediately
13:38or do they need I guess my question is
13:40do you need does everyone need evidence
13:43that this stuff works before they're
13:44willing to give it a try
13:45it's highly varied there are families
13:49and businesses that are very eager to
13:53improve their business processes and and
13:56make more money because of efficiencies
13:58that come with technology there are some
14:00that are highly resistant because they
14:02because exactly like you said that they
14:04need evidence they it's a crop lifecycle
14:07is very long it's not like in software
14:09and SAS where we produce something and
14:12and it's out there tomorrow you know
14:15this is we we do something and it takes
14:18a year you know and it's a huge risk and
14:21their lives depend on it so there's
14:23they're very resistant to risk in many
14:26cases and so you have to have evidence
14:28which means experimental farms and
14:30partnerships with universities and
14:32trying things and and showing and
14:34proving to farmers that this stuff is
14:36actually going to make a difference in
14:38I have a really funny story though I was
14:40trying to help a farmer actually a grape
14:43grower identified leeks using a
14:46combination of electricity sensors and
14:48water sensors on the irrigation system
14:51and the nets and leeks with an A and
14:53that leeks with two E's right okay yes
14:57and they said and and the farmer said
14:59well I know where there's when there's a
15:01leak I just look for the birds and so
15:03that you know they have all of these
15:05wonderful mechanisms because if the
15:07birds are around that means that they're
15:08you know they're thirsty they go find
15:10the water underneath the ground and then
15:12they find the leak so they have all
15:14these mechanisms to solve the problems
15:16already they are geniuses at this and so
15:19we have to come in and be and and
15:22provide them with the evidence they need
15:23to take on something new T Bo how how
15:27open were the arms of the why
15:29industry that sort of get this done but
15:31paradoxically it's not so much the fact
15:33that they cope with farming practices
15:36that are a little bit more obscure or
15:38hard to understand like you refer to
15:41biodynamic it's more of a societal and
15:44psychological predisposition so as a
15:47matter of fact for us it's much easier
15:49to develop our ideas in California which
15:52has a culture of a young society ready
15:55to embrace new technologies compared to
15:58the old world like France or Italy or
16:01Portugal so there is this societal
16:04footprint and then the fact that you
16:07have always been farming with strong
16:11respectful tradition is not at all in
16:13opposition with adoption of new
16:16technologies in fact the main rival here
16:18is how critical it is for me to not mess
16:22up the next crop as Sondra was saying
16:24you only have one chance per year to
16:26make a profit and so if you have two or
16:28three years of bad crop performances in
16:31a row you know you the next year is the
16:34last one you have on on the line and so
16:36if you don't maximize your chances to
16:38make it happen well you're screwed so
16:40that is also these are money incentives
16:44regardless of the psychological or
16:46societal predisposition and and like you
16:49know in the software world you can a/b
16:50test but you can't a B test a vineyard
16:52or an almond tree right said what are
16:56people can how can they sort of turn the
16:58knobs you know from from your
17:00perspective on I'm kind of managing this
17:03live or die situation yeah I mean it is
17:06a tricky technology adoption challenge
17:09because I mean just like in any other
17:11industry all of us are creating a
17:13category right you're not replacing some
17:16on-premise vineyard software with your
17:19new thing you're you're literally
17:21bringing the first software an analytics
17:23package they've used and we're doing the
17:25same thing on corn farms in Iowa so I
17:28think you have all the challenges that
17:29come along with that but then you have
17:31this dynamic we just talked about which
17:32is you get one chance or 40 chances in
17:35your whole career to produce some crop
17:37and you you're not going to just take a
17:42with a shiny software package who said
17:44this can help I can try in my business I
17:47can try Salesforce and if it doesn't
17:49work I'll turn it off a month later and
17:50yes no harm no foul but if I kill a crop
17:54or you know allow you to make some
17:57really big business mistake when you
17:59only have 40 cycles in your whole career
18:01to make money yet that's a really big
18:03deal so it's I mean not unlike medicine
18:06or education or some other things where
18:09there's really good reasons for people
18:11being extra careful about adopting
18:14technology what can we learn then from I
18:17mean you raise a really good point that
18:19there's a lot at stake here for farmers
18:21when they're when they're picking a
18:22technology and they're deciding to get
18:24behind it yeah it's their livelihood but
18:27like you say healthcare education other
18:30places that that need to be careful in
18:32their adoption of technology and then
18:33sometimes we complain that they're oh my
18:35god they're so slow right but yet they
18:38need to be careful what can we learn
18:40from the agricultural adoption of
18:41technology or what you guys have learned
18:43in terms of like building things that
18:45could help other areas and other parts
18:48of other industries and and adopt
18:51technology well I guess it doesn't
18:54necessarily work for our product because
18:56we're a kind of whole farm business
18:58management product but I think what the
19:00classic way of adopting technology and
19:02farming is I'm not gonna put this on the
19:04whole farm I'll try it on ten of my
19:08thousand acres right and we'll we'll
19:10sort of a bee test that way that's been
19:12the traditional way that farmers have
19:14adopted new seeds or new you know
19:17fertilizers or new management techniques
19:19so there is sort of a natural you know
19:23take a small part of the farming and try
19:25it out there approach that I think is is
19:28pretty established I also think that
19:31working with the universities and the
19:33extension systems that they've put into
19:35place is also how adoption happens for
19:38in the act in the AG sector so there are
19:41experimental farms that have similar
19:43crops and similar soil types and that
19:46that we can do these experiments on show
19:49that they work and farmers can then take
19:51you know how to take confidence in that
19:55so they'll be interesting to see if
19:57there would be these sort of forums for
19:59experimentation in in these other areas
20:02that we could bring from the AG sector
20:04since extensions been so successful for
20:07so many decades right that's yeah that's
20:10I mean you can't overstate how important
20:13that's been to the development of US
20:14agriculture so we are like the runaway
20:17leaders in in agriculture and in many
20:20ways some because of the extension
20:23process or the extension infrastructure
20:25and I think you know we need to keep
20:29that going and I think it's under a lot
20:30of pressure from a funding perspective
20:32and then some of the latest technology
20:35is sort of outside of the the experience
20:39base of some of the faculty at the
20:41extensions and so there's I think we're
20:42at a really important crossroads where
20:44the extension Proc you know university
20:46system needs to evolve and keep pace and
20:49support firming for the next hundred
20:50years we have to as a society invest in
20:53that and that will make a huge
20:55difference for Ag and for the just the
20:57technology industry in the AG sector I
20:59see two main points that are going to
21:04benefit from adopting AK tech number one
21:07in the context of global warming the
21:10more data we learn and acquire and
21:12analyze today the better prepared we
21:14will be when temperature will rise with
21:17it we only have half a degree Celsius on
21:19average and the second point if I again
21:23used the example of a vineyard we see a
21:25lot of mismatch between one specific
21:29value at all and the environment where
21:31this value at all it is grown in some
21:34instances there are places where I can
21:37blend gently points that it makes no
21:41sense to grow one specific value at all
21:44or sometimes just grow period right so
21:48but it's only by using analytics and and
21:51an raw number that we'll be able to
21:53select the places and situations where
21:57it makes sense from either an ecological
22:00societal or business standpoint to grow
22:03and that for the future will be very
22:05important as we can no longer
22:09take any chance with land use given the
22:13growing population that's right
22:15and and I think you're exactly right
22:16that it's both a science problem and a
22:18business problem if you look at
22:20Australia that's a very hostile farming
22:24environment from a water perspective and
22:27they have a fully developed trading
22:30system for water so that if farmer a
22:32doesn't use all his water he can sell it
22:35to farmer B through this market and and
22:39I think too for that kind of system to
22:41happen in America you need farmers to
22:43really understand their economics and be
22:45able to say a gallon of water is worth
22:47exactly this much to me and I would
22:49rather leave my field fallow and sell it
22:52yeah exactly crosses that price and
22:55right now you can't number one water is
22:57generally free but but I think the only
23:01way agriculture evolves with the
23:02pressure of global warming is through
23:05those types of market-based solutions
23:07and that that requires farmers know that
23:10Shondra you mentioned the farming for
23:13the next hundred years and I want you
23:14guys to sort of imagine what its gonna
23:15start to look like our drones going to
23:17be out picking you know peaches in the
23:19in the trees and and our is it going to
23:22be self-driving tillers and you know
23:25harvesters how do you imagine things and
23:30you know in your smart farm genre it
23:33almost sounds like the Internet of
23:34Things might hit the hit the farm before
23:37it hits our homes but when you spin this
23:40forward what does it start to look like
23:41I think the biggest change is that
23:44instead of thinking about some of the
23:47things we've been talking about here
23:48water resources amount of land soil
23:51quality the resources that we'll all be
23:53talking about that are going to be the
23:55most valuable to the growers themselves
23:57is their own data we're going to have to
23:59develop the markets and the controls and
24:02the the ways of sharing data that
24:06benefit farmers because in my opinion
24:07data is their most valuable resource and
24:11I think the future is all about that
24:14because there's so much intelligence and
24:16insights that can be drawn from that
24:19locally and globally and so that's going
24:21to be a huge game-changer in the future
24:24and secondly I do believe that there'll
24:27be a lot of automation it will be
24:30hand-in-hand with you know the domain
24:33experts that farmers themselves but they
24:35will be probably requiring much less
24:37manpower because drones and robots and
24:40ground robots and you can take pictures
24:43can do all kinds of interesting things
24:45can can pick and harvest and fertilize
24:48and they can do it round-the-clock in
24:50the rain in the dark you know they are
24:52not limited to what humans are limited
24:55to doing and so I think that's going to
24:58change and then aren't then and then the
25:01crop life cycle is probably going to
25:02change with the changes in global
25:04warming and so different plants are
25:06going to be need need to be grown in
25:08different areas and and you know sharing
25:10that knowledge and intelligence is going
25:12to be vital so I think it's a real
25:14combination of human intellect and and
25:16drone and robotic technology and
25:19analytics and data management that are
25:21gonna change everything
25:22Seana can I just ask you does does I
25:25mean one of the things we hear about and
25:27said you talk about how many of the
25:29farms or most of the farms the United
25:30States are our family farms by sort of
25:33unit then there's massive industrial
25:36scale farms but what technology does a
25:38really good job of is lowering the cost
25:40of things does what you talked about
25:43change the scale of farming or does it
25:45enable you know small scale farms to
25:48even do a much much better job I think
25:50it's just that I think it's the ladder I
25:52think it's smaller farms are now
25:55hopefully economically buckin can be
25:58economically more viable now because
26:01they can be more efficient technology
26:04brings that but the super scale of you
26:07know the field corn and the soybeans you
26:10know it that scale is still an important
26:12factor to have many many acres to
26:15succeed at so so it's really going to be
26:17a balance between those two not to
26:19mention feed the world's population
26:21that's right that's right yeah I mean I
26:24think what I see going forward in
26:26addition to all the things we've talked
26:27about is just the continued
26:29professionalization of the industry
26:31right and and the continued consol
26:33nation of it I mean I think you can't
26:34avoid may used to be that a hundred
26:37percent of the country farmed and now
26:39two percent of the country farms and
26:41it's gonna be one percent before we know
26:43it right it's it is an industry in which
26:46people are moving out of these family
26:48businesses to go do other things so I
26:50think as a society we need to accept
26:52that farms are going to get bigger and
26:55that doesn't necessarily mean they can't
26:56be great stewards of the land and do all
26:59kinds of good things for society in fact
27:01I think the more the the thing we as a
27:03society should care about its
27:05professional management of farms and
27:07whether that's a small-scale farm or a
27:09big farm it doesn't really matter we
27:11just want these resources to be well
27:13looked after and to get the healthiest
27:15you know product into the system so I
27:18think I think to some extent we get a
27:21little distracted by we love big firm or
27:23we love small farms but don't like big
27:25farms and really what we should be
27:26focused on is it's the talent and
27:28management of the business and the use
27:30of technology and I see that you know
27:33it's just going to continue to
27:34accelerate the witness by the fact we're
27:37sitting in South of Market talking about
27:43why do that are gonna be more and more
27:46critical I think it's easy to understand
27:48if you look at medicine for instance
27:51every one understands the value of
27:53monitoring biological indexes over as
27:55many people as possible to better
27:57diagnose trends and statistics and
28:00predict when you are at risk of heart
28:02disease for instance well in farming I
28:04feel because we haven't really value the
28:07meaning of such data it's a little bit
28:09like medicine in in the 17th century
28:11well when you have a problem you gonna
28:14get a bleeding and hopefully is gonna
28:16cut off and cut it off and really in in
28:19the future it's only by pulling together
28:22data that we're going to increase our
28:25ability to jagneauxs the needs of plant
28:28as if they were individual but your
28:31ceil about farming you need getting ever
28:35a greater size is also going to raise
28:39the problem of variability where where
28:41do we get data from and how do we
28:43extrapolate the data to jagneauxs
28:46the best way to to to farm according to
28:50this intrinsic viability that can be
28:52seen at the plant level itself or the
28:55the unit of production itself and and
28:58it's it's it's posing a big problem on
29:00how we're gonna extract the value of
29:02these data so it's not only a matter of
29:04the meaning of the data the data in
29:06itself has a lot of meaning for future
29:07use but how do we get to the data and
29:10how do we make sure that this data is
29:11meaningful which depends upon how do we
29:14deal with viability that's right again I
29:17think there's a lot that we can learn
29:18from agriculture as these questions
29:21start to get answered and as technology
29:22seeps more and more into the fields and
29:25into the vineyards and you guys are at
29:28the front of this and it'll be
29:31fascinating to see how it plays out you
29:32know on our plates and in our economy so
29:35I just want to thank you all Chandra
29:38tebow and Sid we will see you on the
29:41farmer and the vineyard thank you thank
29:43you very much thanks guys