Shahin Farshchi — Lux Capital

Wednesday, May 25, 2022

Lux invests in emerging science and technology ventures at the outermost edges of what is possible.

Lux will write $10- 15M checks and they are not afraid to invest pre-product and sometimes pre-company.
Shahin has fascinating insight on automation, autonomous cars, AI and more.

View More Transcript

Shahin Farshchi is a partner at Lux Capital. Lux is a top VC fund with approximately two and a half billion under management. Lux invests in emerging science and technology ventures at the outermost edges of what is possible. Shahin is an engineer with a Ph.D. in electrical engineering and believes in accelerating humanity towards a brighter future through feats of engineering. I love that. Shahin, thanks for coming on the show today. Great to be here. Thank you for inviting me.

Awesome. Well, I usually like to start with just the basics of Lux so our listeners kind of get oriented on your current fund, stage, that sort of thing.

Totally so Lux started life in the bottom com bust with the hypothesis that the next wave of great companies will be rooted in innovations in the sciences. And since then, we’ve grown to invest across health care and technology. So now we have a team of about 10 investment professionals with backgrounds ranging from stem cell biology to physical chemistry to material science, to myself having a background in electrical engineering. And a lot of what I’ve done at Lux has been on the technology side.

So everything from basic semiconductors and communications technologies all the way to robotics, automation and autonomous cars and software tools.

Is this fund six for you folks? This is fun.

Six. So we are currently investing out of two pools of capital. One is a venture fund and the other is a growth fund. The growth fund is dedicated to doubling down on the winners out of the venture fund.

And we have about 20 percent of that set aside for investing in entry point growth companies. So check sizes can be typically around 10 to 15 million dollars out of the venture fund and double that for the growth fund.

Interesting. So I thought in your venture fund, you still would call yourself entering at really early stages, right? I think you said sometimes there isn’t there aren’t even paying customers.

Is that true? Sometimes there is even a company. So on many occasions, in fact, two over the past couple of months, we’ve invested in companies or behind people who have the expectation to start companies. So we’ve committed to being the first dollars in as soon as they choose to incorporate. And those have been amazing experiences for us as a firm and obviously turned out to be very personally rewarding for us because we feel like if we have a role in the creation and hopefully the ultimate success in these companies by being part of them at their inception.

And so in those scenarios, the checks may be a little bit smaller. It may be five to 10 million dollars.

And so so you’re doing seed investing, maybe series A investing again out the venture fund. But this is really preseed and is that distinct from I mean, not to put labels on things, but is that distinct from sort of the incubation efforts that you mentioned?

You guys are also doing so in some occasions. We do ideate companies. We’ve done that several times and it’s something that we’re very excited about, but that is distinct from getting behind entrepreneurs before there is a company. So there is the incubation of a company where one of us comes up with an idea and we put some people to work and we ultimately find people who would take over these businesses where we recruit leaders to take over these businesses.

I would put that in the incubation category. And then there is building relationships with top entrepreneurs and executives who are aiming to start companies. And we basically line ourselves up or position ourselves to be first money into those companies along with those entrepreneurs.

Has that changed for you, as you’ve seen? I don’t know, deep to get more trendy or more people coming into the space. I guess that’s my first question.

So these companies have certainly become more capital intensive. So I think the the motivation of founders to partner up with investors early has increased, knowing that rather than having to raise maybe, you know, 30, 40, 50 million dollars to get to an exit, they’ll have to raise hundreds of millions of dollars. So having a deep pocketed partner committed to the company on day zero is something that is perhaps more attractive to the entrepreneurs today than it was maybe several years ago.

So because I don’t have a billion dollar fund, you know, often actually it’s attractive to me to look at companies that are fairly capital efficient.

Would you say that you almost have the opposite approach.

I think having to raise a lot of capital is necessary for a lot of these businesses. So, for example, if you’re building a autonomous car company or if you’re building a semiconductor company or if you’re building a communications company where there’s an infrastructure component, then you certainly need to raise a significant amount of money. 

And so I wouldn’t say that we we seek companies that are super capital intensive, but we’ve positioned ourselves to attract the very best people who are building these companies that are off the beaten path.

What do you think makes them most interested in in working with you?

So founders look for people who can complement them, who bring skill sets that complement theirs.

Just because the founder is doing something in semiconductors doesn’t mean they should be talking to me. It looks I have other partners at Lux that have seen these kinds of companies grow and probably would be better fit for those foundries based on their personalities.

Well, then I’m dying to know what your personality and your style is, because, yes, if I were building a semiconductor company, I would probably come to you. But maybe you can tell me more about your your style.

That’s a good question. And I think this is probably a better question for my founders.

 And I think it goes back down to the humanity or the human aspect of this business. If this business was all about just money and Rolodex and connections, then it could be it could be very much automated.

And so I am the outcome or the products of my 15 years at Lux, having invested in deep tech companies, having seen these companies succeed and fail for many different reasons.

And I think founders come to appreciate certain subsets of those experiences.

Do you think your style has changed? And do you think there’s certain investments or events that have been very seminal in sort of changing your outlook or your style?

Absolutely. So I entered the venture scene as a failed entrepreneur slash postdoctoral scholar, trying to pay his bills in Los Angeles, and I observed so many extremely talented scientists and engineers either struggled to raise money or succeed in raising money, but struggled to.

Generates an outcome that made money for their investors or for themselves, and I came to appreciate how technology in itself isn’t a value creator.

I entered venture at a time when there was a investment strategy or an arbitrage opportunity that was starting to go away back in the late 90s, early 2000s.

It feels like as if there was money to be made by simply licensing technology out of the university or out of a lab and wrapping it into a company. And there were buyers of that for something on the order of, let’s say, the low hundreds of millions of dollars.

That was a time when that technology was not as complex when taking a technology to a product didn’t cost as much and didn’t take as much time. And so there were far more willing buyers. 

Those times have changed. Right now, you have to cut a larger check to get modest ownership in a company that needs to raise a significant amount of money. And there won’t be an outcome that would be interesting for you unless it’s in the billions of dollars.

And so that has therefore changed the emphasis from technology to business and technology simply being a means to an end.

So what I spend a lot of time looking for is not just great assets, but thinking about how these assets, how these technologies can yield an unfair competitive advantage for a business that would then generate margin, that can be protected over time, that will grow and generate a great outcome for us. 

That’s a good articulation. Thank you.

Well, let me point out a couple of things in your background and maybe you can tell me more about it.

So you have a strong background in autonomous vehicles and you invested in Zook’s extremely early and they’ve been a lot in the news. And Amazon is about to launch something.

Can you tell me more about autonomy and what you know, what sort of services do you think are going to be built on top of autonomous vehicles and exciting things you think might might we might see in the next decade or so?

I view autonomy as a feature and as a component of different types of businesses and products. If you look at autonomy as a components or a feature in cars that people purchase at dealerships than it is a convenience feature, it makes the car as yet another luxury feature. And as you can see, with Tesla and GM Super Cruze, we’re already seeing some of that today. As it relates to ride sharing.

It has an impact on unit economics. So ridesharing companies live and die by dollars per passenger mile driven. And so if autonomy can reduce their cost as it relates to dollars per passenger mile driven and increased their margins, then it’s obviously very meaningful for them. So these are two very different ways of looking at autonomy. In the case of a automotive company, in the case of Ford, they’re trying to see how they can minimize what it costs to add a feature into their vehicle and maximize how much a consumer would pay for that feature in their vehicle.

That’s very different from how a Uber and Lyft would look at autonomy. And then there’s a question for the Amazons to Shopify as the door dashes the insect parts of the world who are currently using.

People, so, unfortunately, when people use the word autonomy, they they they imply it’s something monolithic. Whereas I view it as a feature. I view it as a capability that has very different implications and very different applications.

And the perspective varies widely for these different applications.

So in the case of of more the ride sharing type networks, do you think do you think that, you know, fast forward a number of years, do you think that we’ll see different services or sort of different ways of interacting with these companies? Because, you know, I’ll go to sleep in my autonomous car and wake up in Vegas or something, and they’ll just be different use cases. So I expect different use cases across different types of companies and there will be new companies.

So an Uber and Lyft today, they sell passenger miles their product to you as a passenger mile and is probably more amenable to a trip from, you know, your your apartment to the coffee shop for a one hour meeting and then a trip from that coffee shop to another coffee shop for another one hour meeting, which is primarily my use case for for ride sharing. Now that, of course, Uber and Lyft may also be able to offer the product where you go to sleep and you wake up in Las Vegas.

But that may not be as easy for an Uber and Lyft to offer than let’s say another company is a new startup to offer that specializes in that kind of experience, that kind of product, and going out for that kind of customer. So there’s going to be, to answer your question, a summary answer is that there’s going to be significant impact to existing companies that use autonomy as a as a tool.

And then there’s going to be new companies that are going to be made possible and created through these technologies that are now available.

I think a lot of what you’re investing in, you know, feats of engineering captures people’s imaginations in good ways and bad ways.

And I was a no, it does. I mean, like, I you know, I has a lot of potential, but it’s scary to a lot of people. And so, you know, how how do you address that sort of aspect when people say, look, you know, our jobs are going away because of automation and that sort of thing?

That’s a very good question. So my response to that specific question on AI and jobs is that if you look at economies that have not embraced automation, it turns out that those economies also suffer from the most unemployment. That’s just basic statistics, whereas economies and countries that have adopted have adopted automation also have the lowest unemployment. And so it’s my expectation that that trend will continue where more automation, yes, it does eliminate some jobs, but usually the jobs that are eliminated are jobs that are very difficult to build and it creates other jobs that are easier to fill and are more attractive.

And you end up with a net effect of more consumption because the unit economics of whatever you’re producing or whatever you’re offering goes down. And as you’ve seen, whenever costs goes down, refrigerators got cheaper. Guess what? People have two or three refrigerators in their house. TVs got cheaper. Well, before we used to sit in the living room and fight over the remote control. And now there is a TV in every room and pretty soon every bathroom.

So with the when the cost comes down, there’s going to be more consumption. And that goes back to the virtuous cycle conversation we’re having earlier, which leads to more jobs, more demand and then more jobs. So it’s my expectation that more automation will create more jobs. But there is a there is a catch to this. And the catch is that there needs to be education that’s broadly available so that the existing workforce will be positioned to take advantage of or to take on these new jobs.

And so I think the onus is on our educational institutions, our governments, to make sure that. These and our existing employers, by the way, so the employer who is adopting this technology, I think the buck stops with them for them to make sure that their existing workforce is going to be trained, to be able to have or take on these new jobs that are now available through automation, through technology. So a line worker, for example, who used to put life and limb at risk to put a piece of sheet metal and a 20 ton press is now operating a robot that does that well.

They should be educated to be able to operate and debug that robot in the event the robot fails.

What about a more vague fear but that, you know, artificial intelligence run wild and, you know, we’re going to be controlled by that sort of the singularity?

Is here any any pieces of that that that ring true to you are interesting to you?

Well, I hate to break it up, break it to you, but I feel like we’re already being controlled by A.I. I feel like a lot of us are already being controlled. Our happiness or our interests are very much controlled by the A.I. that serves up Facebook and Instagram feeds. I feel like a lot of our behavior, a lot of our choices are very much driven by the A.I. that results in the Google search results that we see. And so I feel like that that day has already come.

We are already controlled by AI. Now, at least today we feel like Zipora at liberty to control our own destiny and we don’t feel like we’re under the iron grip of killer robots.

Will that day come? I highly doubt it, but I am more concerned about where we’re going today with a guy with where A.I. is and the way it impacts us today. So I wouldn’t worry too much about the Armageddon, about Terminator is what is it?

What’s the word that that day, judgment day, I would be less concerned about judgment day in twenty thirty and more concerned about twenty twenty where A.I. has level of impact than it does on us.

I think a lot of it is in our control as to how we manage that and how we consume social media, how we consume the news and how we go about finding information and making choices interesting our own values.

So you’re saying a lot of that is on us, the individual consumer user, as opposed to it’s on us sort of structuring our society differently.

I think it’s still early, I think. It’s a combination of both. 

If there are systems in place where it becomes or there is a requirement that when you type in a search, there should be some way of of you being told how much of the search result is influenced by someone. For example, the fact that the search results, if I type in tires, tires for a BMW and the search in the first search results is Continental.

But I want to know, who was it that that that that had? How did that search result become and why? Why is Continental first? 

So I think a level of disclosure coupled with consumer education, I think will address a lot of these concerns.


The disclosure and we saw that, for example, with the markets back during the the Great Depression and with the advent of securities laws, you know, you can still to this day be duped into buying stocks, but it’s more difficult for someone to dupe you from a legal perspective. And people today are far more educated today than they were, for example, you know, versus those people that were buying railroad stocks in the late eighteen hundreds.

OK, so let me take this in the positive direction then, which is I think you talked about, you know, automation is necessary for us in a manufacturing sense to to stay competitive. What about much more on the the desk worker and how how does automation, you know, potentially improve our lives? I don’t know if you’ve heard Elon Musk talking about meat flaps and we’re all made out of meat, but like, how do we improve our own sort of intellectual?

How do you know what I’m talking about?

Yeah. So I’m just as excited about. Automation in factories and cars, as they have about automation for otherwise mundane desk or White-Collar tasks, and if you look at a lot of the great venture outcomes and the recent years, companies like snowflake companies like data breaks. These are very much automation companies. They call themselves data companies that are referred to as data companies. But they are reducing a lot of the work that someone has to sit down and and and manually do so.

And I expect them to be the first of many companies that we see that may be labeling themselves as something else, but they’re very much automating tasks. And that’s very much been the role of software from the advent of software. An algorithm is a sequence of steps that’s used to automate a process. And so.

Anywhere there’s software, there’s automation 

One of the reasons why I decided to become an entrepreneur and later go into venture as an electrical engineer was because I figured that my own job, which was sizing transistors and putting them together into a circuit to build an amplifier or to build a logic gate or to build a buffer or a filter that perform certain tasks, I thought to myself, man, this is going to get automated and I’ve seen it before, and other tasks of engineering and engineering.

There are people that walked around with slide rulers doing calculations that was completely obviated by software in the 80s and 90s. And so it’s my expectation that there going to be a more rapid pace of the automation of these types of technical jobs. And the onus, again, is on the educators and the employers to keep people educated so that they be ready for the next generation of jobs.

I think the system of universities taking you in for four years, charging you tens now hundreds of thousands of dollars and putting you out there for the rest of your life, I think is an anachronism. 

But why is it that our educational system can get away with you not getting a job or you’re not being able to pay your student debt that a few years after graduating? So it’s my expectation that’s going to be more there’s going to be more accountability on behalf of the educators and on behalf of the employers.

And that’s going to help our workforce be constantly ready for those new jobs that are being generated.

Your LinkedIn says San Francisco because Lux is actually based in Menlo. It says Bay Area. Right.

Have you always been in L.A. since you joined Lux?

The office is in Menlo Park before covid. I would be up there Monday through Thursday and some weekends and since covid, I doubt that back to every other week because obviously there aren’t as many in-person meetings going on. However, I do expect now that I’m married, I do expect to to spend more time in L.A. I have a couple of boards that I sit on down here.

The company is doing absolutely fantastic.

I think that there is no shortage of amazing talent in L.A. There’s no shortage of amazing leadership and entrepreneurial horsepower in L.A.. 

Anything else on Lux that I should make sure to take cover here? I feel like you covered it. Is there any other anything else that you’re curious about or you think your your listeners will be curious about? Yeah, I’m curious. If you were to start a company, what company would you start?

Good question. If I were to start a company. So there there’s a few areas that I’ve been tracking along the lines of. How it’s manufactured and the amount of automation and economies that are being brought to bear that would make it possible for a startup to compete in areas that until now have been dominated by large incumbents.

So there’s a couple that I’ve been thinking about in the semiconductor space and also in the in the in the manufacturing space. So it’s still early on. But again, those are the kind of motivators that would make me excited about about the next kind of Lux company.

Interesting. Is that the sort of thing that you would incubate yourself? So if I’m fortunate enough, my my preference is to find a person who would carry it, and in the case of subspace, for example, which is an L.A. company.

You can have a 50 gigabit per second connection cable at home, but still have an awful connection between, you know, Los Angeles and Austin, Texas. And I’m sure a lot of you have mean I’m sure you’ve experienced this with with video conferencing before covid. And it was it was an idea that was that was that was on my mind.

I met an amazing founder of Tofail who is in the process of selling his previous company, who had a track record of building companies in the networking and telecom space. And I was delighted to find him and to partner with him and for this to be his baby, his vision to carry forward.

So if I am fortunate enough to find a founder that is willing to take this on, that I have one hundred percent for partnering up with them. But if I’m less fortunate and can’t find someone, then it’s something that we as a firm would do in-house.

Well, thank you so much for for coming on the show today. And, you know, I wish you a lot of luck in accelerating humanity towards a brighter future. Keep doing that.

That’s the plan. Thank you. Minnie. 

I didn’t ask you about diversity in deep tech, which is something I think about. 

So I think the the the root of that is just not having about there not being as many females in a lot of these technical fields. 

I think some of it also has to do with demographics. What’s not just it’s not just gender, but it’s also demographics. So I feel like as if a lot of these deep technical fields, at least in the undergraduate realm, are or graduate are dominated by foreign students. And it just so happens that it’s more likely that a male would leave their country and come out to study abroad.

For example, when I was at Berkeley undergrad. The vast majority of my classmates were from Asia and South Asia. 

And so I think it’s not just gender, but it’s also just the demographic of the foreign students and then foreign students being predominantly male. Hmm. Yeah. And why is that? I mean, that goes back to the culture as the cultures of origin.