This Week in Startups - Zymergen CEO Josh Hoffman on Biofacturing PLUS Founder U Startup Math | E1231

Episode Date: June 11, 2021

Jason opens with a Founder University segment on "Startup Math" (1:48), then interviews Zymergen CEO Josh Hoffman on Biofacturing (16:45), how DNA sequencing & machine learning are enabling materials science advancements (26:35), what human capital is needed for the industry (42:53) & more. Pod Notes: http://bit.ly/1231notes

Hey, everybody. We've got a great show for you today. I'm going to interview the CEO of Zimergen, Josh Hoffman, on the ins and outs of synthetic biology, including all these recent breakthroughs and what the next couple of decades are going to look like. It's a fascinating discussion. Basically, the future is here. I mean, you start referencing science fiction so many times when you get into synthetic biology. It's kind of crazy. But before we get to that, I wanted to try a new format where I break down a topic that a founder should be familiar with. We're calling these Founder University new. We have this Founder University Two-Day program. Now we're breaking that down into a curriculum and I'm doing little modules here on the podcast to test them. And I want your feedback. You know how to get in touch with me, Jason at Gallaghanis.com for the rest of my life or DME at Jason

on Twitter. And tell me what you think of these fun modules, Founder University, new modules. I'm going to cover back of the envelope math today or what I call startup math and how you can wow investors and be a better strategist just by knowing all your numbers cold and having them all

right in your head and the ability to do back of the envelope math. Stick with us. This week in Startups is brought to you by Secureframe helps hundreds of companies get enterprise ready by streamlining SOC2 compliance in weeks, not months. Get $2,000 off your first year by going to secureframe.com slash offer slash twist and Embroker. The Embroker Startup Insurance Program helps startup secure. the most important types of insurance at a lower cost and with less hassle. Save up to 20% off

of traditional insurance today at imbroker.com slash twist. While you're there, get an extra 10% off by using offer code twist. Founders should be able to do back on the envelope math on important metrics at any time. This makes you, number one, sound really credible at your investors and two, it's going to help you from running out of money. So on episode, episode 663 of this week in startups, I had Jason Lemkin on, and I said this to him in our conversation. What I've learned is that if the numbers are wrong, ridiculously stupid, there's something wrong and how the founder is thinking about the business. If it's really wrong, there are numbers that aren't even remotely sane, and it shows the founders don't understand how the business works,

and that's scary as an investor. And so I'm able to do this little party trick, which I talked about on page 149 of my book Angel. When a founder is talking, I wrote this in the book, I pay attention and I write down the numbers, and I do back of the envelope math to understand where a company stands. When founders often can't do this, they unexpectedly run out of money. Okay, so here's an example of me doing the party trick.

I ask people what their monthly reoccurring revenue is, which means just how much money the company is bringing in. So com.com, when I invested, had, you know, they were selling a thousand apps a month before subscriptions for $10 each. They're making $10,000. Then I asked them how many full-time employees they have, and then maybe what do they spend on their headcount?

But I know in today's money, you know, 10K a month per employee, all in, office space, remote, work, whatever, pretty good benchmark. So if you were calm at that time and you had three full-time employees, you know, in a seed stage company could be somewhere between 7 and 10,000 a month. So it's 20 or 30,000 a month. And they're making 10,000. So if they're making 20 and they're making 10, now I know their burn is 10K. And then I asked them, well, how much money have you raised?

And they say, well, we raised $100,000. Well, if they're burning 10 and they have 100,000 in the bank, they have 10 months of runway. And so I can really quickly understand how much runway the company has based on how much money they've made just by walking through these examples. So once again, how much money you're making per month? Okay, how many full-time employees do you have? Okay, take those two numbers. Full-time employees are typically 90% of the spend.

So if you have three employees, at 10K,000, that's 30%. Yeah, you add a 10% on top of that, 20%. 20% on top of that. Maybe they're spending 35 a month. Okay, they're making 20 a month. They're burning 15. They have 100k and the bank. They have roughly six, seven months left. And boy, I'm accurate almost all the time with this. And the reason this is important is many founders don't know their exact revenue numbers. You should always know the last three months of revenue for your company. You should also know what your payroll is every two weeks. And then you have 26 payroll periods. So if you know what your payroll is per pay period. Let's say your payroll per pay period. You have three people and they're getting paid 10,000 a month each all in. Okay, pretty simple. That means you're spending 15K per pay period, right? 30k divided by two. So you're spending 15K per pay period. Okay, how many week, how many pay periods are a year? There's 26. Okay, 15K, 26, okay, you start to get to that 350, 400K number, right? You can really start to understand what your yearly payroll loss. And if you're paying people 120 all in,

times that by three, and you're at $30060 for the year. You should be able to build these models up and down. And you see how I can just very simply talk about these? Then you could say, well, we're going to add for headcount. So we're going to go from $30k a month, approximately to $70k, but we're growing 20% a month. So every three to four months, we plan on doubling revenue, we're going to grow about 20. If you said, we're going to grow 25% a month. That means every three months you double your revenue. So you're going to go in quarter one from 10K, quarter two to 20k, quarter three to 40k and quarter four to 80k. If you're in fact growing 25% per month, because you know the rule of 72.

If you were growing by, let's say, 25% per month, divide 25 into 72 and you get 2.7 or so. Basically, it means in 2.7 time periods, you would double. So if you were growing 25% week over week, week 1, 10,000, week 2, 12, 5, then 25% of 12, boom, you can just do the math. In three weeks, you will have doubled. That's how the rule of 72 works. Here's just some other points about you understanding your metrics. You can say the word approximately or on average and then give a number. What a lot of founders do is they're afraid to give a number.

And they will give the number last year or the number they think it will be next year. When people ask you what's your cack, they expect you to say, what's your cack right now? And let's have a discussion about that. but founders get flustered and they don't want to say it costs us $100 per user to buy this product that's $60 a year. It's better to own that. Our CAQ is currently at $100 per customer, which is obviously greater than the $60 a year for our subscription. But six months ago, it was $500. So we've gone from $500 down to $100. We think in the next three months we'll get it down to 50 and then three months after that we'll get it down to $25. So it's in process. But always give

a number. The most annoying thing as an investor or as a student, CEO or founder is when people tell you it depends. I can guess on it depends. Telling people it depends is the most annoying thing you can do as a founder. You're the leader. If the leader says it depends or it's a range, that's not helpful. Plant a flag. As a founder, just tell people your best estimate. You could say guesstimate if it really, if it really is more like a guess. But being able to do back of the envelope math is critically important. So if you ask me about inside.com, I can tell you, well, a newsletter takes about $150 a day to publish. So if you were to publish it 100 times a year, twice a week, that's $300 a week, which is $15,000 a year. Now if you go to weekly, well, it's $150 a day,

$750 a week. And $750 a week, 100 weeks would be $75,000. And half of that, 50 weeks a year would be about 37-5, I'll round it up to 40. So it's about $40,000 a year. If you were to pay a freelancer, $125 to do a newsletter every day, which takes three hours, which is $40 an hour, and we pay somebody $25 for an hour of editing it and publishing. Very simple. You see how easily and elegantly I can explain that. And if we get one advertiser to do a 50k ad buy, then the newsletters in the black, and we can pay a writer upwards of $75,000 a year to do, two newsletters or 37-5 to do one newsletter and have a 15% margin business, 20% margin business. This kind of easy crisp back of the envelope numbers makes you credible.

And when you don't know your numbers or you refuse to say them or the investor has to ask you three times, every single time you make them into Colombo, the detective from the 70s and 80s show, your credibility is going down. You need to just, own the numbers. And this is a very important point. Don't apologize for your numbers. A lot of times young founders and nascent companies will over and over again not want to say how modest their numbers are. Your numbers being modest is our opportunity to invest at a low valuation. That's what angel investors want. We want to meet you when you have but 10 customers paying $1,000 a month each. That's the opportunity. I don't want to invest in your company when you have 10,000 customers.

paying you $100 and you have $12 million in revenue, because now your company's worth $250 million or $500 million. I want to get you earlier than that. So don't apologize. Just own the numbers. And small numbers can grow. So when you show us small numbers, we think in our mind, oh, I've seen this movie before. I remember when Uber was in one city with one product.

And then they had four products in 400 cities and they had 1,600 business lines based on geography and Uber Eats and UberPool and Uber Black and UberX, right? This is how our minds work. We think about how can we double, triple, triple revenue in those early years, triple, triple, triple, double, double, double, right? You started at 100K in revenue, got to 400, then you got to 1.2, then you got to 3.6, three triple ups over four years, right? Then we saw you double it, you know, or two and a half times it, and then get to, you know, seven or eight million and then get to 20 million. This is what we're looking for. one of the things I get over and over again is when should I hire someone else to do a certain job, the work at the company, right? So if you're spending X hours of your time on Project X, Y, and Z, it's pretty simple. You just do a formula. Your salary times 1.3. So if you were making, you know, $70,000, well, 1.3 of that is going to be an extra 20K. So we'll just round it up to say $100,000.

Now you take $100,000 and you divide it by 2,000 hours. Where do they get 2,000 hours from? That's how many hours people work. 40 hours a week times 50 weeks, 2,000 hours. Now, some people at a startup might work 50, 60 hours a week, 50 hours a week, and then that's how you get those employees who, you know, do 2,500 hours a year. But this is how lawyers look at their time. I'm going to put 2,000 hours in.

I'm going to get paid $800 an hour. I'm going to make $1.6 million. I'm putting 2,000 hours in. I bill $2,000 at $400. I'm worth $800,000, right? So that's basically how you can get the cost of your project. Now, why 1.3? That's your all-in cost.

You have to pay taxes, and you have to pay for computers, and you have to pay for benefits. So you very simply can look at something. I'm putting 10 hours a week into doing customer support. Okay, 10 hours a week of your time as a CEO, if you're all in cost is $100,000, you're taking some modest raw of $6,000 a month, you're making $70K, whatever.

We rounded up to $1,000, $50 bucks an hour, $50 an hour, 10 hours. So you're spending $500 a week doing customer support. Okay. And what if you found somebody in Manila or an expat or somebody in Canada or somebody in another region, South America, who's willing to work for $500 a week? Now you've got a full-time person. Okay. So now you can start making this tradeoff. Would the full-time person at $26,000 a year, this 10 hours, would that be a better hire for you?

you. Okay, can you find that person? Can you train that person? Now you can start investing. Or maybe that person does, you know, if that person takes twice as long as you to do something, instead of you doing it in 10 hours because you're so good, it takes them 20. Well, now you still have 20 more hours for them to do other stuff, right? Maybe they could write your customer support manual. Maybe they could help with, you know, accounting or something. So this is how you should be cutthroat and think about your own time. Typically when you're starting a company, you might have more of your time available than money and resources, but then once you get funded, you have to be cutthroat about this and radically delegate stuff. This is why in my companies,

I like to hire young people starting their careers off who can quickly learn and who are super motivated. When I have somebody like Presh or Marine on my team as an example, who became associates, they started just doing anything operationally we needed done. And I could look at them and say, you know what, I've got senior people working on this. If I can get that project off this senior person's plate, a Jackie or nationally, a managing director's plate, and get pressure marine to do it? Okay, well, they're going to cost less. And then that enables my senior people to then go do something else with their time.

And this is why operational excellence can be such a catalyst in a second order way for your company. When you hire somebody who is graded operations, what they typically do is they come into an organization. And if you have 10 people in the organization, there might be six people doing operational stuff 20% of their time, and then your new operations person takes 20% off of six people's time, and then they do it twice as efficiently. And now those people can do their core job better. So some sales executive is, you know, doing docu sign management. Some podcast producer is doing

customer support management, tickets, and ad operations. You get the idea. You can over time hire great operations people that then free up the specialists in your company. And as your company grows, you're going to get more specialists to do that work. So just some back of the envelope math for you. Don't be intimidated by math. You don't need to know geometry. You don't need to know advanced math. This is back of the envelope math.

Just try to do basic numbers in your head so you understand ballpark how your business is going. And then every week, you should be checking every expense that comes into your company. You should be checking, you know, especially for a nascent company. Obviously, if you've got a $10 million or $100 million company, you'll have other people doing this. But you should know your bank balance every week. You should know how much cash you have in hand. You should know how many outstanding bills you have, what your costs are, what your payroll are. Just know those top level numbers.

So you have a mental model and you have the heuristics in your brain to do back of the envelope math when you talk to investors, when you do strategy meetings, or you're just trying to conceive of your business. And this is why plans and models are critically important. enjoyed and had fun with this new founder university content. Okay, let's get back to the program. Look, you probably keep hearing about SOC2 compliance and you might think, is this really relevant to me? Well, if you're targeting any large enterprise as a customer, there are all sorts of data privacy and security measures that you need to have buttoned up to close those deals. And you don't want your engineers taking time out to do this stuff. And you definitely don't want to hire a third party auditor. No joke, getting SOC2 compliant can take months.

And it costs a ton. That's where Secureframe comes in. Secureframe helps hundreds of companies get enterprise ready by streamlining SOC2 compliance in weeks, not months. And they monitor over 40 services, including AWS, GCP, and Azure. Secureframe will continually collect audit evidence, run security awareness training, managed vendors, infrastructure, and more all automatically. On average, Secure Frame customers save 50% on their audit costs and hundreds of hours. of time. Their team of compliance experts and auditors are happy to help answer any questions you might have and give you advice. When you think of compliance, don't get stressed out. Just like a

Secureframe, streamlined, affordable, and hassle-free. Here's your call to action. Secureframe is offering $2,000 off the first year for Twist listeners. That's right, $2,000 off your first year at at secureframe.com slash offer slash twist. That's secureframe.com slash offer slash twist for $2,000 off. Today on the program, we're going to talk about the science of bio factoring. And this is a new science to create new products and a range of industries and synthetic biology. We've been talking about this a whole bunch.

And we have one of the pioneers in the space. Josh Hoffman is with us. He is the CEO and co-founder of Zimergen. Welcome to the program, Josh. Thanks for having me. So explain to the audience who are neophytes what exactly you're doing at Zimogen and why it's important. Yeah. So if you think about it, the material world, all the stuff we touch, all the stuff we use, the coating on our glasses, the texture of our shampoo and toothpaste, it's all made from materials that are derived from petrochemicals. That has a couple of problems. The first is that they're all.

all derived from six or seven base petrochemicals. And human innovation, you do anything for 150 years as we've been doing this, it's going to slow down, right? So we've slowed innovation in the material world has slowed substantially. And so our products aren't as good, right? They're not increasing in quality in the way that they once did. And the means of production are frankly torching the planet, right? Climate change is the existential challenge of our time that's directly related to our pulling hydrocarbons out of the ground. And so what we're doing is taking some of the most advanced science and technology on the planet and put it together in a way that allows us to make better products and in a better

way. And petrochemicals, so people are clear, these are basically things that come from petroleum, natural gas, and obviously not good for the environment. So you're making things out of what? So what we're doing is where, so the way it works conventionally is you get oil, get oil or gas, And you heat it up and you do what's called crack it. It fractionates. It separates into a bunch of different layers.

Then you crack it into its kind of Lego-like components. And then you put these back together again. And that's very powerful, but very expensive and complicated, energetically very expensive, horrible for the environment. What we're doing is taking advantage of nature, right? Nature has engineered these little microbes, yeast and bacteria, to eat a carbon source, sugar or plastic, and to inside its body convert it into something useful.

Now, many people are familiar with this process through the making of bread or the brewing of beer and making it wine, right? Fermentation. But that same process works today to make products as diverse as penicillin or statins or citric acid or crop protection agents. And what we're doing is we've created a platform that allows us to, in a general purpose way, identify useful molecules out in the world. and then program the microbes to make these and not just make them, but make them at scale and it costs that work in our economy. And this happens. Microbes are put in some kind of a fermentation tank.

Exactly. And then you build something out of them. The first product you've created is for iPhone screens or smartphone screens? It's an optical film for display screens. That's right. What is that the optimum? What does that film do? What is the purpose?

of it. It allows you to have clearer screens that adhere better together and especially in novel use cases like flexible screens. But basically it's a modern display is a, it's a laminated. It's a stack

of a bunch of materials glued together. And this one happens to be very, very optically it's very clear. So you don't even notice it's there. And so you can make better touch screens. I could get into the chemistry of that, but it's probably not for your audience. And so when you make those better screens and that would have been something that was made from petrochemicals previously. So billions of phones are made.

Billions of screens are no longer made with petrochemicals. And that's better for the environment. Is it cheaper or is it neutral or is it eventually going to be cheaper? It depends on what you're measuring. On a per square meter of film basis, it's probably neutral. On a dollars per unit of screen quality basis, it's much cheaper. Right. So this is we're not selling our products do not compete because they're more environmentally sustainable. We sell better products. We sell better products and that happen to be better for the planet. And this has been going, you've been working on this since 2013 or 2015. It was kind of in stealth for a while. I think a lot of people know that you went public very early.

I'm not sure if it was through a SPAC. I'm assuming it was. No. Regular API. Regular IPO. And so this is a very, promising technology, you've raised over, I think, a billion dollars in the company to date. Yep.

From the seed round back in 2013 all the way up until this $500 million plus IPO. The company, though, is very nascent in terms of revenue, but you've been doing deep tech for a while. Why go public now? I'm just curious as to the thinking. Most people say, hey, you know, stay private longer, get to a billion or 10 billion in revenue. You went public with, I think, maybe 15 million in revenue or something. Sort of in that zone. Little more than that, but that's in the zone.

But I think, look, I think that this is a difference between the life sciences world and the pure tech world, right? There's a long history in life sciences, in biotech, and life sciences tools and diagnostics of companies being able to go public and with an investor base that understands companies. And frankly, in that world, we're a little late. We're right in that process right now. Lots of companies go public when they're really very, very much as a development stage company. We're just rounding the turn from what's called development stage to commercial stage.

And so in that context, it's pretty ir. I mean, we're kind of there, thereabouts. Are these manufacturing plants going to be here in the United States, or is just something that it will be around the world or in China? I mean, we watched manufacturing move to China and, you know, to other places. People have talked a little bit about the dependency on other countries, perhaps even, you know, a country that is not a democracy. It might be authoritarian, et cetera. is this going to move these factories here in the United States? Are you doing this work here or does it have to be done in a low cost place like, say, China, et cetera?

So our first product, we have supply chains in two geographies. You want some geographic resilience. One of them is in Japan, but we're bringing up a supply chain in the United States as we speak. And the supply chain in the United States will be the larger of the two by in terms of volume and frankly the lower cost. And so the answer to your question is there's no reason it can't be in the United States. Got it. And so what will you be making next? What is the plan here? Is it going to be going towards electronics? Is it going to be going towards, you know, protecting, you know, doing bug spray and, you know, things you've mentioned before about preserving food? Where is this all going to wind up? Because it seems like from the description of what you're doing, you could do almost anything. So how does one pick? what they're going to do. Obviously, smartphone's a pretty great category to be in, but is there some sort of a product roadmap here that follows either the opportunity or what's

what the technology is currently capable of? Yeah. So in our S1 and in our roadshow, we disclosed a pipeline that has a product pipeline that has 11 products in it, and we disclosed specific launch dates for, I think, four of them. And so highly, just the product that's in the market now, The next product... That's the biofilm for the electronic displays, micro-lead-d-froids. It's one kind of... It's not for micro-LEDs. That's going to be our next product.

Next product will be another optical film, which we told the world we're going to put on bring out to the market next calendar year. And then the year after that, we'll have still another optical film, right? Because what we're able to do, it costs us, just to give you a sense, it costs us about a tenth of what it would cost to a traditional player to bring a new material to market, and we can do it about half the time. And so what that means is we're able to go after a much more precise product market fit than a traditional way.

Like, if I can bring something to market for 50 million bucks, it would take somebody else almost half a billion, I can really get the exact perfect product, right? I mean, you've seen this in technology in the pure technology industry. And so we've got a series of optical films coming out. And then in 2023, we also intend to launch a product, which will, as you indicated, have our first consumer care product. It'll be an insect repellent, a bug spray. But this is a category that's ripe for disruption because, you know,

you know, half the people I believe in the United States use deed every year and lots of people believe it's a neurotoxin, right? It's a terrible product. The consumer choice is horrible. And yet, I think we all appreciated in the last year or two, the value in insect-borne illness, right? It's a real thing. Yeah. And insects and mosquitoes are just in Texas this past weekend getting killed and somebody took out that off spray or whatever with the deatonin and they were spraying it everywhere and I start coughing. I was like, this stuff is killing people. Yeah, it feels like killing people. And yet, and yet you have a choice, right? I mean, I talk about, I use my poor sister's an example all the time. She lives in Georgia where I grew up. And she has two kids. And

she can either like spray them with stuff that's noxious and like makes them cry or have them deal with mosquito. I mean, it's a terrible consumer choice. And so we're able to partner with nature to create something new and new molecule that the EPA says, not us. It says as effective as deed at repelling it, but it's safer. And, that's, you know, that's a better product, right? What has led to this revolution of this being available now as opposed to 20 or 30 years ago? Is there some technology that enabled this? Why now? Yeah, so there's a bunch of technologies. I would say from our standpoint, there are probably three or four that are most important.

Number one is the rise of low-cost gene sequencing, right? the second is the development of a set of tools that have allowed you to edit the genome. Think of sequencing as transcribing. You're not really reading. You're transcribing, right? Gene editing tools allow you to set type, right? But again, you don't really know what you're doing, right? You don't know the words yet.

We're human understanding biology. I mean, Daphne Kohler, I know a friend of yours and has been a guest on the program. She talked about how complicated biology is. Biology is not a human invention. It's massively complicated and very badly understood. So we know to transcribe it and we know how to set type. What has then happened, what we've done is taking advantage of developments in large-scale cloud computing, a whole series of development on the pure technology side with things like

NoSQL databases, with advancements in machine learning that have allowed us to begin to understand maybe not exactly how to write, but how to write much better. And the convergence of those technologies has allowed us to be able to build the technology. stack that supports our business. So the sequencing of DNA and then being able to set type and maybe change it a little bit, then being able to take in all these sequences, put it into a database, and do some sort of analysis of it gives us some sort of insight into what could be changed or what's actually going on. Would that be a good way of describing it?

It gives you insight into what you might change that would have an effect that you want. Got it. That's not the same as saying you understand what's going on. Ah, so we actually know there's something going on, and then we kind of know an area where you might be able to make a change, and then you have to test to see if it actually made that intended change. Yeah, but the thing about machine learning that's beautiful is you try things where the human has no idea why it matters. In our strains, they have gone into scale. I'll just give you a sense. We find on average 40 to 60 percent of the genetic edits we've made defy human explanations.

even after the fact, and up to a third are in genes that have no known function. Wow. Right. And what we know is that it's easy enough to put what's called a pathway, the set of genes that allow you to program a microbe, to make the initial bits of the product. But, you know, when we put the initial pathway in, you know, on that phone screen, for example, it would be like $250,000 a phone screen, right? Like, there's just no business there.

And so when you're making the edits, that allow you to hit cost targets, even for very valuable products like optical film, you have to find ways of finding these parts the genome that defy humans can't even imagine. A third of the edits in genes that have no known annotation, we don't even know what it does. And so are we basically doing a process of elimination

or is the machine learning saying, we think there's something over here, let's try that next? Again, simplifying a little bit. The machine learning is saying, based on all the programs you've run before, if we try these thousand edits, the ROI on that portfolio and the thousand edit is going to be the highest portfolio, highest ROI in the portfolio. So we're not making point predictions.

We're making probabilistic predictions across large sets of builds. And this is why you need automation, right? Because if you don't have automation, you both can't do enough, you can't get enough throughput, and your data fidelity is not clean enough to deal with to pull the noise out of the signal. or separate dismalism. When you have a thesis that, hey, this change might result in the outcome we want, does it then go into this fermentation tank and we test it? And you have to do 1,000 fermentation tanks to get where you want to go.

And that's where the $50 million goes to develop it? Or is it all done in a simulation? No, no. It's not done simulation. You do have to test stuff in the lab. Right. So you'll say, I'm going to build 1,000 strains. You'll build 1,000 variants, right?

And again, this is not hypothesis-driven. It's a little bit more nuanced that. Sometimes it is, but mostly it's not hypothesis-driven. And then you have to grow the strains up in these little 96 well plates, which are, I don't know, yay big about the size of like the old 4x index card. I don't know if you remember those. Kids these days probably don't. No.

And each while it's got a little 2-mill deep well. It's about the size of the final little joint of your pinky. And then you'll grow them up. And then we'll actually use another set of machine learning to figure out which ones to evaluate performance because performance is super noisy. It's a classic like case of pulling signal from noise. And then those get put into fermentation tanks. Got it. And when you started this, you were going to be more of a service business and do this for other companies?

Or was it always the intention to just, hey, let's go right to market with products? It was always the case that we wanted to go to market with products. But we believed that there were two or three things we needed to do. We need to demonstrate that our technology, which is pretty different than anybody else's technology, we wanted to demonstrate that it was going to work especially for scale up. It's all well and good to develop a product, but if you can't put microbes into large-scale fermentation that are going to work at scale, and to give you a sense of scale, it's like a fermenter the size of an eight-story building, right?

Like, this is not, right? This is not pristine lab work, right? You have to show that it was going to work at scale. And we've got a machine learning driven system. where are we going to get the data? This is not data that exists. And so, oh, by the way, now we get to demonstrate for third parties that our platform works, and we own all the data we generated off of all the work that we did for that.

Got it. And so the bug spray comes next. And then was there a fourth product? Or did they get three optical films, right? Three optical films. Three optical films and a bug spray. And then there's additional products that we've talked about, but we haven't disclosed a release date. So, for example, we're working on a product that's a film former.

Film formers are the things that give your toothpaste and shampoo the texture, but they also clog the oceans. And so we're looking at biodegradable film formers. We're looking at biobased adhesives for electronic assembly that both are environmentally friendly, but also allow you to do assembly cheaper and faster, right, which has a lot of value if you think about how a phone is assembled. We've got ag products. We're working with a partner on a bug that's in-field trials that reduces the amount of nitrogen,

basically increases nitrogen fixation and reduces the amount of nitrogen that farmers have to put down, which is hugely both value-creating for the farmer, but also for society because nitrogen runoff basically has made a huge dead zone in the Gulf of Mexico. I mean, there's some other goodies, but that gives you a flavor the stuff we're working on. And so it will take a couple of years to do each of these products, Is that the sort of the path? Today it takes us about five years to go from product idea to being in the market. To give you a sense, right, though, that seems slow in the world of software.

You're like five years. Yeah. A billion-dollar companies get built in five years. But to give you a flavor, it takes 10 years for a traditional petrochemical player to bring a product to market. And we can do it half the time, and we've got a bunch of plans to reduce that. And so, yeah, half the time, a tenth of the cost. is there some tipping point that's going to happen when enough, enough trials have gone through machine learning that we're going to hit some sort of scale?

Because it seems like you're just scratching the surface here or society in this category is just scratching the service. Is there another tipping point that's coming? And what would that be or look like? Yeah. So this is a frequent topic of conversation among biologists who are taking something fermented that they can hold in their hands. hands and ingest. I eat beer. Sorry.

People really disagree about this. There are some people who think, I mean, this is a, the question you're asking a little bit is like, how close is the singularity? Because there are people who will say it's never going to come, right? The biology is so complicated that we're never going to have enough data.

And it doesn't obey. I mean, if you think about a modern, the computers are using to record this on, right? You can debug every single thing all the way down to the semiconductor is human designed and understood. And at every level, like, no matter, if you have a bug, you can't understand at one level, you just go one level below that. The entire thing is debuggable. Biology does not, is not like that. Biology has evolved over billions of years. And we have the barest understanding of the rules. And we have no ability to

monitor lots of what's going on. And so some people will say, well, in five years, we're going to have enough data to make it all work. And other people say it's never going to happen. I honestly don't know. It's a religious debate. It really is. Like, we're sort of getting to the creation of the human species. We're sort of tipping over into alien and Prometheus for people who've seen that movie of, you know, like, where did we come from? And how did this all manifest itself? Yeah. That's right. And that's even more a speculation for, you know, that's, that's where you take the fermented thing and you distill it. Yeah. Every startup needs business insurance, please get your business insurance tight. And you don't need to look any further

than my friends at him broker. If you don't have insurance, you basically failed the first step of running a company. Prices are 20% lower and you're going to get better coverage than incumbents when you use them broker. You can go from sign up to quote and purchase in just 10 minutes. It can take weeks when you use the large, slow incumbents. The process is so transparent. There's no opaque pricing. You're not going to get jerked around like on these other, you know, incumbents. I'm telling you, I've been through this before. And there are four types of insurance you need to know about. Cyber insurance. Hacking. Everybody gets hacked. If you have cyber insurance, you're protected. D&O insurance, directors and officers. This means if somebody does something

dumb in your company, you're bored or the management team has attorneys to protect them. Errors and omission. This is super important. When you're scaling and you have major customers using your platform, they're going to ask you, do you have, you know, it means if you make mistakes, you're covered. And finally, EPL. Sadly, this is very critical. Employment practices liability. This covers harassment and wrongful termination and other type of employee issues. And there's no better place to get it taken care of than with my friends at Mbroker. To instantly buy custom-built insurance just for startups, go to imbroker.com slash twist. Let me spell this for you. E-M-B-R-O-K-E-R.com slash twist. And while you're there, you're going to get an extra 10% off using the offer code.

You know it. Twist. TWIST. What are we going to figure out in our lifetime? What is, you know, obviously discoverable and implementable? Obviously, your company is, you know, getting us off petrol chemicals. But what do you think the entire industry is going to figure out over the coming years? You know, we obviously have seen, you know, you know, MRNA sequencing and we're seeing new Alzheimer's drugs announced today. What does this next decade look like in terms of material science? Are we going to just recreate every, be able to recreate most of the stuff that we build in our lives? So I think, I don't know if it's going to happen in the next decade. You know, what's the old quote?

It's easy to, you know, you always overestimate progress in the short term and underestimated in the long term. But I think that in our lifetimes, which obviously, you know, touchwood will be longer than a decade, I do think we're going to have a fundamental reinvention of material science. I think you're going to have a fundamental reinvention of the things that humans can do with respect to the material world. There was this profound reinvention of society, the most profound reinvention from 1860 to 1960. And it was a whole bunch of technologies that engaged in our physical world. The oil and gas was one.

The petrochemical revolution was one. internal combustion engines, the modern network electric grid, the telegraph, telephone and telegraph, and the modern sanitation, right? You put those together, and human existence changes in much more profound ways, right? Somebody in 1960 could not go back to 1860 and know what to do. Somebody today could easily go back to 1960. Life would, there'd be fewer seatbelts and, you know, more cigarettes, and that's kind of it. Yeah, but you can still have cities. You can still have society, yeah. That's right. Whereas 1860 and 1960, I believe that the kind of technologies we're working on have the potential to create another industrial revolution

of the same level of impact. I mean, we live in a material world, right? I know that lots of people who listen, you know, most of the startups you cover are purely digital and life has definitely changed because of it, but we still exist in a material world. So I think we're going to reshape material science. I do think you're seeing, you know, reshaping of medicine. I think you're going to see a reshaping of ag. I think you're going to see huge advances in the human interaction in the material world.

in the next time, you know, in my lifetime. We are talking about a lot of the things that we use as consumable products, whether it's toothpaste, shampoo, bug spray, et cetera, or something we hold in our hands. Do we think this is going to also impact construction? We have this nanotechnology revolution that kind of never happened, where we thought that buildings would be made just with nanotubes and it would be incredibly strong and, you know, we would be able to put them up for a fraction of the price. never really happened.

Do we see this affecting construction materials and eventually buildings? The answer is yes, but at least the initial use case is the stuff that I think is closest are not necessarily things that would be obvious. So plywood, super widely used construction material.

It's glued together, right? Those glues are problematic. There are lots of opportunities to improve those adhesives with biobased adhesives. Foams, right? there's a lot of opportunity to use bio-based materials to create better foams. The glazing, the coating, take those optical films I talked about. You can imagine tweaking them, and now you can put them on the windows,

and you've got smart windows in a better way. So these are all ways that, I mean, I've measured in years. There are people at startups that I'm aware of looking at doing stuff around concrete. But I think the thing that's amazing about humanity is, like, this is just stuff I know of and can imagine. the world is far more imaginative than like one guy sitting in an office. So yes is the short answer your question. Yeah. And being able to produce things that are stronger and lighter or less damaging, all of these things,

the implication could be more sustainability on Earth and population growth becomes less of an issue and climate change becomes less of an issue. All these second order and third order challenges we've had with this dependency on oil and fossil fuel goes away. and these things are all net neutral? I mean, look, I think that, I think we're going to see populations decline, right? I think that demography is very, you know, demographics are very predictable over a long period of time,

and I think when you're seeing in Asia where you're seeing fall off in birth rates, I think we are going to see global population rates decline. I think Africa's the wildcard there. So I don't think anything we're doing there's going to affect that one way or the other. I mean, we certainly want to take a bite out of climate change, and we want to take a bite out of the sustainability challenge that comes with many of the materials that are made from petrochemistry. I certainly hope that we're going to do something that isn't just net neutral there. But as you point out, there are often unintended second order effects.

What are the degrees and the career path here to work in this space? And are we producing enough of those talented people? If you're a young person in college right now, what are you studying to then go work at your company? Yeah, I mean, so the people, if you, you know, the standard answer is a mix of a hard science, biology, or chemistry, and then computer science, right? So that crossover specifically, having both over those skill sets.

That crossover is incredibly valuable. Now that said, we've built a company with best of breed folks where we've tried to create the culture that allows them to talk or at least fight constructively with each other. Because so it's so hard to be an expert in both. These are both very deep technical domains. It's really hard to find somebody who's an expert in both. But what I would say is more than that, the people who are most successful, the people have shown an ability to learn new things and really learn new things because this is a new industry we're building. Our head of automation was a guy who was a physics PhD student doing astrophysics, who then was a professional photographer who's now like a senior automator.

Like his ability to learn new things to be curious and creative and not scared of technical domains he didn't know and to be able to apply the lessons from one place and another, that's far more valuable than any individual domain expertise. And when we think about these new products, there are incumbents in the chemical space. Are they in favor of this? I know you have a partner with a Japanese chemical company to do these screens, if I'm correct? The films, yeah.

Yeah. So how are they looking at this? Are they interested in this space or is it just so far out of their wheelhouse that they're not even paying attention to it? Yeah. What I would say is the closer you are, you can think about the chemical material space as having three kinds of companies. You can have people who make base chemicals. Those are like the six or seven core base hydrocarbon, like ethylene, benzene, things like that. And you can think about the specialty chemicals companies that take those and make something more expensive and higher margin.

And then the materials companies that ship final product. The closer you are to a materials company, the more interested you are in what we do, it still seems like science fiction. They don't, they don't. So I would say the more advanced ones are interested. But the closer you are to the intermediate chemical. the corner immediate, it's the more that you can't, you can't even imagine. You just bought a company, Lodoh Therapeutics. Yes.

Tell me about what that company does and why you bought it. Yeah, so one of the things that we're always on the lookout for are sources of novel genetic diversity. And we have a capability called metagenomic sequence. We have a metagenomic sequencing capability that allows us to take a soil sample or an ocean sample or an environmental sample and to reassemble the genes of all the microbes that live in that, most of whom the vast majority of which you can't culture in a lab. And Loto had a similar but complementary technology

that allows us, we think, to be able to, without getting into it, to really, excuse me, 10xR collection of metagenomic diversity. And this becomes useful for creating novel genes or finding novel genes, creating clusters of genes that make novel products, etc. Tell me about plastics, you know, and the function they have in society and then how that gets replaced. I know packaging of foods is something that we use a ton of plastics for to try to, I guess, extend shelf life or protect them or in some cases, just aesthetics to make the fruit still look beautiful. That's all going to be replaced at some point, you think?

I do think so. I think that in Europe, the regulator is requiring that it be replaced, not because it uses so much carbon in the creation, but because the end of life, I mean, plastics last forever. I mean, plastics are light and cheap and they're a very effective oxygen barrier and they last forever. That's great until you start really thinking about the implications of the last forever, given the volumes that we use them at, right? We've almost created something that's too good. Like, the innovation of plastics to make something that lasts forever like that, and then to make it at scale and to make it so cheap. we are now victims of our own success there with straws that will outlast the human species possibly, unfortunately. That is exactly right. That is exactly right. And so I think regulators appropriately recognize that we can't go on.

And so are forcing us to innovate our way to new materials that offer the performance, right, that we're used to, but don't have the end-of-life characteristics that traditional plastics do. In other words, in another couple of years, there could be a wrapper on food that when you throw it in the garbage, goes into a landfill and then just dissipates. It has some sort of window of life that's not, you know, a million years. I mean, there's a couple ways you can do.

One is you can have it compost in the way that you describe, you know, biodegrade. You can also look, I mean, one of the problems with lots of food packaging is that there are laminates of stuff. They don't work in existing recycling streams. So even if you wanted to recycle them, so maybe you can find some way of making the laminate recycle. I mean, there's all sorts of innovation. But it's interesting, right?

I mean, it's innovation that requires you to extend beyond those core six or seven hydrocarbon molecules that people have making materials with the last 150 years. Got it. And then what about food itself? Is that an area where material science is overlapping with what we're seeing in the production of food and people making changes there? Or is this two different sciences? Well, I mean, I think that the material science,

is not so much. There's some around crop protection agents, but I will say that, for example, Impossible Foods is basically a material science company with a biomolecule at its core, right? Right. I mean, I'm reframing what they've done, and they've been incredibly impressive, right? But I think that what they've done is they've taken the same technology. They make he, right, which is the molecule that creates it. They make it in a microbe, and they've used that to assemble a burger, right, that looks and tastes like a burger. And so it's the same basic concept. What they're doing is by increasing molecular diversity,

you can create products that in some cases have features you wouldn't have had otherwise. What's the biggest challenge in running this company and trying to solve these problems? A couple challenges. One, I mean, growth is hard, right? My respect for anybody who's grown, even the most basic company, has only gone up as I've done this. And to grow in a business that has deep technical domains across multiple disciplines where you've got to kind of create that and to do it in a set of markets that have very

slow by venture standard adoption, customer adoption cycles. I mean, just keeping the culture together, that's hard. And I think managing, that's what I would say is the hardest thing. Yeah. And do you intend to, with like the bug spray, be the brand and release it under your own label or do you go to

somebody who already has that channel and say, hey, you can take this chemical and make your own products out of it? Yes. Yes. Both. Both, you have the optionality of both? Yeah.

No, I think, I mean, look, one of the things that's interesting, for example, about Tesla, right, is, as I understand it, I mean, this is just from the S-1, they set out to go make power trains for the industry, right? But the industry wasn't actually interested in adopting the power train.

And so Elon said, I'll make cars. Yeah, he had no choice. I mean, I think the story was there were venture capital. I mean, I think the story was there were venture capitalists who said, well, back the company, but we don't want you to make a car. We want you to sell these batteries and power trains and technology to Mercedes, who was an original investor in Toyota, who weren't ready to produce electric cars. So they had no choice but to make them. But they, they, I would only, and I don't know, because I wasn't there. You said they weren't ready?

I don't think I would say they weren't ready. They weren't interested. I think that's actually, I think you're accurate. Yes. Right. They weren't interested. And I use this as an example, and it's going to come back to the point I was making, which is in the material world, the trick is getting people to want to try, right?

And what we know in industry, disrupted industry after disrupted industry, is that the incumbents are not actually interested in adopting the new technology. anywhere near as quickly as the technology innovator can deploy it in some way. Right? Yeah. I mean, there is no reason why Salesforce should have been the dominant kind of SaaS.

Like, Oracle was well set up to do that, and yet... A bit of an inventor's dilemma. I mean, you... Yeah. And a big crossing that chasm of, I have an existing revenue stream here, to protect the ice engine and I'm going to create a car that's $150,000 and only goes 150 miles. Yeah, that's kind of hard to get up for if you're Mercedes or Toyota.

Yeah, and I mean, this often gets categorized as an innovator's dilemma. And I think it's more fundamental and cultural than that. You literally can't imagine that you could do something in such a foundationally better way. Right? The people that are making the judgments, they're not the CFO who's kind of looking at the revenue, profitability of a business unit and saying, well, how can I do this? And by the way, the investment dollars are usually so small relative to total cash. They could totally afford it, right? It's, they cannot imagine that you could actually make a product that's a better product because they can only see,

they're like, oh, I'm going to spend $150,000 for a toy roadster that goes 95 miles. Like, whatever, that's never going to be a threat, right? Yeah. Yeah, I mean, I have that roadster. It goes about 100. I think there's 190 was the listed, but if you drive it, you know, in an aggressive way, you're probably going to get closer to 125. And yeah, now I had the battery replaced on it. Now it goes 300 miles. So it's very interesting to see over 10 years that when you get the battery replaced, it now triples the range. But it's interesting. I, um, you know, I'd buy a car like once a decade. And so it just got a car and test drove a bunch of them. And all the electric cars were just better than the ice cars.

They were just better cars, right? There was no comparison. And so, and this was true, like, the Tesla's and the Volvos, like, they were just better cars, right? In all real world conditions. It is interesting. There were all of these fears people had about them, range anxiety, the battery bricking and just what happens if they get an accident, the weight distribution.

And then all of those went from being question marks. I think, to explanation points, like joyful explanation points. Like, wow, these things are actually better. Look at the, look at how fast it is. Look at the pickup. That's right. And that gets back to the point that we're doing. You ask the question, like, how is it cost us?

It's better. We're making better products, right? Nature gives us access to molecules you can't get otherwise, and we can make better products with them. And yes, they're better. Right. Yes, they're made in a better way, but they're just better products. That's the whole point.

The ability to make chemicals that make plants absorb nutrients faster, I know this sounds like science fiction, but does this mean that we will be able to grow plants faster than nature currently, they currently grow in nature? In other words, a cannabis plant or a tree or a lemon tree, whatever it happens to be that you're growing in your backyard could grow faster or be more plentiful. I mean, what's true is that there are opportunities to create growth stimulants, right, that allow you to grow trees and other plants faster more productively than they would otherwise do so. So the answer is yes. The answer is, yes, sorry. Yeah. Yes.

And the answer is, I mean, just to let that sink in for people, the ability to grow a tree faster and to be able to tweak the, I mean, we've been doing this for centuries, obviously, with hybrids and hybridization and I guess certain chemicals to keep pesticides or whatever to keep bugs away. But this is literally growing things better and faster than nature builds them. Yes. It's pretty amazing. Yeah. The impact of that.

We could rebuild the rainforest faster or something or maybe even someday coral reefs or something. Yeah. Yeah. Coral reefs. Yes. I mean, in theory, sure. I certainly am not going to put a timeline on when I think we're going to have that technology or how much faster it's going to be.

Yeah, I would think growing things underwater is slightly higher challenge than doing things above ground. And so you mentioned before being a public company is something that you see in this space, or at least in the drug discovery space and some other spaces, has that now become challenging in that we have this meme stock era where people are loving to buy speculative things or things that are, you know, I mean, essentially the public seems to, retail investors seem to want to be venture capitalists more than buying a growth stock, you know, buying Amazon in year 30 or 20 or 30 versus buying your company and year one, year one as a public company and you're, you know, whatever it is, as a company. How is that impacting what you're doing? I mean, so we don't have a lot of Reddit investors in our stock.

stock base. Yeah. And so far, so far, so far we've avoided that. I mean, we were one of the things that's great about a regular way IPO, a success one anyway, and I think GORS was quite successful, is you get to pick your investors. And we were quite careful to pick what we thought were really tier one blue chip investors who understood we were doing, understood the scale of what we're trying to go after, understood that we are in the early innings yet, and they're going to be, you know, ups and downs along the way, and hopefully will minimize the likelihood of the impact if somebody with diamond hands decides we're exciting.

It does seem that this analogy, or the most apropos analogy, correct me if I'm wrong here, is the 70s and 80s in the computing business was really about getting the hardware and the infrastructure to just function and operate correct. and it feels like in your industry, it's kind of like the 70s for PCs and like just getting an operating system

in memory and storage and the monitor and the video car. I mean, video cars didn't even exist at that point, really, as a separate product, getting some kind of connectivity onto it. Is that kind of where your industry is now? I think that's right. I often say that where the where semi was in the early 70s. Right?

Yeah. And the metaphor is imperfect in a lot of ways. But I think that I think your metaphor of like trying just to get the whole thing done, right? Like before the home, like we're before the home brew club though, right? I think we're, so we're earlier than that. And does this eventually, what scale does this get to in terms of accessibility? If we're going to go with the homebrew club and then Bill Gates is sort of concept of a PC on every desktop or a PC in every home, Is that something we're going to see here where this kind of bioengineering is going to be available,

you know, maybe not in everybody's home, but maybe in everybody's home, like the replicator that we see on Star Trek is the idea, eventually that these chemicals could be in some sort of a device that could make a tea, you know, an Earl Grey tea at a certain temperature like Picard would order? Or is that like way out there? Yeah.

I was about to say it was way out there, but then I was trying to imagine that I'm, you know, Bob Noyes in 1973, right, mocking. I mean, I don't know if he was mocking, but like people looking askance at Gates's comment. I think what we're going to see is, I don't think we're going to see in every home, but I think you are going to see much more distributed manufacturing, right? And I think we're going to see much more general purpose manufacturing assets that allow you to make a far broader range of these products. But the finished products still require at-scale manufacturing, right? So you can make the core intermediates in a distributed way, but a chemical is not a material, right? And the plastic wrap, the biodegradable plastic wrap around your apple is still going to need to be cast on some line that's going to be where there can be big economies of scale.

And that can occur in the United States. We could actually take this next wave of manufacturing and have it occur anywhere in the world. Does it require a ton of human beings in this process, or is this all automated in a way? I think there's no reason why it couldn't be in the United States. Yeah. This seems to be a great opportunity for us to bring manufacturing back without bringing back, you know, things that are in the review mirror. We've skating to where the puck is going. I think that's exactly right.

I think this is a huge opportunity. And certainly policymakers on both sides of the aisle are very interested in taking this as an opportunity to bring manufacturing jobs back. I completely agree. It's so hard to talk about this without referencing science fiction or sounding like we're futurists, but what's the downside here? Are there things that we are messing with that we don't understand enough? And what are the safeguards that you think about in terms of what you're making, i.e., you know, how safe is this bug spray? Are there unintended second of their consequences to the screen you're making, etc.? I mean, first thing I would say is that there are existing and very, I actually think quite good regulatory

infrastructure and protocols for testing new chemicals and new materials. So you got to make the EPA happy, right? Like the EPA requires human safety testing, right? I don't get to just like launch my product in the United States without a regulatory approval. And so I think that, I think those processes work pretty well to be direct. I know that some people find them burdensome and they certainly cut across the, like, move fast and break things. But, you know, when you're talking about stuff which can leach into groundwater and which people are, I am okay with going slowly to make sure that you've got human safety. We never want to take risk on human safety. So for the final products, I think we're actually the regulatory infrastructure is generally

in pretty good shape. I think the question is, you know, and we feel, generally speaking, that the process around engineering the microbes is also quite safe. There are a set of regulations around that. And there are, at least the way we do it, incredible safety protocols around working with killing bugs, making sure that stuff doesn't go out in the wild, making sure that you're creating microbes,

which would not be able to be, you know, to survive easily in the wild. So I think we feel generally speaking, it's all very, very safe. And what is the chemical or the compound, I'm sorry, that's being used in these biofilms? Like, obviously beer we know is made from like yeast and hops and water and whatever.

What, what, how do you know where to look in nature for these precursors? If that's the right term to you. Yeah, that's, that's the great term. So we, we don't talk about this as much, but we've built a huge, you know, another value of what we've done is we built a huge machine learning, basically a giant library of novel biomolecules. And it's growing all the time. So today it's about 75,000 molecules.

It started when we started the company eight years ago, 700 hand-coded ones. we've just added again in the way that data-driven companies do. We've been adding decision rules along the way, right? We've been taking experimental data and finding false positives and now it's 75,000. We expect it to be in order of magnitude more by the end of this year. And then because it's now big enough, now we've layered on all sorts of genuine machine learning that helps us identify the...

So you come in and you say, I want a molecule that's going to allow me to make a material that does this. And we've got a set of machine learning tools that allow it to sift through all of those and come up with, you know, 50, 100, a thousand of those things, at which point we start testing them. That's wild. And so the molecules could be found anywhere from somewhere in nature, in the earth, in a plant, in a species. They might never have been seen by a human before. They might only be, they might be wholly novel things that we are predicting can be made. Now, our predictions are pretty good at this point because we've actually modeled the reaction rules and the thermodynamics. So we're actually coming up with new molecules based upon what we've learned about planet Earth.

Based on what we've learned about enzymatic chemistry. Got it. What is enzymatic chemistry? So an enzyme is the protein that a gene codes for, and think of it as like a step in a chemical recipe. Got it. So we know about those. can make predictions about what comes, what might. How you can string them together and therefore what the molecule they can make by stringing

them together would be. That's right. Literally like probithias. It's literally like, it's not literally. Not literally, you know, because that's fiction. No, but this is, but actually, Jason, I mean, this is a good example. You asked earlier, right? In order to build the system that did this, we had to have deep understanding of biology.

We had a deep understanding of chemistry. And we actually had to have deep understanding of how to build a very computationally, intensive process that you can then update. I mean, the first time we ran this to saturation, it took like 10 days, right? Like, you know, I mean, this is not, this is not simple stuff on the compute side, but it's only as good as the ability to pull these enzymatic levers.

Now you asked about the load of therapeutics. Now you've got a whole bunch more enzymes that we can start to put in that system, right? And so what we build is very real, I mean, I know it's a bit of a cliche, but there are very real flywheel effects up and down our entire platform in this way. And eventually, as computing continues to expand, I mean, I don't want to bring quantum computing into this, but even as networks of computers, I'm assuming you use some cloud

provider to run this stuff, over time, computing is going to increase, and then the data sets going to increase, and then the possibilities increase. That's right. And I think that way, we've seen those kinds of, I mean, I'll tell you that in 2015, 2016, 2014, I don't remember. There was a pretty material increase in the product quality of one of the large cloud providers. And their tools got a lot better.

And that allowed us to migrate from, you know, bare metal, right? On-prem bare metal into the cloud. And that created a whole bunch of ability to do stuff we couldn't do otherwise, right? And as they've continued to improve their speeds, like stuff that was too computationally expensive to run is now borderline trivial. Right? So you're absolutely right. We've already started to see that.

And so as you do that, you're able to then gather more data, which means you can now have better algorithms, which means you can now find stuff. In our case, it tends not to be, we find stuff faster. We find stuff better. Fascinating. Listen, we're going to be monitoring what you're doing going forward.

And congratulations on the IPO. I know it's just another financial event, but it does give people the ability to have insight into what you're doing and getting that first product to market. It seems like those foldable, phones are going to be the future. And those are ready for prime time? Those,

that's my customer's judgment, not ours. I'm not in the cell phone. That seems like a really good product for Apple. I wonder why they haven't come out with one yet. I think they're having, as we're taping as, I think they're having a keynote. Are those films going to allow any other form factors

that wouldn't be obvious, aside from folding? Well, I mean, rollable is also, I don't know if you consider rollable obvious, right? But you can also then imagine things that you could put on your arm, right? So I don't know if that's rollable. You can imagine a device that's a surgical sheet that you lay across your body and now has the x-ray. And so a surgeon can start to look in real time in an overlay on your body.

Wow. So I don't know if that counts as a novel form factor. I think it counts as extremely novel. Like we have seen in movies and science fiction and MIT, media lab, curve displays as bracelets, like a Wonder Woman bracelet kind of situation or a sleeve. But the idea that you could lay this on top of a person who's being examined and it's doing the x-ray in real time and showing what's going on in their body. Wow. Or imagine somebody's being operated on, right? And think about how many surgical layers there are because people

are like, they can't quite see what's under there and they can't get the real time and they're looking on a screen. And now what you've done is you've just created, it's like, you know, in Google Maps when they put the data layers on, it's not just a data layer on top of your body. It's incredible, actually. But this is like a, this is a great example. Sorry, and I know we're, you know, more or less at a time, but this is a great example of how new materials, changing the boundaries of the material world,

start to open up things in ways that truly are mind-blowing. And why, if you can open that boundary up, you change society. And this is really the lever that moves society, I believe as much or more than any digital lever. Well, I mean, the ability for batteries and a screen and a processor to fit into the form factor of a smartphone and be carried with every single person

on the planet has had dramatic impact. And as Steve Jobs was said, that was as much a material science revolution as anything else. The ability to touch the screen and have it do things was also way up there, let alone sensor technology. But that's all material science.

Do you know how many adhesives are in an iPhone? God, it's got to be dozens. Yeah. I mean, some people think over 100, right? Wow. And again, this is, and they use the adhesives to replace screws, lighter and smaller and thinner. And so again, all of these things that we take for granted are foundationally in part material science. It's unbelievable. It will continue success with it.

And we will be watching everything you do like a hawk because it's just so fascinating the possibilities here and continued success and understand you're hiring. So if you're in computer science, biology, or generally an adaptable person who is into photography, and biospace and any, just if you're a super geek nerd, smart person, this is an interesting place to do meaningful work,

I would say. I think it is a very interesting place to do meaningful. And so if you were thinking about maybe making an ad network, get people to click on ads 15% more efficiently, don't do that with your science degree. Don't do that work.

It does nothing for humans. We have a lot of people in our technology department, which is like 160, and 70 people. And, you know, there are a huge number of people who walked away from jobs at large ad tech or large social media because they don't want to be responsible for increasing click-through rates by 10 basis points.

Yeah. Like, we get to figure out incredibly cool stuff. We work with robots and figure out how, you know, how to make new materials. Like, who doesn't want to do that? Yeah, it's mind-blowing and it's science fiction becoming reality. And my gosh, just the ability to do something positive for the human species, and all of these technologies starting to compound on each other. AI, material science, chip technology,

all of these things are compounding now to create an overlap that I think nobody could have predicted. And this is one of them that is just absolutely fascinating to me. So continued success and we're going to be watching every move. And we'll see you all next time on this week's startups. Bye bye.