Latest / Automated with Brian Heater / Rohit Sharma on Venture Capital, Robotics, and Long-Term Thinking

Transcript

Rohit Sharma on Venture Capital, Robotics, and Long-Term Thinking · Automated with Brian Heater

• 0:00
  SPEAKER_00: In the farmer scenarios, they're just exposed to this sort of nature global level climatic and other changes and have to respond to them.
• 0:09
  SPEAKER_00: It's like primal.
• 0:11
  SPEAKER_00: It is very primal.
• 0:12
  SPEAKER_00: For roboticists going into that domain, they need that visceral understanding of what those sets of challenges are.
• 0:19
  SPEAKER_00: Agtech is the double black diamond of robotics.
• 0:23
  SPEAKER_00: It's the hardest thing you can do.
• 0:30
  SPEAKER_01: Hey everyone, my name is Brian Heater.
• 0:32
  SPEAKER_01: I am the managing editor at A3.
• 0:35
  SPEAKER_01: We are back with another episode of Automated.
• 0:39
  SPEAKER_01: This week I'm reaching back into my TechCrunch days and doing a bit of VC talk.
• 0:45
  SPEAKER_01: We have Rohit Sharma of True Ventures on the show.
• 0:50
  SPEAKER_01: Spoken to him a number of times over the years.
• 0:52
  SPEAKER_01: Always a fascinating conversation.
• 0:54
  SPEAKER_01: He's very good at putting uh big ideas into large very large context.
• 1:01
  SPEAKER_01: I think you're going to enjoy this interview.
• 1:03
  SPEAKER_01: I did.
• 1:04
  SPEAKER_01: If you like the show, please like it in the way that you like things on YouTube or the platform of your choice.
• 1:11
  SPEAKER_01: Rate it, review it, all of those things as well.
• 1:14
  SPEAKER_01: And uh follow the newsletter over at automated.fm and I will see you on the other side.
• 1:21
  SPEAKER_01: I feel like no two people have the same journey into the world of VC.
• 1:29
  SPEAKER_01: Your yours is uh certainly pretty unique, but how did you end up where you're sitting right now?
• 1:35
  SPEAKER_00: There's no established paths to venture.
• 1:37
  SPEAKER_00: Um, maybe intentionally so sometimes I think, because it's it's it's a it's a torturous circuitous route for for most people.
• 1:44
  SPEAKER_00: Mine was literally falling backwards into entrepreneurship after my grad studies up in Canada uh in engineering, um, finishing my PhD, running into good people at a conference in San Jose who left me um who gave me an opportunity to come to the Bay Area here and work.
• 2:02
  SPEAKER_00: And within a few months of that, I had started a company, and that eventually led to sort of climbing up the stack, if you will, to venture.
• 2:09
  SPEAKER_00: Um, but my introduction to venture was through creating a startup at the point where I know knew nothing about startups, venture equity, blah, blah, blah.
• 2:17
  SPEAKER_00: I was just very lucky to be with great people.
• 2:20
  SPEAKER_00: Uh, and that really started uh me on the road that eventually led to venture.
• 2:25
  SPEAKER_00: Um, and I think I've been lucky in venture as well to end up with very good people.
• 2:28
  SPEAKER_00: The partnership we have here currently at True Ventures is just phenomenal at the people level and have been.
• 2:34
  SPEAKER_00: I I just don't get to call this a job ever, which I think is is the absolute best it could be.
• 2:40
  SPEAKER_01: Yeah, I always say uh given what I do for a living, there's always I mean, you know, it it's helpful if you're smart and know a lot of things, but there's always also power in not knowing certain things because when you don't know certain things, you ask the right questions.
• 2:54
  SPEAKER_01: And I'm wondering you coming into VC from this kind of back door, whether that was almost an asset.
• 3:02
  SPEAKER_00: I think it it it could be an asset.
• 3:04
  SPEAKER_00: Maybe it was an asset, because I think it let me be very skeptical about a lot of things the way I think a good engineer always is.
• 3:11
  SPEAKER_00: Um, a very, very skeptical optimist.
• 3:14
  SPEAKER_00: You're an optimist because that's what in the end you're trying to do.
• 3:17
  SPEAKER_00: You're trying to bring something new into the world.
• 3:19
  SPEAKER_00: And I think continuing that point of view into venture um certainly has has been the way that I have worked in venture.
• 3:27
  SPEAKER_00: I hope it's been good for the people who worked with me as founders to have that framing, to have that viewpoint rather than somebody who came up through the ranks on finance or working early in their career in a venture firm.
• 3:40
  SPEAKER_00: I certainly did nothing of that sort.
• 3:43
  SPEAKER_01: This is something that I've I've never thought to ask, and I've talked to a lot of VCs because um I guess it's never really occurred to me, but in a certain sense, do you have to be an optimist to be a VC?
• 3:54
  SPEAKER_00: I think it helps having an overall optimistic kind of anchoring to it, if you will, right?
• 4:01
  SPEAKER_00: Because in the end, you are trying to look longer term, you're trying to get on the side of constant change, right?
• 4:08
  SPEAKER_00: I don't I don't think venture as an industry works if there was not much change in the domains we are involved with.
• 4:14
  SPEAKER_00: And I think there's always an element of irrationality alongside that I think only it works if you have the optimist framing on top of it.
• 4:23
  SPEAKER_00: Because the rational answer to probably most startups at the point of financing is no, right?
• 4:28
  SPEAKER_00: Uh, it's not a project that should exist, it's not an effort we should finance, um the at least for the very good outcomes and and companies that end up changing the industries they go into.
• 4:39
  SPEAKER_00: So I think it helps having an optimist view all the time of what if this was to come true and what could happen on the other side, rather than coming up with 15 reasons why it won't work, which I think is always possible, by the way, right?
• 4:52
  SPEAKER_00: So I think you you need that that irrational optimist framing to some of this.
• 4:59
  SPEAKER_01: Yeah, I'm and and maybe this is the useful, one of the useful things about having an engineering background is you obviously you also have to be a pragmatist, right?
• 5:06
  SPEAKER_01: There's there's it you can't be too optimistic about investing in a company.
• 5:11
  SPEAKER_00: Pragmatist goes hand in hand.
• 5:12
  SPEAKER_00: And that's why I said very intentionally, this is a very uh my framing is a very skeptical optimist.
• 5:17
  SPEAKER_00: I I want to be very skeptical of why something is relevant and why something will work better than other efforts, uh, rooted in technology, rooted in technical thinking, rooted in engineering, perhaps my engineering education and its practice.
• 5:31
  SPEAKER_00: Uh, but then in in the end, you are trying to then look past those and say, okay, assuming you can get past those, what is the change you bring in here and how does this upend an existing industry or create a new one?
• 5:44
  SPEAKER_01: That's interesting.
• 5:44
  SPEAKER_01: So those two are those are prerequisites for anything you're looking at.
• 5:49
  SPEAKER_01: Is I I hate to say this because and I've certainly contributed to this in my career, but disruption?
• 5:54
  SPEAKER_00: I think disruption is a very loaded word, but certainly the new way of reimagining a market has to be a fundamental possibility.
• 6:02
  SPEAKER_00: Um, and also, frankly, that's our bias as very early stage investors.
• 6:06
  SPEAKER_00: If you want to write the very first one to three, three and a half million check a founder raises, often with a lot of PowerPoint at their disposal, not a lot of reality yet.
• 6:15
  SPEAKER_00: So I think if they go into fully formed markets, one, by definition, we and they are likely too late.
• 6:22
  SPEAKER_00: Um, because this is not, we're not contemplating a late stage$50,$100,$150 million check.
• 6:28
  SPEAKER_00: So at the early stage, we are biased towards filtering for things that would hopefully create new markets or change the market that they will be in, and they'll mature seven to ten years after a point of investment, which is an amazingly long time frame in technology.
• 6:43
  SPEAKER_01: That's something that I really wanted to talk to you about specifically as it pertains to um robotics and AI.
• 6:49
  SPEAKER_01: So Because this is an issue, right?
• 6:52
  SPEAKER_01: I mean, obviously it's your job to invest in technology, but also the people who are investing in you want a return for that money.
• 7:03
  SPEAKER_01: What kind of what kind of runway are you generally looking at?
• 7:07
  SPEAKER_01: And how do you sort of monitor progress and success along the way?
• 7:13
  SPEAKER_00: The timelines for our investments are around a decade by design.
• 7:18
  SPEAKER_00: And eventually the saying is every good 10-year fund is likely a 12 to 15 year fund, right?
• 7:23
  SPEAKER_00: So that's our framing for the timeline.
• 7:26
  SPEAKER_00: When we sit down with a roboticist or a team of roboticists trying to do something, it's interesting.
• 7:32
  SPEAKER_00: I think we've had some short-term thinking creep into the value over the last decade, decade and a half, because you've had software projects that get to maturity in two, three years, five years, maybe massive value inflection.
• 7:42
  SPEAKER_00: But something hard and meaningful does take time.
• 7:44
  SPEAKER_00: So I think the conversation we've had are often like, okay, let's start here, but where do you think it gets to five to seven years from now as the initial traction point?
• 7:55
  SPEAKER_00: And then where does it get to in seven to 12 year framing for maturity?
• 7:59
  SPEAKER_00: And that's not an easy conversation to have because I think our ability to look past one tech cycle is severely limited to begin with.
• 8:06
  SPEAKER_00: Um, so it's an unfair question almost to ask the founder what might happen in seven to ten years, because uh I think both of both of these sides in the conversation are likely able to look at about three years, maybe three and a half years kind of time frame.
• 8:20
  SPEAKER_00: But I think that question again frames the discussion well.
• 8:23
  SPEAKER_00: So it lets us look past near-term disruptions or near-term technological limitations that might um encumber one's idea and really look at what could be possible long-term.
• 8:35
  SPEAKER_01: It's wild though, because you get started and then there's this like decade-long period of waiting, and then you get to see what all that return is.
• 8:42
  SPEAKER_01: But how many, I guess, how many sort of balls in play do you have to have in the meantime?
• 8:49
  SPEAKER_00: So by design, our venture fund, we it uh in every fund, we invest in 40 to 55 companies.
• 8:54
  SPEAKER_00: And we know 10 years from today, likely just our history and patterns tell us five to eight of them survive in about 10 years.
• 9:03
  SPEAKER_00: So if we have a tremendously high failure rate.
• 9:05
  SPEAKER_01: Um and I think that's how does that, I mean, I should I should know like stats like this, but how generally how does that compare to other early stage investment firms?
• 9:14
  SPEAKER_00: I think that's about average for the whole industry.
• 9:17
  SPEAKER_00: Yeah, about on par with the industry.
• 9:19
  SPEAKER_00: I think now what we are trying to do by design is to have a high ownership in those that survive as well.
• 9:25
  SPEAKER_00: That's why we want to start at the seed point.
• 9:27
  SPEAKER_00: We want to write our first check, which acquires often 20 to 22 percent of the company.
• 9:32
  SPEAKER_00: Um, and then while that might get diluted, and we'll be participating in nearly every funding round a company might raise over the next five to 10 years.
• 9:40
  SPEAKER_00: Uh, and our strategic goal is to be one of the largest investors seven to 10 years from now.
• 9:46
  SPEAKER_00: That's how we make money.
• 9:47
  SPEAKER_00: So if you do the fund level math there, if you have three to five companies in 10 years, and each one of them, let's say we have a 12 to 15% ownership and the exit at a meaningful high number, we would have returned a multiple of the fund to our investors.
• 10:02
  SPEAKER_00: And I think then the cycle works.
• 10:04
  SPEAKER_00: And we are fully aligned with our investors, our LPs, with the founders in value creation, and then as a partnership.
• 10:10
  SPEAKER_00: And that when it works, venture is beautiful, absolutely.
• 10:14
  SPEAKER_00: Because you've helped bring something into the industry that otherwise would not likely uh have come into existence, and you've made money for your investors and you've enabled the founders to hopefully go realize what their dreams were.
• 10:27
  SPEAKER_00: Um, so I I do think it works, and now it's direct answer to your question.
• 10:31
  SPEAKER_00: I think in our industry as a whole, from seed to final maturity, uh, that is about the hit rate.
• 10:38
  SPEAKER_01: Looking in your portfolio, looking in the true portfolio, like there are some companies that have matured.
• 10:43
  SPEAKER_01: There are companies that are like out there with products.
• 10:46
  SPEAKER_01: I mean, diligent, right?
• 10:48
  SPEAKER_01: That is a company that you had invested in.
• 10:49
  SPEAKER_01: That is a company that will be on stage uh with us at our upcoming event that uh I'm plugging that may or may not be after this podcast comes out.
• 10:58
  SPEAKER_01: Um so, so like that's an example of somebody that you invested on early on.
• 11:03
  SPEAKER_01: That it took how long did that take to mature?
• 11:07
  SPEAKER_00: I I think it's been about nine years, um, if I remember correctly, for Diligent and then several others as well.
• 11:13
  SPEAKER_00: I mean, we had invested right around that timeframe in Abbott bots, which is not really hitting maturity.
• 11:19
  SPEAKER_00: Bear flag robotics, which is a great outcome for the founders in the team.
• 11:22
  SPEAKER_00: After around a round and a half of financing acquired by John Deere for 250 million, that was one of the earlier uh outcomes in robotics broadly, in terms of the time frame from seed to an acquisition.
• 11:33
  SPEAKER_00: I think it was about four years for them.
• 11:35
  SPEAKER_00: And for several of the others we are involved with, um, which are all task-focused robots, by the way, um, it has been sort of seven to eight to nine years to get to maturity when you have a hundred plus robots in the field.
• 11:48
  SPEAKER_00: I mean, that's mentally how I frame or define maturity.
• 11:50
  SPEAKER_00: When you have product in the field, your robots, your autonomous machines are learning from the environment that they're in.
• 11:56
  SPEAKER_00: They're doing a valuable task for the customers that they are deployed with.
• 12:01
  SPEAKER_00: Um, so that that's what I mean.
• 12:02
  SPEAKER_00: I and uh and you can sort of say first prototypes probably took two, two to three years.
• 12:07
  SPEAKER_00: Maturing those prototypes into second, third refinements took another couple of years.
• 12:11
  SPEAKER_00: And then delivering those in first few tens and then eventually hundreds takes time.
• 12:17
  SPEAKER_01: There are differing points of view as far as whether an acquisition is necessarily like a positive outcome.
• 12:23
  SPEAKER_01: Obviously, it depends on sort of the circumstances behind it.
• 12:25
  SPEAKER_01: But it but it strikes me that especially in the case of a robotics hardware company that is so expensive to operate that that generally can be a positive outcome, especially when you have somebody like the size and the resources of a John Deere.
• 12:41
  SPEAKER_00: I think there are many positives to an acquisition.
• 12:44
  SPEAKER_00: Um and frankly, for startups, I mean, if you look by the numbers, um 80 plus percent of outcomes are MA or mergers and acquisitions.
• 12:55
  SPEAKER_00: If you survive and become independent publicly traded, perhaps at profitable companies.
• 13:00
  SPEAKER_00: So we know the odds are against us when it comes to having an independent, profitable surviving entity 10, 12 years from now.
• 13:08
  SPEAKER_00: So MA is always a possibility.
• 13:10
  SPEAKER_00: In some cases, it happens for technology, some cases for product, some cases for product technology and revenues, right?
• 13:17
  SPEAKER_00: So all kinds of reasons.
• 13:18
  SPEAKER_00: And I think certainly for beer flag robotics, thinking back to that time in I think 21 when they were acquired, the amplification of the impact they could have with John Deere was clear.
• 13:28
  SPEAKER_00: Uh, it was just perhaps that was not the time to say, okay, I'm gonna from the ground up stand up not just an autonomous machine for agriculture, but an OEM at the same time, right?
• 13:40
  SPEAKER_00: So I'm gonna bend all the sheet metal necessary, make it autonomous, and um and make it um employable across the board in agriculture.
• 13:49
  SPEAKER_00: I think at that time that was a great combination.
• 13:51
  SPEAKER_00: And I think certainly history now, last five years, has has proven that they've been able to drive that product to completion and have an impact in other groups and projects within John Deere.
• 14:00
  SPEAKER_00: So it's been a very good outcome.
• 14:02
  SPEAKER_00: And certainly the founders, um, and I'm in touch with both of them often, E-Geno and Aubrey, uh, are very pleased that they were able to do that.
• 14:08
  SPEAKER_01: I think this might be changing.
• 14:10
  SPEAKER_01: And I, you know, this is my own personal bias.
• 14:12
  SPEAKER_01: I hope to a certain extent that this is changing.
• 14:14
  SPEAKER_01: But people who don't follow ag tech robotics closely might underestimate how much of a presence John Deere has been over the past like five to ten years, especially with all of the acquisitions that they've been making.
• 14:28
  SPEAKER_01: That like, uh, you know, people know John Deere, you know, they recognize the heavy machinery, but like they've been owning agriculture robotics for like a decade now.
• 14:38
  SPEAKER_00: Yeah, I think it's been admirable the way they have prosecuted this opportunity, not just with uh tractors in preparation of the fields or looking at harvesting, but also a very big push early into spraying, which is a very key component in the agriculture cycle for growing various kinds of crops from fruit trees and so on.
• 14:56
  SPEAKER_00: And they have recognized what robotics and automated uh and then autonomous operation on top of automation could deliver there and have consistently put together teams.
• 15:05
  SPEAKER_00: And if it didn't exist within JD, I think they've gone and acquired those and done that pretty coherently.
• 15:10
  SPEAKER_00: So, yeah, and they have a significant presence in addition to tech and products, which I think startups often miss, right?
• 15:16
  SPEAKER_00: If I'm a farmer with a thousand acres on the West Coast or in the Midwest doing broadacre crops, I need the full cycle of product capability.
• 15:24
  SPEAKER_00: I need the pick product technology capability, servicing, maintenance, the capital and the leasing uh options for me to acquire all this equipment and operate them, right?
• 15:34
  SPEAKER_00: So that's what I get when I work with a large outfit like John Deere, which I likely won't get working with a startup.
• 15:41
  SPEAKER_00: So I think it's it's a it's a big and complex area.
• 15:45
  SPEAKER_00: I think JD has a commanding position there.
• 15:47
  SPEAKER_00: And then there are several others that are notable as well.
• 15:49
  SPEAKER_00: I think Case New Holland, Holland uh CNH continues to put out great machines.
• 15:54
  SPEAKER_00: Uh, Kaboda has been um interesting around the edges, trying to figure out a better-than linear path to owning more of the specialty crafts, perhaps in the US and elsewhere.
• 16:04
  SPEAKER_00: And then several other contenders next to them, including Mahindra out of India, which has been very tech forward the last five, six, six years.
• 16:11
  SPEAKER_00: I think it's not a name that is commonly heard in the US yet.
• 16:14
  SPEAKER_00: But if you look at the plans they have, and certainly the public announcements coming out of them in the last few years, point to a very coherent strategy around how to do more agriculture uh across nearly every corner of the world.
• 16:27
  SPEAKER_00: Big emissions, big emissions.
• 16:29
  SPEAKER_01: Yeah, I hear this a lot from from I guess sort of seasoned robotics founders.
• 16:35
  SPEAKER_01: Like one of the hard lessons that it seems like everybody has to learn is um obviously really getting the feedback of the people who sort of work and live in that space, right?
• 16:45
  SPEAKER_01: And and and that's especially like we talked to Tessa Laura recently, and she was talking about the experience of really sort of like living in and around construction.
• 16:54
  SPEAKER_01: But more than any other sort of subrobics category, I hear this about AgTech.
• 16:58
  SPEAKER_01: And it makes sense, right?
• 16:59
  SPEAKER_01: I mean, that these are these are farmers, the generational, like they eat, breathe, sleep, that.
• 17:06
  SPEAKER_01: So when we're talking about like bear flag and your involvement with them initially, was that something that you feel like they had baked in from the beginning?
• 17:13
  SPEAKER_00: I think what Igenob and Aubrey brought to the table was just an unusual degree of understanding of their eventual customer.
• 17:22
  SPEAKER_00: I think they were very keenly aware that a farmer today is almost like an everyday scientist, right?
• 17:29
  SPEAKER_00: In the face of changing weather constantly, not holding up to the patterns they had learned over generations, perhaps, or the last few decades, variables that are not really in their control.
• 17:41
  SPEAKER_00: And then we forced to make a decision in near real time like, do I go plow this field over there, get it prepped because it's gonna rain next weekend?
• 17:48
  SPEAKER_00: And do I have the labor available to go do that?
• 17:51
  SPEAKER_00: It's a near continuous set of challenges in front of them.
• 17:54
  SPEAKER_00: Uh now, growing up in India and growing up in a family that was half farmers and half teachers, I certainly had a kind of view into that.
• 18:01
  SPEAKER_00: My father, um, who ended up um later uh as a physicist in the US and went back to India, was the last generation still to be tilling his fields.
• 18:10
  SPEAKER_00: I mean, they bought a tractor, got to work.
• 18:13
  SPEAKER_00: I learned to drive on a tractor.
• 18:14
  SPEAKER_00: Um, so in a weird roundabout way, when it came to meeting with the bear flag team, I think, I don't think they realized it, but they were talking to somebody who had literally learned to drive on a tractor and had an unusual interest in the field, and I think an affinity to understand what those problems were.
• 18:28
  SPEAKER_00: And I was surprised because looking at the background, just like, okay, you you one of them grew up in the Bay Area, the other one grew up in in, I think, Colorado somewhere.
• 18:37
  SPEAKER_00: Um, you don't come from farming families, you haven't been around this issue.
• 18:40
  SPEAKER_00: So I was skeptical to begin with, right?
• 18:42
  SPEAKER_00: So, how do you understand this problem?
• 18:44
  SPEAKER_00: And then the depth of understanding they were able to describe was phenomenal.
• 18:50
  SPEAKER_00: And they had done that because they spent the time in the field with the growers day in, day out.
• 18:55
  SPEAKER_00: Like, okay, tell me how do you make this decision?
• 18:58
  SPEAKER_00: And then it turns out, on top of being a scientist, an everyday scientist, the farmers also are required to be economists on the fly.
• 19:05
  SPEAKER_00: What level of input cost can I bear?
• 19:07
  SPEAKER_00: What kind of forward future pricing can I get around this crop or that crop?
• 19:11
  SPEAKER_00: Looking at the yield I might or may not get.
• 19:13
  SPEAKER_00: Um, and I think that was their insight that was clear at the table at the time of our seed check, that yeah, they understand the solution reception side extremely well.
• 19:26
  SPEAKER_00: And of course, they are very talented analysis.
• 19:28
  SPEAKER_00: So their ability to then describe why that machine made sense and what capabilities might the machine require, I think had just a higher degree of authenticity to it.
• 19:38
  SPEAKER_01: Like I said earlier, I'm a big believer in, you know, you can get pretty far in this life, just knowing like the right questions to ask.
• 19:44
  SPEAKER_01: Obviously, like that's been the entire basis of my career.
• 19:47
  SPEAKER_01: But that that seems to me from the outside to be like quite the jump.
• 19:51
  SPEAKER_01: How does one go from farmer to physicist?
• 19:54
  SPEAKER_01: I'm asking about your dad specifically.
• 19:56
  SPEAKER_01: I want to hear that.
• 19:57
  SPEAKER_00: That was a phenomenal journey.
• 19:59
  SPEAKER_00: Um, so he's nearly 90 now.
• 20:02
  SPEAKER_00: And uh the the story there was his dad, uh his my grandfather was a school teacher and farmer on the side because you had to supplement whatever income you had.
• 20:11
  SPEAKER_00: So he grew up pretty much as a farmer, went to school, got a degree, and thought, hey, I could teach and farm at the same time.
• 20:17
  SPEAKER_00: And that's what he did.
• 20:18
  SPEAKER_00: Till he met a couple of really smart people who said, Well, and we had really bad floods in India, northern India, right around the time of first year he was into almost full-time farming.
• 20:29
  SPEAKER_00: And they said, Look, you're at the mercy of things outside your control.
• 20:33
  SPEAKER_00: You're a bright, talented young person.
• 20:35
  SPEAKER_00: Why would you want that?
• 20:36
  SPEAKER_00: And that really stuck with him.
• 20:38
  SPEAKER_00: Like, okay, what am I doing here?
• 20:40
  SPEAKER_00: And to his credit, he looked around and said, Well, being in that point, he had a general BA degree and was a high school uh teacher for a few months.
• 20:50
  SPEAKER_00: Um, went back to school, decided to choose physics instead.
• 20:54
  SPEAKER_00: That was the toughest you could do.
• 20:56
  SPEAKER_00: Uh, I was he was pretty good at studies.
• 20:58
  SPEAKER_00: I think the first time around he'd finished college by 17 or so he had done high school by 14.
• 21:03
  SPEAKER_01: Do you feel like he intentionally went for the toughest thing that he could find?
• 21:06
  SPEAKER_00: Oh, yes.
• 21:07
  SPEAKER_00: Yeah, well, yeah, he th his calculation was at least if I can do this, there'd be an economic outcome that is proportional to the toughness of what I'm going into.
• 21:16
  SPEAKER_01: There are fewer people going into it, or there are fewer people making it through.
• 21:19
  SPEAKER_00: Yeah.
• 21:20
  SPEAKER_00: So he went to physics, um, did masters, and then was a Fulbright scholar, um, and went to Harvard for Grass Studies on a school scholarship.
• 21:27
  SPEAKER_00: Uh and sidebar note, I mean, you can see now the environment today we have here, the US is ripping out all kinds of assistance for university RD.
• 21:36
  SPEAKER_00: The long cycle of that is decades long.
• 21:40
  SPEAKER_00: The reason that changed his life and probably our families, and then many other things in concatenation from that point on was um those floods uh also made the tractor they had, and they had to borrow big money from the bank, uh, knocked out the tractor.
• 21:53
  SPEAKER_00: It was a it was um a company that's the precursor to Massey Ferguson in Canada, and that's the tractor they had bought.
• 21:59
  SPEAKER_00: And And the flood waters were so high, uh, the tractor was not usable, and the bank loan was still there at pretty onerous terms.
• 22:06
  SPEAKER_00: So he actually paid back the bank loan from his scholarship at Harvard, and that lifted the family out of poverty quite literally, right?
• 22:16
  SPEAKER_00: Otherwise, there was no no outcome for the family in that scenario.
• 22:20
  SPEAKER_00: So, on one hand, his physics thing pan out, but also I think the scholarship part, which is largely the US income taxpayers, um changed the arc of that family and changed his arc in life.
• 22:32
  SPEAKER_00: Now he went on to be not just a teacher for three decades and then uh contributing on the scientific side.
• 22:38
  SPEAKER_00: Uh, he had invented instrumentation that is still in use today, 40 years after he invented it.
• 22:43
  SPEAKER_00: Plenty of royalty that went to all of his lab uh uh co um uh sort of classmates and and and co-workers.
• 22:50
  SPEAKER_00: And then in uh when he was in the US, he worked on the F-16 fighter avionics for general dynamics up in Rochester.
• 22:56
  SPEAKER_00: So pretty direct paybacks.
• 22:57
  SPEAKER_00: And then um, I mean, not to make too much of it, but I grew up in India, went to engineering, came here, started a company that contributed, I mean I can say without boasting, more than a billion dollars to the US GDP, 750 jobs, um, and probably have paid back, I don't know, thousands of what the original Food Bright Scholarship was back to the US Treasury, but after 50 years.
• 23:23
  SPEAKER_00: And and is hopefully a spreading economic impact.
• 23:26
  SPEAKER_00: So the arc of these things is long.
• 23:28
  SPEAKER_00: I think the the question you asked about his journey, uh, yeah, it wasn't a straight line, but I think it was a great outcome.
• 23:35
  SPEAKER_01: So if I'm reading between the lines here, perhaps um uh investing in very smart people abroad and having them come into our country uh to study might ultimately impact the country positively in the long run.
• 23:49
  SPEAKER_01: Is that fair to say?
• 23:51
  SPEAKER_00: I think it's very clear.
• 23:52
  SPEAKER_00: We see a day in, day out in the valley.
• 23:54
  SPEAKER_00: There are there are thousands of examples of that, but not just a few.
• 23:57
  SPEAKER_00: I mean, this these are not outlier people.
• 24:00
  SPEAKER_00: Uh, these are um this is a continuous flow of talent and energy into the valley.
• 24:05
  SPEAKER_00: I mean, the valley thrives on its young.
• 24:07
  SPEAKER_00: I think Kafka had that line in a short story.
• 24:09
  SPEAKER_00: I remember this about the cities being like city feeds on its young or something like that.
• 24:15
  SPEAKER_00: And to me, it's like there's a direct map.
• 24:17
  SPEAKER_00: No, the Valley, the Valley's engine uh is the constant stream of young people and the ideas that they bring with them.
• 24:24
  SPEAKER_00: We don't have ideas here by default, right?
• 24:27
  SPEAKER_00: We just have the environment that welcomes this young talent continuously, and I think that's that's it, that's the primer for the whole mechanism.
• 24:35
  SPEAKER_00: That's what gets it going.
• 24:37
  SPEAKER_00: If you look at the history, there's been a near continuous flow since the beginnings of the valley.
• 24:42
  SPEAKER_00: And I hope it continues.
• 24:44
  SPEAKER_00: I hope it continues.
• 24:45
  SPEAKER_01: It sounds like ultimately uh your father perhaps didn't stray too far from the farm because you, like you said, you've spent some time there, and you know that's how you learn to drive.
• 24:54
  SPEAKER_01: So the farm the round.
• 24:56
  SPEAKER_00: Again, these things go in long arcs.
• 24:58
  SPEAKER_00: So the the family still farms around 25 acres.
• 25:01
  SPEAKER_00: And I'm and my younger brother, after a full 15-year-plus career in IT in Bangalore, India, actually went back to do organic farming on the family land.
• 25:10
  SPEAKER_00: So he actually is today a farmer in North India.
• 25:12
  SPEAKER_00: So, weirdly enough, I get very routine reports from what's happening on that side, and not that different from farmers out here in Salinas and Central Valley in California.
• 25:22
  SPEAKER_00: The challenges are similar, the input costs are similar.
• 25:25
  SPEAKER_00: Um the the diesel costs and fertilizer and insecticide sprays go up.
• 25:32
  SPEAKER_00: Economic uncertainty uncertainty has only increased.
• 25:35
  SPEAKER_00: Uh, he's had a couple of wipeouts at minor scale of the weather changing and kind of just the crops drying out in the field or getting too much water and getting waterlogged and the yield going to near zero.
• 25:46
  SPEAKER_00: So that's the constant reality of the farming world, right?
• 25:49
  SPEAKER_00: It's it's unlike, it's not an industrial warehouse scenario where I can put in robots and most more or less things won't change.
• 25:55
  SPEAKER_00: Agtec is the double black diamond of robotics.
• 25:59
  SPEAKER_00: It's the hardest thing you can do.
• 26:01
  SPEAKER_01: It's funny the story you were telling earlier about somebody uh coming up to him and saying that, you know, why are you uh putting yourself in a situation where you're at the mercy of these things outside of your control?
• 26:11
  SPEAKER_01: The first thing that jumped in my head is, well, you know, there's no scenario in which we're not, you know, at the whims of forces that are out of our control.
• 26:19
  SPEAKER_01: We just hopefully, to a certain extent, have some control in deciding what those forces ultimately are.
• 26:25
  SPEAKER_00: Absolutely.
• 26:25
  SPEAKER_00: But I think in the farmer's scenarios, they're just exposed to this sort of nature global level, climatic and other changes and have to respond to them.
• 26:34
  SPEAKER_00: It's like primal, yeah.
• 26:36
  SPEAKER_00: It is very primal.
• 26:38
  SPEAKER_00: Uh so I think for rotators going into that domain, they need that visceral understanding of what those sets of challenges are and align with the farmers first before you invent a piece of metal that would be propelled somehow by hopefully electrically and and be autonomous in perception and carry out great tasks for which hopefully we won't need human labor.
• 26:59
  SPEAKER_01: Yeah, I mean it's interesting that that was the other thing that jumped in my head is it's another way in which your family was kind of ahead of the curve there, like dealing with the flooding in the farm.
• 27:08
  SPEAKER_00: It's it's not easy.
• 27:10
  SPEAKER_00: I mean, uh look, it's it the the scale at which these things could occur uh is is overwhelming for individual farmers.
• 27:17
  SPEAKER_00: So I think to the extent technology can help, uh, I'm a big believer that technology will and should help them tackle these very sort of existential questions.
• 27:27
  SPEAKER_00: And and it goes it spans the whole arc from predicting weather better to having um more intelligence in in play at every level to having better seeds, to having sort of an understanding of soil dynamics and microbes that live there, and how do we encourage those to flourish?
• 27:45
  SPEAKER_00: And of course, do all the other mechanical work that is still going to be always needed.
• 27:49
  SPEAKER_00: And also, there's no scenario in which we don't need farming, right?
• 27:54
  SPEAKER_00: Six billion people on the planet, we gotta feed.
• 27:56
  SPEAKER_01: For your father specifically, and and I guess for you as well, um, coming from these two different worlds, how how does something like that impact your concept of success?
• 28:08
  SPEAKER_01: And did he have a really sort of clear concept of what success would look like for his own children?
• 28:13
  SPEAKER_00: My parents were both unusual in the sense that, unlike probably most middle-class Indian parents, they did not compel us to think in the ways they did, but they opened up this huge world to us.
• 28:26
  SPEAKER_00: My mother was a professor of English uh at the engineering school where my dad taught physics.
• 28:31
  SPEAKER_00: And I think they just kind of the the biggest thing they they did for us as kids was they made sure we understood the world was big and full of possibilities, and you didn't have to be constrained to the local realities.
• 28:42
  SPEAKER_00: And I think that was such a great gift to not be constrained by what we saw around us.
• 28:47
  SPEAKER_00: Um and and and I think I well, I went the full spectrum of initially I wanted to be a journalist because my my uh mom's side of the family has plenty of those, and that looked very exciting and appealing.
• 28:57
  SPEAKER_00: So that's what I was gonna be.
• 28:58
  SPEAKER_00: Then I wanted to be an architect because building big things seemed very compelling and interesting.
• 29:04
  SPEAKER_00: And then eventually I went to electrical engineering because that seemed like, wow, that this is just amazing because communications.
• 29:09
  SPEAKER_00: I mean, how could you not sort of do more towards that?
• 29:13
  SPEAKER_00: So I think it was interesting, and we had the freedom to do that.
• 29:16
  SPEAKER_00: And in hindsight, I realized that was unique and lovely to grow up with.
• 29:20
  SPEAKER_00: Um, so suddenly they gave us the love of new things, the possibilities that exist beyond just what we could see locally.
• 29:27
  SPEAKER_00: Um, I think that was great.
• 29:28
  SPEAKER_00: That was wonderful.
• 29:29
  SPEAKER_01: It sounds like though, that the way in which you sort of went really from one career to another was fairly organic.
• 29:35
  SPEAKER_01: I think in hindsight it appears organic, but it really wasn't.
• 29:39
  SPEAKER_01: It was more calculated than it looks.
• 29:41
  SPEAKER_00: It seems calculated, but it really wasn't.
• 29:43
  SPEAKER_00: It has been, I mean, my life really has been a series of very impulsive decisions, and I'm just tremendously privileged to not have suffered from that because it could have turned out really badly.
• 29:54
  SPEAKER_00: I think I've just been lucky to be with very good people at the right time in my career who freely offered, if not advice, an insight into how I might think about something and were accommodating of this newness that I certainly had.
• 30:07
  SPEAKER_00: Like after undergrad, I went to Canada, right?
• 30:09
  SPEAKER_00: Like I said, University of Alberta Edmonton.
• 30:11
  SPEAKER_00: And the big biggest attracting point for me there was it was fully paid, of course, because I couldn't have afforded to go without it, full scholarship.
• 30:18
  SPEAKER_00: But it was also off-campus.
• 30:19
  SPEAKER_00: I actually never went to campus for either a master's or a PhD.
• 30:22
  SPEAKER_00: It was a research lab off-campus run by the government and the industry together alongside the university.
• 30:28
  SPEAKER_00: So it let me just bug this endless number of engineers down the hall who had all come from industry.
• 30:34
  SPEAKER_00: So while I was a grad student, I could ask them about how this made sense or not in the real environment.
• 30:40
  SPEAKER_00: So I think that was such a privilege to go through that kind of education and then switch.
• 30:44
  SPEAKER_00: So my master's was thesis based on cellular mobile communications, signal processing.
• 30:49
  SPEAKER_00: Um, and then I switched because fiber optics seemed more interesting.
• 30:52
  SPEAKER_00: So I switched to optics and was able to do that for my PhD and finish it in in less than three years, which is, I think, a good thing.
• 31:00
  SPEAKER_00: Because again, they accommodated all the things I pushed for and let me get through it.
• 31:05
  SPEAKER_00: And then I was able to apply it to telecom switching uh in my startup here in the Bay Area, O and I systems.
• 31:12
  SPEAKER_00: But I can literally see there were two to five people around me at each of those steps.
• 31:18
  SPEAKER_00: That made those steps easier for me, even though there was no coordinated forward thinking on my part.
• 31:25
  SPEAKER_00: I think they they they were accommodating, um, patient with my questions, uh, and I think gracious with their time.
• 31:33
  SPEAKER_00: And I really, really appreciate that.
• 31:35
  SPEAKER_00: I think without that it would have been a messy outcome.
• 31:38
  SPEAKER_01: This I think is really important, and this is a thread I want to pull on, especially for you know I'm gonna make myself sound old, and I'm gonna make us all sound old, but I think especially for um for younger people coming up now, because obviously um the way we've interacted socially has evolved a lot and you know, even before the pandemic, but the pandemic played a big part in that as well.
• 32:02
  SPEAKER_01: So I mean, what is your advice as far as being in a position to cultivate the right people, to keep them around, and to make sure they're there at the right time?
• 32:12
  SPEAKER_00: For any young innovator um stepping into college or stepping from college into a career, um talking to more people continuously seems to help.
• 32:23
  SPEAKER_00: Now, you don't have to follow everything they might say to you, but you have to really listen.
• 32:27
  SPEAKER_00: Because I think you're getting for nearly for free, except for investment of time, this condensed set of observations, life lessons, um, and maybe advice, but you don't have to take it.
• 32:39
  SPEAKER_00: But you should listen to it because I think it it gives you one additional data point, um, nearly for free or at a very low cost that you learn from.
• 32:47
  SPEAKER_00: So why won't you learn from it?
• 32:48
  SPEAKER_00: It's a sort of like the LLMs today.
• 32:50
  SPEAKER_00: Like, would I throw more data at it so that they know better?
• 32:52
  SPEAKER_00: Of course I should.
• 32:53
  SPEAKER_00: There's no scenario where the LLMs are so smart that they don't need data anymore, right?
• 32:57
  SPEAKER_01: What's the thing?
• 32:58
  SPEAKER_01: Garbage in, garbage out.
• 32:59
  SPEAKER_01: I mean, there is there is bad data that can be fed into the system.
• 33:02
  SPEAKER_00: So I think that's the only way to overcome that is scale.
• 33:05
  SPEAKER_00: You want a massive scale of data input all the time.
• 33:08
  SPEAKER_00: And yes, you're gonna get into some questionable inputs for sure, but I think you'll have the ability to discard that.
• 33:13
  SPEAKER_01: This is something that I've struggled with.
• 33:15
  SPEAKER_01: I wonder, you know, I wonder if you have as well.
• 33:17
  SPEAKER_01: You're obviously a very affable person, and and we've talked a few times, and it's always been a very uh open and transparent conversation.
• 33:23
  SPEAKER_01: I appreciate that.
• 33:24
  SPEAKER_01: But for people that do you know what you do, for people certainly that work in in robotics, that work in coding, like socialization is not necessarily something that comes naturally.
• 33:36
  SPEAKER_00: It can be done, though.
• 33:37
  SPEAKER_00: Maybe you may have to frame it in more transactional terms for that personality type, but it certainly can be done.
• 33:44
  SPEAKER_00: If this, then that.
• 33:46
  SPEAKER_00: Yeah, maybe, maybe.
• 33:47
  SPEAKER_00: I haven't thought of those terms, but certainly if that works for people.
• 33:51
  SPEAKER_00: But there's an obvious reward here for doing so.
• 33:53
  SPEAKER_00: So it is worth the engagement time and effort.
• 33:56
  SPEAKER_00: And also, I have never ever thought of myself an extrovert who just sort of goes and talks to people.
• 34:02
  SPEAKER_00: I think it's I am an introvert, but also I realize that if I don't continuously listen to other people, uh it's it's not relevant.
• 34:12
  SPEAKER_00: My thinking isn't relevant because you have to continuously, you test it against the the insights that others have.
• 34:19
  SPEAKER_00: You have to form it with the folks around you.
• 34:23
  SPEAKER_00: I think there's a big myth in the Valley that you lock, go lock yourself in a garage and you appear in two years and you'll have this massive number of insights.
• 34:29
  SPEAKER_00: And I just don't think that's how innovation works.
• 34:32
  SPEAKER_00: I think innovation works within constant engagement with your ecosystem, whether the ecosystem is technologists or customers or a planner, both, hopefully.
• 34:40
  SPEAKER_00: Out of that interaction has come, I think, just about everything of importance and meaning in our technical careers.
• 34:47
  SPEAKER_00: And I'm a big believer that's how it happens the next couple of decades as well.
• 34:51
  SPEAKER_00: So, if anything, if I was researching OpenAI today, I mean, in addition to my day job, yeah, I should be talking to everybody else.
• 34:57
  SPEAKER_00: And I think they they do.
• 34:59
  SPEAKER_00: Uh, I don't I don't think that has gone either generationally out of fashion or somehow um easier or harder than it used to be.
• 35:07
  SPEAKER_00: It's about the same.
• 35:08
  SPEAKER_00: But the tools certainly are there, like it today.
• 35:10
  SPEAKER_00: I mean, on WhatsApp groups, I'm connected to the folks I went to grade school with, high school and engineering school, across the three decades removed, right?
• 35:18
  SPEAKER_00: And it's it's brilliant.
• 35:20
  SPEAKER_00: Uh so I would hope for the generation today, um, it may seem like they have too many inputs, but uh, I think that the the cost to getting those inputs has has nearly gone to zero.
• 35:30
  SPEAKER_00: So I hope they do a lot more.
• 35:32
  SPEAKER_01: Something that I've learned, you know, having been in journalism for uh as long as I have, is that every single industry is a lot smaller than you think it is looking from the outside.
• 35:42
  SPEAKER_01: And and uh I this is like again, this shouldn't be viewed transactionally, this should just be like something that is inherent and innate in all of us.
• 35:51
  SPEAKER_01: But you have to help people where you can because there's going to be a time uh probably when they're in a position to pay that back.
• 36:00
  SPEAKER_00: I think that give and take is continuous or ought to be continuous.
• 36:04
  SPEAKER_00: Um and I think going back to all the way to founders, I mean, there's there's no downside to a potential group of founders just spending an outrageous amount of time with their potential customers.
• 36:15
  SPEAKER_00: Just talk to them.
• 36:16
  SPEAKER_00: Talk to them for three months before you ever figure out what machine you might build, right?
• 36:20
  SPEAKER_00: And I think on average, if they did more of that, I think we'll have higher quality ideas in startups.
• 36:26
  SPEAKER_00: Uh I I don't think it happens as much as it it could.
• 36:30
  SPEAKER_01: That's probably a question of resources, too.
• 36:32
  SPEAKER_01: You know, obviously, like you still need to make ends meet while you're doing that.
• 36:36
  SPEAKER_01: Yes.
• 36:37
  SPEAKER_01: I I think in much the same way that there are no two stories of the path into VC that are the same.
• 36:46
  SPEAKER_01: I suspect that no two VCs have uh similar calendars day to day.
• 36:52
  SPEAKER_01: Um in terms of like your breakdown, how much of that time is really just sort of spent talking to people, how much of that time is sort of sitting there hearing pitches?
• 37:00
  SPEAKER_01: Like what do you let let me rephrase that.
• 37:02
  SPEAKER_01: What do you do all day?
• 37:04
  SPEAKER_00: Well, I I I get to spend every single day with brilliant people, right?
• 37:09
  SPEAKER_00: So just this morning, I I look at my calendar.
• 37:11
  SPEAKER_00: This morning was two and a half hours of a board meeting with a brilliant leading it, leading uh in its domain satellite company based in Finland.
• 37:20
  SPEAKER_00: They have operations across Finland, Spain, Poland, UK, US, of course.
• 37:25
  SPEAKER_00: And then hearing them solve challenges that just a few years ago seemed just impossible is just tremendously energizing.
• 37:32
  SPEAKER_00: So there's that to the day.
• 37:34
  SPEAKER_00: So by no means it's sort of always a two-hour board meeting I'd get slot through.
• 37:38
  SPEAKER_00: It's just an amazing set of new inputs into your brain.
• 37:41
  SPEAKER_00: Um, and I love that part.
• 37:42
  SPEAKER_00: And then there's gonna be, I think, today's three new pitches from the young founders who are trying to do something.
• 37:49
  SPEAKER_00: And then in between, every time I can find uh I go read.
• 37:53
  SPEAKER_00: Uh I read in an absolute sort of uh continuous amount of inputs, news, analysis, political news, which you can't escape these days, uh, what's happening in our industry.
• 38:06
  SPEAKER_00: I don't get to read as much of our own technology um uh tech publications as much as I used to, but I try to.
• 38:14
  SPEAKER_00: Um there's a full stack over there making me feel guilty, but I get to them when I can, just to keep up with what's going on, not to learn firsthand there, but to really sense where the directional vectors might be and how they change.
• 38:28
  SPEAKER_00: And I think if you don't do this continuously in our business, you have to do it continuously because it compounds over time.
• 38:34
  SPEAKER_00: So you can't just say, well, one day I'm gonna go sit down and do all this.
• 38:38
  SPEAKER_00: I think you do this work in tiny increments all the time.
• 38:40
  SPEAKER_00: And it's a year from now, you would have found you've changed your thinking, you've changed the way you think about something.
• 38:46
  SPEAKER_00: Um, and there's a continuous the luxury of this position is there is a continuous interaction possible with bright people bringing new ideas to the table all the time.
• 38:56
  SPEAKER_00: In fact, we are always in a very unfair position of learning from the best, right?
• 39:01
  SPEAKER_00: Right?
• 39:01
  SPEAKER_00: So we might fool ourselves and say, Well, I'm thinking at such a brilliant high level.
• 39:06
  SPEAKER_00: And the right answer is 90% of that is likely because you met 10 brilliant founders in the last 10 days.
• 39:13
  SPEAKER_01: As a journalist, I can't be an expert on every single thing I write about.
• 39:17
  SPEAKER_01: As a VC, certainly like you could you could be the robotics guy, you could be the AI guy, but as far as secondary isn't the right word, but you know, ag tech, right?
• 39:26
  SPEAKER_01: Construction, all these other things.
• 39:27
  SPEAKER_01: In terms of those, and in terms of, I guess even beyond that, in terms of really sort of deep technology, in terms of like physics, in terms of these things that maybe you're not an expert about, like how do you that technologies?
• 39:42
  SPEAKER_00: I think two things help.
• 39:44
  SPEAKER_00: I mean, having an engineering education really does let you think of these remarkably complicated machines from first principles, right?
• 39:52
  SPEAKER_00: So often getting to those in the meeting is very educational, right?
• 39:56
  SPEAKER_00: I mean, just sort of let's figure out the basic physics energy level.
• 40:00
  SPEAKER_00: How does this machine work and scale?
• 40:02
  SPEAKER_00: And I think there's a if there's an ability to understand that, that's a good one to have.
• 40:06
  SPEAKER_00: And then if you were in a meeting with Elizabeth Holmes, you feel like you would have been like, uh, I think if you look at, well, I I have since deleted my Twitter account, but if you look at my Twitter feed, the very first time I found out about it, um, I actually referred to this talk I've I'd heard from a Stanford uh double E professor who was also an astronaut, I think.
• 40:24
  SPEAKER_00: And he had he had this really elegant curve that to look at these kind of statistical phenomena, what kind of volume would I need?
• 40:32
  SPEAKER_00: And and he was working in blood.
• 40:34
  SPEAKER_00: That had just that curve had stuck in my mind.
• 40:36
  SPEAKER_00: So when I looked at first at their claims and was like, well, there's no way a single drop of blood at a few milliliters will give you just statistical confidence in those kind of phenomena.
• 40:47
  SPEAKER_00: So I think there was there's always that, but I think what that education helps with is not so much it answers the question, but it lets me ask the right question sometimes.
• 40:57
  SPEAKER_00: And I and I think that that goes a long way.
• 40:59
  SPEAKER_00: And then the other part is is the other side we covered earlier, which is how much empathy do these groups of founders have for the customers and the customer scenarios they're likely going to solve.
• 41:10
  SPEAKER_00: We have found good founders have an unbelievably deep empathy level for how their product will be received, used, and put to work.
• 41:20
  SPEAKER_00: And they may not put it in those terms or those words specifically, but they have that.
• 41:25
  SPEAKER_00: And it's very clear.
• 41:27
  SPEAKER_00: And they and they demonstrate it by talking about how their product might change the way things are done, but but they invariably have that element to their pitch, which is I can describe what an hour in the life of my customer is going to be like if I ask them that.
• 41:42
  SPEAKER_00: So I think that part is useful.
• 41:44
  SPEAKER_00: So for us, most meetings are equal part that let's look at the innovation you have and at least sort of sanity check it, but then quickly shift to the other side of okay, what does this product and the potential market might create and the customer behavior it might entail?
• 41:58
  SPEAKER_00: What does that look like?
• 41:59
  SPEAKER_00: And do you have a good picture of that?
• 42:01
  SPEAKER_00: Um, and the right founders, the really good founders would either outright be honest and say no if they haven't done that, or if they have done that, they'll have a pretty good image in their heads, they're able to describe what that scenario might be like.
• 42:13
  SPEAKER_01: So you have there been cases of pitch meetings where you've just been like, hey, I like you, you're smart, but you need to go.
• 42:20
  SPEAKER_01: I'm just using as an example because a diligent for obvious reasons is top of mind, but you need to go to the hospital and hang out with nurses for a few more years.
• 42:28
  SPEAKER_00: In robotics, that happens often.
• 42:30
  SPEAKER_00: Now you you brought up Tessa.
• 42:31
  SPEAKER_00: I think Tessa is just a remarkable problem solver.
• 42:34
  SPEAKER_00: But now we met her just too late when she was initially raising money, otherwise, that would have been an easy check to write.
• 42:40
  SPEAKER_00: But her ability to understand in those days what a general contractor might be facing on a floor under construction commercially and the markings and the errors inherent in it, and how might you solve for that?
• 42:53
  SPEAKER_00: But she had this great understanding of how it was done prior manually and with this antiquated technologies.
• 43:00
  SPEAKER_00: And she could go toe-to-toe with anybody on well, how do you have sort of this GPS GNSS fix over there and you translate it floor by floor and how the errors build up?
• 43:10
  SPEAKER_00: That understanding, and never mind the machine she was going to build to solve that, that understanding just stood out.
• 43:15
  SPEAKER_00: So I think founders have that ability.
• 43:18
  SPEAKER_00: And I'm sure there are many.
• 43:19
  SPEAKER_00: I mean, I've we've been lucky to work with Andrea and Vivian at Diligent who have this.
• 43:23
  SPEAKER_00: I think the founders at Bear Flag, we've talked about certainly had that in spades.
• 43:26
  SPEAKER_00: Uh Pablo and Faizon over at AbbottBots.
• 43:29
  SPEAKER_00: Yeah, I mean, I can name all of the folks that we work with, uh, Davis and Jack and company at uh at Scythe Robotics with Lawnmowers.
• 43:36
  SPEAKER_00: They they just they knew the domain they would be solving for.
• 43:40
  SPEAKER_00: And I think um that has been great to see.
• 43:44
  SPEAKER_00: And they understand the environment in which their machines will exist.
• 43:48
  SPEAKER_01: As I was talking to Andrea in the early stages of putting this event together, I I feel like she and I I suspect probably Vivian too, that they're taking this, I will say, uh well deserved victory lap.
• 44:00
  SPEAKER_01: In terms of um I can't remember what their exact wording is, but you know, they they have they are one of the few companies out there that actually have a humanoid like really deployed at scale right now.
• 44:11
  SPEAKER_01: And and then like that ultimately is what every single startup, sure, but every single VC is looking to do is to not to not chase the trend, uh, is is to be a few steps ahead of that.
• 44:24
  SPEAKER_01: I mean, that feels like from the outside for you, a case where you know you saw something happening in the future.
• 44:32
  SPEAKER_00: I I think we certainly saw Andrea's and and Vivian's vision around how might these machines work, how might they learn.
• 44:40
  SPEAKER_00: And I think what they ended up really doing, if you think about it, was they described nearly perfectly what a humanoid-like function would be and what ecosystem it needs to exist in.
• 44:51
  SPEAKER_00: That's not always visible.
• 44:52
  SPEAKER_00: I mean, we have a remarkable amount of energy and effort going into humanoids, where a lot of that are machines that just look like humans with bipedal motion or with sort of two arms and two end effectors and so on.
• 45:06
  SPEAKER_00: And their views were very evolved even a decade ago, and how might that work?
• 45:10
  SPEAKER_00: And simple questions which are very hard to answer.
• 45:13
  SPEAKER_00: So, okay, I might get something that looks, walks, and feels to me visually like a like a human of some sort.
• 45:20
  SPEAKER_00: And then uh, even in those days, they were beginning to talk about how you might train these machines, right?
• 45:25
  SPEAKER_00: Because these are not industrial robots, right?
• 45:27
  SPEAKER_00: An industrial robot I can teach to do the same, weld on a on a car chassis a million times a day.
• 45:33
  SPEAKER_01: Because everybody at the time, and and this still applies now, they were it was all teleop, right?
• 45:39
  SPEAKER_00: I think the people had certainly done teleops, but beginning to have autonomous perception of the surroundings you're in, and then autonomous action taken on those inputs, that was the big shift that I think we have gone through over the last decade now in robotics, or we should have more and more of.
• 45:55
  SPEAKER_00: But it also requires this really hard questions to be asked, question to be answered, which is okay, who trains this robot?
• 46:02
  SPEAKER_00: I mean, if I've if I put a humanoid next to me, which is doing imitation learning of some kind, that's a very complex software scenario.
• 46:09
  SPEAKER_00: Should it learn from me?
• 46:10
  SPEAKER_00: Should it learn from the person next to me?
• 46:12
  SPEAKER_00: Should it learn from some so-called expert?
• 46:14
  SPEAKER_00: And how does it learn on the job?
• 46:16
  SPEAKER_00: What incremental learning is on the job versus already in there?
• 46:19
  SPEAKER_00: Those are not easy questions to answer, but I think they are as important as battery life and the physical phone factor and the weight it might have and the charge cycle it might need, all of those.
• 46:32
  SPEAKER_00: And the right answer needs to be solved.
• 46:36
  SPEAKER_00: Solve all of these things together.
• 46:38
  SPEAKER_00: I mean, that's what engineering is in the end, given the constraints, given the potential application scenarios, try to solve as many of the power problems you're gonna have, not just one or two.
• 46:48
  SPEAKER_00: So I think on one hand, it's a tremendously exciting domain, um, humanoids of many kinds.
• 46:54
  SPEAKER_00: Um, I think the other part is we haven't yet honestly thought through how the environment would be modified to permit these machines to work as well, which is which is going to be at the table.
• 47:06
  SPEAKER_01: Well, we'll be discussing more of this in the not to distant future.
• 47:10
  SPEAKER_01: Always a pleasure.
• 47:11
  SPEAKER_01: Thank you so much for joining us.
• 47:12
  SPEAKER_00: Anytime, Brian, it's a pleasure to talk to you.
• 47:14
  SPEAKER_00: Um, I hope some of this is useful to you by way of a rambling conversation.
• 47:18
  SPEAKER_00: Um, I'm sure you make some meaning out of this and and looking forward to the occasion um and the and the event in September.
• 47:27
  SPEAKER_01: Really great conversation.
• 47:28
  SPEAKER_01: Thank you so much to Roheet, and thank you to True Ventures.
• 47:32
  SPEAKER_01: Thanks to you for tuning in.
• 47:34
  SPEAKER_01: Uh, I hope everyone is staying warm and having a great whatever it is you do or don't celebrate.
• 47:42
  SPEAKER_01: All we're asking for this holiday season are some likes and subscribes.
• 47:46
  SPEAKER_01: And uh we will see you next week with another episode of Automated.