 Good afternoon and welcome to today's energy seminar. Today, I'm excited to announce our speaker Greg Mahalan from Citrine Informatics, where he is the CEO and co-founder. This quarter, we're going to end with four corporate talks, interestingly, two from large energy companies and two from outstanding successful entrepreneurs, rather smaller companies of which Greg is example number one. I would like to give a little bit of background without stealing Greg's thunder about Citrine. Greg himself has a degrees in material science from Cambridge and electrical and computer engineering from NC State, but also was a Stanford MBA. Right here at Stanford, I think the team was formed and it was the pathway to reality and the real world success that they've had was helped by the innovation transfer program at the Tomcat Center. And in fact, Brian Bartholomew was the first of many to recommend Greg as the speaker. Greg, among many other honors, a young guy with a long resume is a member of the World Economic Forum Future Production Working Group was named the Forbes 30, to the Forbes 30 under 30 list in 2015 and was named a world economic forum technology pioneer in 2017. So without further ado, I'd like to introduce to Greg Mahalan to talk about his experience and doing a very, rolling out a very successful energy startups to train informatics. Greg. Well, John, thank you very much and it's a, it's a real a real pleasure to be here and I'm going to share my screen so folks can see. It's a real, it's a real pleasure to be here and you know I think you highlighted nicely, you know, the foundation of Citrine really came from some time with with a classmate of mine at Stanford. And we both had this deep interest in material science and and the reason we had such a deep interest in material science was with partly scientific. I'll admit there was there was a huge motivation for me at least in this idea that, you know, we could use materials to really change the direction of humanity and I and I mean that no, no small way. But also, you know, I think our realization was that materials is a really big business and, you know, it's like so many things. If we can use the incentives and and possibilities of the business world to transform some of the behaviors of the people around us and the companies around us. We believe we could actually truly affect a lot of change and the way we use energy and we use resources. Today I want to talk a little bit about where we got our start and what our philosophy has been and a little bit about what we do to use the the science and technology and business that we picked up at Stanford and in other locations to really try to change this incredible industry. So I guess to start with, you know, when I, the reason I got into materials in the first place was because I had a realization actually, as you mentioned was studying electrical and computer engineering. And I had a realization that when I was doing transistor design, which was kind of my direction that I was engineering around materials and awful lot of the time. And that my, my hope was that I could figure out a way, how can I use material science to make it so I didn't have to engineer around things. I could use materials to solve problems, not to be the limitation. And this led me down a really interesting path. In particular, it became super interesting because there was this realization this moment that I had, which was that materials really are the thread that connects every physical thing on earth. And when we talk about this to materials companies, you know, I always say, you know, materials are the enabler, or the limiting factor. When I talk to the rest of the world about it, what I like to say is, no one cares about materials. There's no one who cares about just materials, we care about is what we can do with materials. And that's what makes it so exciting. For a long time, we've always talked about kind of materials, eras, right, you have the bronze age and the stone age, this is sort of the classic recruiting pitch to the material science domain is that we name eras after materials and that's how important they are. But I actually think we're really moving away from that in a pretty meaningful way. And the reason I say that is because we've gotten so good with materials for so long that now when somebody invents something new, it doesn't revolutionize every industry. It revolutionizes one industry, you know, we don't have the invention of steel anymore where it just sort of takes over everything. And that actually creates a really cool opportunity for us, because what we realize is that for a long time we were trying to, you know, the material was our hammer we were trying to use everything in every possible way. And in my world, it was really more about this, you know, kind of how do we use the right material for the job. Before I get into that particular line of thinking. I want to touch on exactly why this is relevant to energy because I think to a lot of people, the possibilities of the possibilities of materials and how they connect to energy isn't always completely obvious. So the first thing I want to highlight is this this image you see on the screen. It's sort of the, the very high level of what is the life cycle of a material. When you pull it from the ground, you design it you make something of it right we don't use iron ore all that much anymore we use iron ore that is then converted to steel or aluminum that's converted to steel, or sorry aluminum is converted to more advanced aluminum. You then scale it up and manufacture make it into some intermediate thing a sheet or a block or what have you. And then you hand it over to somebody else, you have an OEM company that starts to make a product you know Apple makes the MacBook or Ford makes the car, and ultimately, we get it into our hands as consumers. And what's really interesting to me I'm about to go talk about all kinds of stuff with with respect to energy but if you look at each of these blocks on the screen you can see that mining is an incredibly energy intensive activity. Making materials smelting purifying and producing is unbelievably energy intensive. Making products you know actually making the final device uses a lot of energy and affects how energy is used tremendously and so I'll talk a little bit about this but our goal at Citrine is to connect the design of materials and the materials we can make with those end products and solutions that actually allow us to live in a more efficient sustainable world. So, when we started Citrine actually it kind of, you know, what really kind of shocked me into starting the company was actually a course I took when I was in my MBA program. And the course was talking about sort of energy flows and energy trading and energy markets, but what I realized was every area of materials, or every area of energy is affected dramatically by materials and so this is a Sanky chart that for those of you who are interested in energy this is the 2019 one from Lawrence Livermore. This is probably something that's familiar right we generate the materials here on the left, or sorry the energy here on the left and then we use it in different ways, either good or bad over the course of his life. But what I started to realize was every stage of energy usage and energy creation is really affected a lot by the materials that we use and I've come to understand this more and more. I mean, solar is probably pretty obvious right solar cells are basically just a couple of materials, and then you're trying to take the energy out of them. But with nuclear materials right you're trying to optimize these things lighter stronger blades, and there's an article in the New York Times about the most recent GE turbine that was released that's, it's for offshore wind but it's just truly shockingly enormous. And what that does is, it means you have to have super strong blades that can handle storms and handle their own weight and all these things, and the materials you make them from really matter. So it turns out for even even fossil fuels. With the traditional fossil fuels it turns out hotter is better. So you can burn cleaner if you're burning hotter, but there's a lot of problems that you start melting the metals your turbines are made out of, which is a thing that actually can happen. So we realize is that in generation materials are such an enabling, enabling thing and for those of you who are studying materials studying chemical engineering chemistry or even even systems. You can. This is not new, but from, for me at least, it wasn't new but it was it was shocking to think about the breadth of these technologies and how limited we are by the materials that we have today, even though we understand them so well. And by the way this skips over all kinds of things like catalysis and recycling and all kinds of other energy intensive things that that can bring these things together. So where things are used where energy is used materials actually play a massive role there too. So, you know, I just use heating and cooling as a pretty interesting example. You know residential and commercial better insulation materials, better glass, better roofs and these sorts of things to make sure we don't lose energy to the environment or we don't want to. We'll talk about agrochem in a minute but we also talk about things like lighter safer vehicles. It turns out that transport is a huge volume of our of our energy consumption, and if we can make our vehicles lighter, the better off we are. And so, I like to highlight a few things. First of all, the opportunities for materials innovation are truly enormous. And some shocking statistics for me. Haber Bosch alone, so the process by which we make ammonia consumes 5% of global natural gas, 2% of our global energy footprint annually, and generates 1% of our CO2 emissions every year. That's just an insane amount of energy. And when you think about it, it basically links this agricultural world and makes it a petrochemical product because of the amount of energy it consumes. Whereas, there are people now doing work to try to create the same process but using catalysis using a more efficient catalytic process, so that we do consume less energy but we still can feed the planet which is obviously a huge priority for us. Cement creates 8%. It's almost a 10th of our global CO2 emissions. It's a crazy amount and actually will probably get worse as, as some developing economies accelerate and accelerate and build more and more cement is a huge part of that world. And so the opportunities to come up with new CO2 efficient cement is huge. And I'm not going to talk about the other two in depth, but just suffice it to say that everywhere we look, there are opportunities for more efficient energy applications by being lighter, stronger, hotter, more efficient, all the way around. These materials enable us to live in a way that is frankly quite sustainable in a way that will allow us to survive for many, many years. In fact, there was at one time, Bryce, my co-founder and I were both at the GSB, he said, is there a credible argument to be made that the story Citrine wants to tell can actually save the planet can actually have a positive impact on both energy and environment broadly. And my assertion was yes, because our goal is to accelerate the development and deployment of the next generation of materials and chemicals. And in my view, if we do not do that if we do not come up with new materials to solve the problems we need to solve to solve today, we are going to basically be extinct in less than 100 years. And so the faster we can do it, the better off we are. But it's not easy. And so here's where I'll talk a little bit about what we do and how I believe it's really changing things. So, our industry, the materials industry is constantly under tension, right, as you can probably as you probably know, the development speed of these, of these, these products, you know, these are companies right they want to get things into the market as fast as possible. But at the same time they're always looking for the higher performance thing right and, and I think most of you who are involved in scientific research know that the further you're trying to push the frontier, the riskier it is and the longer it's going to take. And so we have this natural tension between performance which can be sustainability it can be efficiency, it can honestly be nothing to do with energy at all right it can be, you know performance could just be you know the car drives faster. So I think of sustainability and energy efficiency as a, as a performance characteristic as well. And the product cycle demands of the worlds we see it today so this development speed is really a challenge. And here's the problem. And you might not think it's a problem at first but the real problem is, we have more control over the materials that we make than we ever had at any point in human history. If you work in materials or mechanical engineering or chemical engineering labs, you know that we can basically place atoms perfectly one by one. I mean it's not very efficient but we can control materials incredibly well. And that's wonderful, except what it has done is it has given us a new, a new search space, a new set of haystacks we have an entire planet of haystacks in which to find the needle that we're searching for. And that's a huge problem because if speed is our goal, then all of a sudden this new searching we can do is a limiting factor for us, we were relying on serendipity to get there. And so that's what citrine set out to do. We set out to figure out how do we reduce the search space to just to the most promising haystacks. And when we do that, can we actually create a new material that achieves a new goal, because in my view, we're headed to a world of mass customization where we use the right tool the right material for the job, every single time. And to enable that you have to know what your palate looks like. Before I go into exactly how we do it, and exactly why it's interesting because, by the way, the best part about being an entrepreneur and the best part about doing what we do is that not only do I get to work with materials which is an industry I love and a topic I find fascinating. But I get to work with some of the best minds in thinking about AI for these scientific applications, which is obviously a really exciting thing to do and so this is this is the the slight dilettante in me, saying that being a CEO of a startup is wonderful because you get to see a little bit of everything. So, I always like to start at the end. So, we were talking with a company. This is actually a real project that we did with a major chemical company, and they were mixing two things together to try to create was actually this is actually a catalysis problem. So they were using a two part catalyst. And there were about 3.8 million candidate pairs they come up with, you know, in different modalities different combinations different ratios different shapes, all these sorts of things, a massive search space and by the way, in our world a relatively small one so 3.8 million, it would take them roughly the rest of eternity to test that. And so they then by bringing all their data together in our system so we do two things we bring materials data together. And then we use AI, especially tailored AI that I'll talk about in a minute to help identify what materials hold the most promise. And they were able to down select to about 90,000 that we're promising, not not the rest of eternity but probably a good few scientists lifetimes to test that entire space. We then said well, we should be able to use computers to do this we should be able to figure out how do we down select even more promising materials. And we were able to immediately just based on some real data science it wasn't even true AI at this point, about 27,000 materials. We put it in the automated system we put it in the AI system, and it was able to select 76 of the most of those 3.8 million, the most promising 76. And then we realized we, you know, we took the list back to our collaborators. And they said, Well, actually, these of these 76 13 we've seen before. And so there were 63 materials in this in the sea of 3.8 million that had never been considered by this group, and was actually ultimately ended up hitting their performance goal. And so this is the goal this is what we want to do we want to enable this level of materials responsiveness all the way around. So, you know, of course this is where you know the presentation is basically over right we just, we have AI to the rescue. And here we are. But that's actually not true right you knew you knew where that was going. Because here's the problem, you might have heard the, the, the old saying, I guess, old ish saying data is the new oil. Well, it turns out that's not exactly true. Because when people talk about big data and they talk about them, they talk about AI and all the things we can do. And, and by the way, all these companies, many of which are within 50 miles of Stanford. Doing this breakthrough AI work, what they fail to realize is their world looks very different from the rest of ours. So let me give you some examples. We have data. We talk about huge volumes. So, these are just these are actually pretty old examples right 2014 Facebook is collecting on its users 46.3 gigabytes per second. Tesla in in as of April 22 2020 had driven 3 million miles. And by the way they basically collect video for that whole time. It's not exactly video but they they're doing a lot of data collection I don't even know what that that compares to in terms of on disk size. Google is obviously collecting enormous amount of data as well. And so this is really exciting right you know these guys are doing amazing work and they're doing all of this, all of this thinking about what is possible. Why can't we apply that in our world. Well, I've talked to a company that had one of the largest product catalogs at any company we've ever worked with. We've had 100,000 products ever that and that's not even in their product like for sale product catalog that is the number of things they have tested. And so even with that this is the biggest number I've ever seen. It is 23 million times smaller than what your Facebook's your Google's your Amazon's are working with. That kind of makes sense, at least makes sense to me, because when you start talking about new materials new chemicals new formulations new paints, new coatings, all of these things. A new tests can cost a million dollars actually can cost $10 million. If you're talking about aerospace engine materials, you know those can be really expensive to test. So the reality of this industry, which which is so critical to our success in being a more sustainable planet is really problematic right because there isn't that much data. So let me give you some examples of what the data scale is that we typically run into. We worked with a big, well known adhesives company. And this company had about 2000 was just just shy of 2000 products in their catalog. We worked with Panasonic, you know the the company you probably know well on a project that was incredibly successful. They started with 32 example molecules. This obviously not a machine learning normal number right if you go to Google with 3232 of something. They're going to say hey we have a product for that we can help you, and it's called Google sheets, because you can enumerate that almost on fingers and toes right I mean that's just just a small amount of data. And we worked with a global petrochemical company who showed up with about 280 example formulations of one of their products. You end up in this place where the promise of AI the promise of using computational power in the way that everybody talks about in Silicon Valley isn't real to this part of the industry. And so you start to think about the materials and chemicals industry in a very different way. The reason for that is that these companies, it's not that they're outmoded in that you know they've been they've been leapt by other people it's that their reality is very different from the reality that most of us think of in the tech world. And so when Bryce and I set out, we set out to say, we want to bring this new way of thinking to the materials and chemicals industry, but build it in a way that they can adopt. This is one of the most important questions you can answer if you decide to start a company. One of the most important questions you can answer is, how do I uniquely serve my customer. Because if somebody else can do it the same way you can, you don't have something defensible, but if you have that insight you can meet your customer where they are. You have something that can really change the world and you know I think really create huge business opportunity and real technical opportunity. So as with anything, there's real challenges and opportunities here. So first, you know some of you might know there's a pandemic going on right now. That's why I'm not on Stanford's campus. We have these these challenges right we see in 2020 we saw the chemicals industry grind to a halt for two quarters and slowly start to come back in Germany they've talked about no recovery until 2022. And frankly the, the EPA and regulatory bodies are starting to say, hey, these pollutants these products that these companies are putting out are really challenging to the materials and chemicals are really challenging to the environment and so we need to start paying attention to them, which in my view creates a huge opportunity. The CEOs of the global chemicals companies are starting to listen frankly to a lot of activist investors, who are saying sustainability is the key to consumer preference and sustainability whether it's an energy, or in sustainable products are things that that people really want to see in the companies they work with. I'm sure all of you, especially those of you who are in the Bay Area. You see Tesla is running around but you know now you see electric cars from VW group you see electric cars from GM, all of these new products that are ostensibly more sustainable have this big big asterisk next to them. How do we make a battery that doesn't require rare earth elements. How do we make a motor that doesn't require rare earth elements how do we make a battery that's recyclable. Things are tremendously challenging to the environment, and you know, we're starting to even see oil and gas companies come around and say hey, we're about to go the way of the dinosaur if we're not careful because consumers aren't going to want to buy our products and regulators are going to remove us from existence. So this is the context in which Citrine was started we believed that there was an opportunity to help these companies change. And so we came up with one word that was so important. And that was agility. We started to believe that the best companies are the ones that could respond to actually changing world. You know as much as I would like to say we avoided climate change we didn't. And so the world is more volatile than it ever has been before. The idea that a company isn't just the plant in the ground and that's all it is it's a group of people that can learn and respond became our mantra became the idea behind Citrine. And so, when Bryce and I started the company. We had this view that there was a way things were done today, and there was a slightly different way that things would be done done in the future. Scientists and engineers. For those of you who are in grad school and technical disciplines particularly lab disciplines, you know, the way you plan experiments, you probably feel like it should be something that can can change, or it can can be sort of more systematic as you get more industrial. I hate to tell you my experience in industry was basically the same as my experience in grad school was very intuition driven. It was very personal. It was not a systematic way of making decisions we believed we should use data to make those decisions in a way that hasn't been done in the past. Very oftentimes, even strategic decisions made at the very highest levels are made based on kind of a scattershot approach. And you should rationally be able to determine what are your likely paths to success, and actually attack them, figure out where should I invest to have the most bang for my book and have the most impact environmentally, or I guess least impact environmentally depending on what you're thinking about it. And finally, I have been at companies where an adhesives group and a lubricants group worked on the hall from each other, and they never speak, because this is an area where where they don't do the same thing. And yet the physics of what they're doing is very related. And so our idea was, if we started to bring these things together, we could start to create a really good way of collaboration, actually create a scientific discourse in some of the world's best companies because these people have expertise and we need to take advantage of it. The first thing, as I mentioned, is data. Well, the problem with materials and chemicals is that we know a lot about them, right? We understand structures, you know, there's some examples here of sort of hard-sign and soft-sign and polymers. And the issue that we run into is that when we try to store materials data, we often just write down the things that we remember. So we actually, a lot of times, use baking as the example within Citrine because it's sort of something everybody kind of understands. And you might remember that, hey, I baked that at 350, but you might or might not remember that you were doing it at altitude. Or you might or might not remember that you're using your mom's oven, which actually runs 10 degrees hot. Gathering that context is critically important in the context of a chemicals industry. And so we started by saying we needed to build a way to store chemicals and materials data. And this was great, right? We actually had someone with computer science expertise join us. And their earliest contributions were to say, how do I start to map these things? This is actually a screenshot from our system, actually, where you can say, I have these ingredients and I'm mixing them together along the way to get one particular outcome. And what this means is you can actually start to put together all of the recipes you've looked at in the past and compare them truly. I had a mentor of mine who used to say she was a Bulgarian scientist. Her name is Tanya Pascova and she was one of the most brilliant people I've worked with. And in working with her, she used to repeat to me, Greg, you cannot compare uncomparable things. Which at the time kind of was, it was sort of like an obviously, you know, sort of a tautology. But what it reminded me of is you have to constantly be thinking about how do the experiments you're putting together connect. And this allows us to be much more flexible in what we can connect than what other people might do, and sort of a traditional approach. But that's not where we stop, because data is the truest currency for AI. In science, it's only a piece of the puzzle. It turns out that we have hundreds of years of expertise. And for those of you who live in the chemical engineering material science mechanical engineering chemistry worlds. You know that you do not teach a student by putting a bunch of tables of data in front of them and say, Hey, here, learn. You teach them the Arrhenius expression you teach them the, the ideal gas law you teach them all of these relationships that we have learned as a scientific community to give them a jumpstart. That is why we say we stand on the shoulders of giants. So, at Citrine, we actually enable that in a very particular way and we call it domain knowledge integration. Because expert teams at the world's materials and chemicals companies don't need to use data to teach these these systems, how things work. They actually have learned it themselves over many years. And so this allows us to integrate things like equation, rules of thumb, business requirements even cost, all of these things together where we understand trade offs more explicitly than just data. And then we use data to better understand the non idealities of those assumptions. So it's a good way for those of you who think of it this way. We can put the laws of thermodynamics in and let the data help us really learn the messiness of kinetics. In other ways, this is the way scientists have tried to work for for centuries. The problem is the human mind can only hold so much in it once. And so, in our view, the way we will develop materials in the future is by using computers as leverage, and actually enable better decision making based on data that's been brought together, the knowledge you have as an expert, and the people around you who might be experts in different ways. When we talk about AI, right, you know, I'm sure everyone on this on this webinar sees AI on a regular basis, either in hype or in practice. And at the end of the day, most AI is about finding similar things. So, this is a goofy example right we have, we have this you know this this cat wrapped up, and the computer can figure out that is not a burrito. It is a cat, and great we're all happy. In chemistry. We sometimes want to do that right we want to find, how do I find the lubricant or the adhesive or the paint. That's 99% the same as the thing I sold last year, but 1% more efficient 1% uses 1% less toxic chemical, or we remove phthalates or something, you know some some pretty strict target but with a performance frontier that's pretty similar to what we've had in the past because it does the job well. Well, this is something that we excel at this is something that we that AI can really help with because humans are bad at figuring out subtle relationships between things. And a computer is very good at that. So we can allow the human to elevate above the searching for the needle in the haystack, and that we can enable that human to have five needles placed in front of them and then they get to choose which is the best one for the problem at hand. And so, in some ways this is the very easiest problem we like to deal with. But it turns out that most of materials and chemistry is not actually about finding things that are a lot like what you found before. For those of you who are in grad school, try publishing a paper on something that is virtually identical to the paper that was published five years ago. And it's not going to get through all that effectively at least not without some really good storytelling and some pushing on the peer reviewer. Generally, what we want to do is we want to push the frontier out, we want to discover new things we want to leverage ourselves. And so for this, we use a technology called transfer learning. And what we actually do is we bring data together with this domain knowledge, and it allows us to project out to say, here's what it would take to get five or 10 or maybe even 20% outside of the performance frontier that we have today. And that's a really cool thing to do, right, because it gives us the opportunity to actually branch beyond the, the, what we're exactly what we see today and it gives us the opportunity to optimize into new space. And so, building on these relationships as we go with time as we learn we can teach the computer more and more, and they can say here's a recipe that I believe is going to make you successful in achieving your next goal. And for us, that's scientifically interesting. I think it's really important for the energy economy, because whether it's storage or generation or even or even use. It's, it's so critical that we be able to refine things and make them better over time. Scientifically, that's fascinating. But you might know that, you know, especially since I got my MBA and they sort of pounded it in my head, you have to create an incentive for people to actually use this stuff. Right, because I can have the best science project in the world, but it turns out a science project is not what most companies exist to do. So we had to come up with a way to articulate this to the materials and chemicals industry that quite frankly at the time this is in 2013 was not that adopting of new digital technologies. Things have changed somewhat now but at that time people looked at us a little bit like we were crazy. We realize there are five ways that this thing could help. The first is kind of what I described it's, you know, the faster product development, the faster product development angle, can we actually get new products into the, into the market faster. The second is regulatory resilience. You know, if you Europe has been talking about outlawing plastics, full stop, which clearly they don't understand that would sort of send us back to the Stone Age, but the sort of regulating plastics out of the environment is going to be very challenging. And if companies can can actually respond and replace those plastics with biodegradable versions or lower energy versions or more sustainable versions, that's better. This year was huge supply chain disruptions as COVID moved around the planet. We ran into issues with so many factories shutting down that many companies had to reformulate their products, not because they wanted to, but because they couldn't get their ingredients. And so just like when you go to, when you go to the grocery store, and you pick up flour and you get home and you realize you grabbed the wrong stuff, and you start Googling how do I replace all purpose flour with bread flour. That's, think about doing that at an industrial scale at a plant in Houston. This is something that companies had to do within weeks of COVID hitting and starting to see shutdowns happen. And then finally, you know, dollars and cents is where it's after these companies so being more responsive to your customers and being able to optimize cost out of your supply chain, really, really important. So, I want to talk about two quick case studies and they'll be very quick a minute each. This is one of our proudest projects. We work with the HRL labs. For those of you who who were around, you might remember Hughes aircraft, this is the this is the research lab that spun out of Hughes aircraft, which is now owned by Boeing and GM. They're trying to create 3D printable aluminum powders for aerospace, so to make planes lighter, how do we print parts that are exactly the right shape, such that we can create a better, a more efficient plane a faster plane whatever your goal is. And this group needed to search 11 and a half million additives that they could put into this, this powder. And, you know, the result was that they got down to about 100. They were able to use our platform to get down to about 100. And then they disappeared quite honestly, when we started working with them, you know, they got down to 100 and then they wouldn't tell us what they were doing. And they just disappeared. And two years later, we got a call and they said, you know, Greg, this is the nature paper comes out next week. And I looked at my co founders and we said, what are you talking about that we don't even know what you're doing and I said, oh, we just created two new 3D printable aluminum allies for aerospace. And that was really exciting. Now they're selling it to NASA. So this is a commercial product today. And what was really cool, I mean, this is this is the guy who was doing this research for the time I think was a PhD student is now quite senior. He said, what would have taken years, citrine narrowed it down to days. And the reason that's so important is that when an expert tells you that it works, you can really believe that it works. And for me that was so important because, you know, it was always it was an idea that Bryce and I had when we demonstrated early I'll talk about that in a minute. But really when we talk about the impact, having someone who does alloy development professionally, confirm that it works to us was such a huge breakthrough for us as a business and it really was the foundation upon which we built a lot of our early decisions at Citrine. So the other example I'll give is a new class of thermoelectrics and this was a goofy one. We actually this is before Bryce and I started the company. We were students at the GSB and we, in the summer between our two years, we said we're going to try to see if this can work. We have this idea we're going to see if it works. So we started a partnership with a group at UCSB, without Santa Barbara. They had a kind of a database of thermoelectric materials, which if you don't know they use them to make refrigerators they, you can apply a voltage and make things cold, or you can absorb heat and put off a voltage. And we were able to with technology that was substantial substantially less fancy than we use today. We were able to show that we could invent not just a new thermoelectric, not just two new thermoelectrics, but an entire new class of thermoelectric materials. And ultimately we had it demonstrated by synthetic scientists they actually made the material that worked exactly as expected. This for us was was proof that AI can actually drive this change. And in an energetic materials class which for both of us was very exciting. So, I want to leave time for questions but you know I think this is a case of, you know, an opportunity being created, not because of, you know, a scientist sitting in a room, thinking about their one thing. But most of the best opportunities come from boundaries. And Citrine exists the boundaries of energy efficiency sustainability. The bound with with computer science and AI with materials and chemistry, and by bringing these together in a way that really no one else has been able to. To establish ourselves as as a real leader in the space and so as you're thinking about how to go solve these problems I'd encourage you to think about them this way. No one is ever going to solve a problem at least not the world's hardest problems by running at the problem in the same way everybody else is dying. You have your own experience you have your own view. You can combine all of those aspects of your knowledge and come up with a unique approach. That is the thing that's going to lead to success and frankly, it will lead you to the most interesting types of success. So, with that I think we have a little bit of time for questions now. Thank you very much I hope this was interesting and it's been a real honor to speak to all of you. Great, thanks a lot that was really terrific I expected to see a state of the art entrepreneur startup dude, and you fulfill that but much more than that for me I now understand why you're a world economic forum superstar. You were able to wrap into this talk which I thought would be more technology worded massive insights regarding strategy and even the way we think about doing policy analysis which is what I spend most of my time doing. So we did get a number of questions keep the questions coming. One precursor student who I know pretty well actually had two questions right out of the box. What are sustainable alternatives to cement if you know that and then maybe more generally, how can developing countries contribute are you working with people in developing countries who might have a different perspective, and a different way to use materials than we do in the developing world. Happy to so so I think with with concrete I would say it's it's less a sustainable alternative to concrete and more sustainable versions of concrete I mean concrete is truly unique and it's in its own sort of capabilities right you know there's a reason that we've used it since Roman it's less about that and more about the way we make it we haven't really changed the recipe all that much and so rather than thinking about it as replace it wholesale. There are many companies now working on low CO2 emitting basically when it cures. Can you can you figure out how to make it not image so so much CO2 and that's that's a reaction kinetics problem. It's a materials problem. There's a lot of very, very interesting interesting problems there. The, the sort of thinking about developing countries is a fascinating one actually a classmate of mine at the GSB. Even though she's not a materials person started a materials company whether she likes it or not that's what it is. It's a company called earth enable. It's this idea of they use local materials to create hardened floors in I believe they're they started in Rwanda now have branched out because one of the major public health problems in that part of the world was that parasites could come up through the floors because they were they were dirt floor huts. And so they came up with a way to integrate these some material into the into the floor of the hut, tamp it down and it made it an impermeable surface. So, you know, this was a case of, you know, thinking about what do I have access to in my own space that I can use sort of very locally to solve a very immediate problem and I think, you know, we talk about concrete actually the perfect example. The biggest cost in concrete is transport. And so we talked about this reformulation problem. And the opportunity that we have at Citrine is to say look, you know, Brazil's lime is different from the US is lime which is different from China's lime which is different from the lime, you know, dozens of types of lime over over the total area of Africa. And so, be able to respond to those and come up with the right materials for the local environment, reduce on transportation costs reduce on transportation pollution, and let's us create the right thing at the right moment, which, you know, I think there's a lot of difficult things about the on demand that we've created. But I actually think in the materials world the more on demand we can be the better, because it does really mean we have a lot less wastage. And so, hopefully I answered your question about the developing world it's a it's a fascinating topic that I'm starting to get deeper on it as to reflect. Yeah, then probably you hear this a lot. People were fascinated by your use of the term agility as kind of a comparative advantage type killer thing and moving into transfer learning in particular. There's a whole simplistic version of the question about agility which covers a couple of others and that is in your view and experience is it is it is it about the people. Is it about the technology how you use the technology the business models you use all the above none of the above or something completely different. So, so I mean I think business model is sort of a price of admission, you have to figure that out and honestly we get pushed all the time and whether ours is the right one because everybody looks a pharma right everybody wants you to patent the billion dollar thing and then go sell the billion dollar thing. And it turns out our industry is pretty different from that and their, their, the path before us was littered with sort of the corpses, or still existing companies but but some companies sort of tried and failed to to build some kind of crazy industry around this stuff and we're a pure software company. So it's a very straightforward thing. The bigger question though, I think if you have a good solution your business model, you can sort of iterate with it. For us the problem is the culture of the chemicals industry, you know, things have been done, the way they've been done for hundreds of years, and trying to convince someone they need to change things. If you do it the wrong way just just insulting right it's like you don't want to go in and say, you guys are idiots because they're not. I think people who are the best in the world at developing new chemicals. I'm just teaching I'm just saying hey, you know, rather than trying to lift the car with your bare hands. I've got a jack. So why don't I show you how to use that. And people start to understand that this becomes a source of leverage for them so you know I always talk about people process and technology is sort of the three things you need to change. And I think you can't avoid any of them. And it turns out that we in Silicon Valley are kind of the best at technology and that that's what that's kind of where we excel. And we've, you know, my team and I have had to learn how do we help coach these companies to adopt new, new cultural practices new new processes so that they can adopt the technology in the most useful way possible. Terrific. Another similar question I think all of us at Stanford, who are not in computer science at times feel totally overwhelmed as in the board through line and Star Trek about AI data science and machine learning. I have a lot of friends over there and I think we would do they would do better and we would do better if we work collaboratively and it seems like you've definitely come to that conclusion and I and I very tangible meaningful way as opposed to me who mostly theorizes about such things so how how have you found that. So if I remember that part of your talk it was really about not just the data but the quote unquote expertise that we have by going through all our long years of engineering education. Could you could you expand upon that a little bit more and it have you been able to work in a collaborative way with data science, people, actually you would be the quarterback in between the data science people and the people who stayed for the PhD and materials or whatever. Yeah, so so that's exactly what we do you know I like to say we are not. You know if you want to go work for the most cutting edge AI company, it's probably not us there are lots of companies are doing really cutting edge stuff and if you want to work for the world's most pure materials company that's not us either. We don't make materials we're only a software company, but you know what I've found more and more is that nobody. I mean there are people who go into data science or just you know sort of chasing a paycheck. But in large I don't see that I see people who want to use these tools to have a great impact and you know for so for some people that's better advertising or better Netflix recommendations or whatever it is. But a lot of times people come to a talk like this and they say, actually I want to use my skills to to advance the world to change things for the better, and you know a lot of times that core motivation really creates a nice bridge between the, what I call computer science technical people versus the chemistry technical people, and I think, then then it's just a matter of teaching them to facilitate a discussion with each other. Teaching them the same language so they can actually communicate and once we do that. It's usually just beautiful harmony bumps here and there but but I think in general it's, you know, people want to have impact, you know that's what this generation is all about and I think, as a scientist, an on ramp into doing that in an industry where where a lot of people overlook what an impact it has really motivates people to do a lot of cool stuff. Okay, I remember like it was yesterday and talk I saw with the entrepreneurial thought leader series are kind of sister Wednesday seminar by young guy probably not much older than you named Jeff Bezos, who said the reason people think we're going to fail almost immediately like Barnes and Noble is I think we're a book company. There's no way we're a book company we're an IT company and they have no idea that that's what that's what our game is. It's interesting very precocious probably for him for sure and probably for you as well. Actually we're kind of running out of time so as a transition as I usually do I wonder, you've done a lot of this during the talk. Do you have any last words of wisdom and advice to give students who are interested in getting into your part of the energy and sustainability future of the world. You know, I'd say two things the first is, you know, for those of you who are who are interested in materials who are interested in data science in materials, come and talk to us. You know, this is a this is an exciting area and we want more great people that you know come work says my that's my shameless recruiting pitch, you know, we're hiring all the time. I would say, you know, if you have an idea. First of all, I did not come into this believing that I wanted to start a company. Actually, I've done I've worked with relatively small company had suffered the pain, and I honestly went to had planned to go work for one of the big tech companies. And what I realized was, I had a calling to do this. And I realized I had an insight right this boundary problem of how do you do AI and materials at the same time. I saw something a lot of people didn't see. And so my my exhortation to you would be, don't just go try to do something because it's there. Do something where you really believe you have a different perspective than the rest of the world because that's the biggest opportunity to have impact and I mean, you know, yes, I would love to be Jeff Bezos at least the money aspect like you know, I'm not a person on earth but you know, honestly, he found a scene that no one else had seen. And, and I believe we have to and so you know as you're thinking about how you can change the world. You know, it is my belief that it's these overlaps these boundaries that create the greatest opportunity and, and it's been a huge honor to be able to share it with all of you so thank you very much for your time and and john thank you for inviting me it's been Thank you so much we look forward to welcome you on campus as soon as we're able to do that. And actually on your comment about materials being exciting people didn't think that before the seminar I think they probably do now. So thanks once again for a terrific inspiring and impactful seminar. It's not the good work as they say we need you.