Endgame with Gita Wirjawan - William Chueh: 1 Hour of Pure Academic Gold on Energy Transition

Episode Date: July 24, 2025

This episode revolves around two big questions: How can we scale up renewables? And how does Stanford approach the energy transition?About the Luminary:William Chueh is the Director of Stanford’s Precourt Institute for Energy, Professor of Materials Science and Engineering, of Energy Science and Engineering, and Senior Fellow at the Precourt Institute for Energy. His research focuses on energy storage, particularly materials for energy transformation, including batteries, fuel cells, and electrolyzers.About the host: Gita is an Indonesian entrepreneur and educator. He is the founding partner of Ikhlas Capital and the chairman of Ancora Group. Currently, he is teaching at Stanford as a visiting scholar with Stanford's Precourt Institute for Energy; and a fellow at the Harvard Kennedy School's Belfer Center for Science and International Affairs.------------------------ Berminat menjadi pemimpin visioner berikutnya? Hubungi SGPP Indonesia di:⁠https://admissions.sgpp.ac.id⁠⁠https://wa.me/628111522504⁠Playlist episode "Endgame" lainnya:⁠Technology vs Humanity⁠⁠The Take⁠⁠Wandering Scientists⁠Kunjungi dan subscribe:⁠SGPP Indonesia⁠⁠Visinema Pictures

At Stanford, I'm constantly challenged by my students. My students would tell me, I don't like to learn this way. Like, this is outdated, or this is not useful. This is not impactful. Stanford realized many decades ago that our superpower is to scale through our students. Most of the students, when they leave Stanford, they often don't end up practicing what they learned here. But the mindset, this entrepreneurial spirit,

This desire to not only change textbooks, but also change the world, is ingrained in the students. So what is the role of academia in innovation? The dominant part of our impact is we have the unique opportunity to influence their mindset when they're still very shapeable. Hi, friends. Today we're graced by William Chu, who is the director of the Precourt Institute at Stanford University. Well, thank you so much for gracing our show. Gita, as a great honor and pleasure.

Thank you. Thanks for having me. You're one of the very few great products of education from all over the world, right? You grew up in Taiwan until early age, and you moved here to the U.S. Tell us about your educational upbringing. Well, thank you, Gita. This is going to be a long conversation. First and foremost, I really value myself as an immigrant to this great country, as many of my colleagues here at Stanford are.

I immigrated from Taiwan to the U.S. in 1993 at the age of 10. And when I came, I landed in Los Angeles. I spoke one sentence of English. It's, how are you? I'm fine and you. That's all I knew. And humbled by the opportunities afforded to me over the past 32 years to advance my career, to contribute to society. And it was really humbling that I started at age 10 fresh off the boat with no English skill in mind. I think that speaks to the ability for the country to nurture immigrants and enable them to really reach for the stars in terms of achieving

the American dream. I think that is one of the most unique things about this country. The role of your parents, how did your mother, your father shape your educational trajectory? It is that very notion of coming to the United States. You know, my father was my age when he came to the U.S. I cannot imagine myself leaving the comfort of my home to go to a country in which the language and the culture and the people are entirely foreign. So I think they made a huge bet and a huge sacrifice to leave what they had in Taiwan and to come to the U.S. with almost nothing in their hands. So the role of my parents really making that big decision to leave Taiwan and come here. And, you know, at the time, Tamwa has a flourishing economy already. It's well on its

way to be a fully developed country in the late night, in the early 90s. As I understood it from my parents, they moved here because they really see what's possible with the American education system versus that of a more prototypical Asian education system. So to them, that was the reward, is that my brother and I could receive an American education instead of a Taiwanese education. And I can't compare because I never went through the Taiwanese education. But that was their motivation. That's why they made the move. And I think they sacrificed hugely so we can have the career that we have today and be able to contribute.

You got into Caltech and then you studied applied physics. then you veered off to material science. Why physics? You know, I... Because it was the hardest? So my father was an electrical, it is an electrical engineer. My mother is a piano, teacher, a musician. So I come from a family of many backgrounds.

Both sides of the brain. And I've always had a fascination with science and technology. as early as an elementary school. And I was deeply inspired, and I think this points to the role of educators, by my middle school science and math teachers. I had a really wonderful science teacher who taught chemistry and physics, and I had a really great teacher who taught math, ultramal. And I think they enable me to see the predictive power of science and math to describe most things in nature.

The reason why I gravitated toward, say, physics compared to life sciences is because we don't have as good ability to predict the world with biology. We're getting there. But it's much more predictive in physics and chemistry. And I chose physics. I could have easily chosen chemistry as where I wanted to go, because I really appreciate that predictive power, because you can describe things, our physical world, with a set of equations, and you can predict the future very straightforwardly.

So that's why I studied physics. Taiwan has been a highly, highly successful. industrial nation. And it's been a backbone for some of the most sophisticated precision engineering and all that, right? Is that likely to sustain in the long future? And are you also in a camp that believes that for any nation to be developed, STEM, is crucial or the mastery of STEM for any kind of development, for any nation.

Guida, I think this is a crucially important point. Having a workforce that is both great at being creative and great at building and executing as important. I would argue, well, this is my personal. opinion, I think Taiwan is really great at building and executing, as you mentioned, is one of the top in terms of industrializations of any technologies in the world. I think where, say, the U.S. perspective comes in a great deal, perhaps to complement industrialization efforts in countries like Taiwan would be the creativity part. You know, how do we really think outside of the box?

And coming back to your earlier questions about, you know, how has a U.S. education changed me? I think in the U.S., we're really trained to be highly creative. We can ask any questions. We're not bound to textbooks. In fact, one of my main teachings to my students is that's rewrite textbooks. That's our job here. Not just to understand them, we have to do that, but to rewrite them. And I think the coming together of, say, the industrialization spirit of Taiwan and other countries

with the creativity of Silicon Valley and the U.S. and elsewhere, I think that's a very potent combination. And it's very complementary, I think. Not a single system has all the attributes needed. So I think this points an opportunity of partnership. You know, Southeast Asia, which is where I'm from, right, has underinvested, broadly speaking, with some exceptions in education, generally speaking, but more so than not in STEM education. And this has been manifested in how we've not been able to sort of like follow the steps that have been taken by industrial giants such as China in Japan, South Korea, Taiwan and Asia, right? I think, you know, Taiwan has been a real inspiration for many people in Asia. Now, how do you think that could be broadened as to include places like Southeast Asia to play a part,

whether it's tiny or bigger, in supporting, you know, the manufacturing capabilities of the world? and, of course, the demand for goods and services is a rule. Gita, I think the working fluid as people, end of the day. And I think enabling this free flowing of talent across borders is critical. Since we're talking about my home country of Taiwan, you know, in the 70s and 80s, you know, how did Taiwan build up. Of course, there was influx of Western capital, but there was also the returning of expats from the U.S. and from the West back to Taiwan. You know, there are many prominent example, Morris Chang, being one of them, the founder of TSM. And you yourself, Gita, have done

exactly the same thing being Western trend, but returned to Indonesia to tackle on big challenges. I think promoting this moving of people so they can have very diverse experiences and return to Southeast Asia. I think this is going to be a very critical step to create the incentives for people to return because, you know, I've always had a strong desire to contribute to the growth of Taiwan as something that's very much on the top of my mind even today. you're at the very core of cutting edge innovation at Stanford, right? How do you see Stanford or an academic domain being a catalyst for innovation going forward? And how do you see that being able to promote, encourage collaboration with entrepreneurship? I mean, Stanford has been known as one of the very first. few universities that have been able to basically spark up entrepreneurialism.

How could that be exemplary to other parts of the world? Spark up is exactly the right words to describe this. So what is the role of academia in innovation? Yes, we develop a lot of interesting technologies. We do a lot of cutting-edge research. But I would argue that is not the dominant part of our impact. The dominant part of our impact is we get to interact with talented people very early in their career, whether it's undergraduate students or graduate students. We have the unique opportunity to influence their mindset when they're still very shapeable.

So you ask, you know, how has Stanford being able to do this? You know, Stanford realized, you know, many decades ago that our superpower is to scale through our students. Most of the students, when they leave Stanford, they often, some do, but many of them don't end up practicing what they learned here. But the mindset, this entrepreneurial spirit, this desire to not only change textbooks, but also change the world, is ingrained in the students. So when you spend, you know, four, five, six years at Stanford as undergraduate or graduate student, I think the biggest take home is the mindset. Then decades later, you say, what did you learn from Stanford? It's not going to be about, you know, this physics, that engineering, it's going to be, well, I learn this spirit. And I think this recognition of our power and influences and our responsibility in education is to instill that mindset very early on.

Because when you do this 20 years later, it's a lot harder to change. So we have a unique opportunity and responsibility here. Do you see that as a need for changing the way or the pedagogical approach to learning within campuses? You know, I, what I want to do is, you know, rather than preach, I think just sharing, right? For example, I recently, along with several colleagues, taught a course on scaling sustainability solutions. We didn't go through any equations. We didn't go through any formula. We barely had any plots.

But what we did was we invited a lot of practitioners from the world to talk about how they got there, the experiences, with the goal of shaping the mindsets of the learners in the class. So that is one output we always strive for is what are the take-home mindset shifts students should have, right? In addition to the granular details that it learned. But I think having a clear view on the objectives of the class being changing mindsets, I think can be very powerful. And it's a departure also from the standard science and technology and math and engineering curriculum, sort of focuses on specific knowledge, and very much so, those are needed. But then at the end of the day, I think educators like myself really enjoy when we see students adopt a new mindset in addition to their specific knowledge areas.

I'm worried about campuses that are not necessarily as open-minded as you want them to be. And I think they're at risk of being left behind, right? I think education is all about being open-minded, about combining that force of preservation with the force of innovation. And how do you make sure that that stays, you know, that energy to be able to continuously combine the force of innovation and the force of preservation? Is not a structural challenge for any university? You know, with education, you start with the two most important ingredients, the students, the learners, and then the teachers. It's a bit of a chicken-and-egg thing.

I think at Stanford in places like Sanford, we have been very privileged to recruit some of the best talents. And these talented teachers and faculty bring in their experience as well, this risk-taking nature. So I think universities can really pretty straightforwardly modify its culture by bringing the right people as teachers. The flip side of that is also students. Students often come with an open-minded desire of wanting to be.

to learn. So I think the key is to harmonize the two. You have teachers who want to teach in a new way. You have students who want to learn in a new way. You put them in a place in a classroom together, and magic will happen. You know, I think that's really fundamentally, this creative approach to education is not structured. It's not teaching by the books, so to speak, but it's really just welcoming, okay, how can we re-envision education? So at Stanford, I'm constantly challenged by my students. My students would tell me, I don't like to learn this way. Like, this is outdated.

Or this is not useful. This is not impactful. And my mindset has always been, I want to embrace the best ideas. So as I hear these ideas, there's something like, well, this is a really good idea. I should embrace it. And I go ahead and make the change. So I take substantial amount of risk in my pedagogy. as well to make sure that I am taking my student's suggestion, the best ones, and trying

them out. If it doesn't work, no problem. I'll come back to what I had before. The iteration time of academia is really good. It's on an annual or quarterly or semester rhythm, so you can keep trying new things, rather than fall into the same routine. Well, we've talked about this before, but I'm just curious as to what your views are with respect to energy transition. And I come from a developing economy, which is somewhat slow in building

power generation capabilities. Countries like Indonesia and India, at the rate that they're electrified at around 1,300 kilowatt per capita, at the rate that they need to go up to 6,000 kilowatt, at the rate they built only India, 19,000 megawatts per year, Indonesia, 3,000, it'll take more than 100 years. What's the answer or remedy in terms of scale, which you've aptly pointed out? And what's the remedy in terms of doing it on an environmental, environmentally friendly basis? I think the energy transition is exceptionally difficult because you have to balance a number of things at the same time. You have to balance sustainability. You have to balance economics or affordability.

And you also have to balance security and reliability. And navigating or co-optimizing for all three is incredibly. challenging. Incredibly challenging. You know, I think when we see certain things not happening fast enough, it's probably

because one or more of those dimensions are not being met. So then the investments is not being unlocked. Often, it's economics. Not always, but often it is. When things are not affordable enough or competitive enough, it's difficult

to unlock the capital. You know, we need experts would say $15 trillion a year for the next 25 or so years to transition our economy to clean energy. We are now spending approximately $1.2 trillion. Now, as you know very well, the capitalization of the world is more than sufficient to deliver this. The capital is there. We have a very solvent system. But the capital isn't flowing at the right speed. And my feeling is that in many regions, it's because the economics doesn't make sense.

Now, in other regions, there are other considerations. Security, for example. Taiwan, my home country, is a great example. Security is a great consideration because of the geopolitics of the region. and it's simply not acceptable to give up security. So the transition moves slowly there, or not as fast as we hope. So I don't have a solution, but I think it's good to recognize the challenge as co-optimizing for these three things. And what are the opportunities to develop a strategy that achieves the right balance?

And I think that will enable the energy transition to move faster. So, for example, in the U.S., yet I do agree, building more natural gas with an optionality for sustainability, perhaps through carbon capture, is a good idea because it leverages our low-cost natural resource in the United States. As one example. So I think the key is we have to start somewhere. We have to recognize this challenge, and we need a portfolio of solutions to help us get there. It's unlikely one solution can deliver all three attributes.

at the same time. Well, capital is definitely in abundance, but metaphorically, it's in solid state. It's not liquid. It's not moving. It's not being distributed. We can melt the gold. We can heat it up so it becomes more liquid, and it flows into other places that needed much more than some other places.

But that's a structural challenge. You know, I mean, the 15 trillions that are needed and 1.2 is flowing. I'm not so sure if the remaining 13.8 on a yearly basis is going to get liquefied. You know, as easily as one might think. It just seems to be pretty static. Yeah. Right. You know, I think having success stories.

Yeah. of economic games will help ease investor concerns. For example, we have seen cases where big investments were made too early, and that yielded poor returns. And that's a bad example. Obviously, we have to keep trying. But in order for the whole capital necessary for the energy transition to be unlocked, I think we need more success stories.

So I think we should be much more thoughtful than we are today in deploying capital with an eye on higher chance of success. Because if we have too many failures, the world will come to think, well, this is being tried, right? Say tens of billions of dollars has been lost on this one project. The second one is going to be much harder. So I think one thing we're trying to do at Stanford is to really think about how do we, properly guide what to invest, right? Because investing in all of the above will also result in many failures at time. Now, I agree we should try, but we should think about the scale of capital.

So we can make many small bets, and many of them not working out as fine, but that's also a just our risk profile. So as we go to the bigger bets, let's do our best to make sure that they are successful. And this can be reinforced through government policy. It can be reinforced through good decision-making on the corporate levels. It can be reinforced through partnerships. But we need more of the big projects to be successful. If you look at traditional energy, the success rate is very high.

The return as an asset class is very stable. The risk is very low. But for new technology, the risk is up. automatically very high. So that's all to keep the capital in trunches. So then that we are putting small amount of capital into the very risky beginning, so hence venture capital. But that's also be mindful. As we scale, then we are adjusting our risk profile lower and lower. So as a whole, we see the aggregate of the capital delivering good returns. I think that's going to help, in my humble opinion, unlock the capital necessary. And we look back over the last 25 years,

There had been many successes, but there was also asking a few failures as well in terms of these large projects. If you're a typical coal fired entrepreneur, you make a lot of money at five cents or kilowatt, right? But if you're a geothermal, solar, hydro, nuclear, what have you, it's way above that. That's the break-even point. So how optimistic are you in terms of being able to bring cost down so that more and more people are going to be able to enjoy the economic pie? Because as you bring cost down, the entrepreneurialism is inevitable. And the risk profile will definitely change for the better, for everybody or for more people. how do you think that cost of 15 cents per kilowatt will come down in the next 10 years?

Will it come down to the 5 cents level anytime soon? I'm very optimistic on our ability to learn as we deploy solutions. I see there are two kinds of learnings. One is learning through high volume manufacturing. I think solar and batteries has been a huge success story there. But it could also be learning through iterations of megaproject. Fracking is a really good example of this, natural gas. And, you know, all of that happened over the past, you know, quarter of a century.

It's not that long ago. So I am very optimistic. Yet another example in terms of energy efficiency is LEDs. that took about 20 years and the cost came down and it completely displaced the incumbent. So I'm very optimistic. However, with that said, we should not rely on learning for everything

because what we ought to do is to bet on the most, the solution sets that have the highest probability of being competitive at the end. Because something that is not competitive, then we need to stop working on that. So in terms of an innovation structure, battery, it's very much a stage gate approach, is when you need to make that big investment, well, that's don't wishfully think that the cost will come down simply through learning,

right? But that's analyze it carefully. See where the cost floor is. See what drives the learning rate. So Stanford as a think tank has been thinking about this aggressively over the past few years, is how do we help innovators think about the promise of their solutions when you put in 20 years of effort. How do we allow the market forces to shape the learning? And, Gita, I want to also highlight a very important aspect. Someone has to pay for it. Yeah.

You have to finance the learning curve. So if you look at solar, who financed the learning curve? Was the German government through the feeding tariffs? It was the U.S. government. Subsidized a big time. electric vehicles, same thing, we're seeing sort of the earlier stage of that. Batteries, a different story. There was no subsidy to begin with, but there were certain beachhead markets that provided

financing at different cost level. So I'm someone who knows batteries very well. 25 years ago, lithium-mount battered was over $5,000 per kilowatt-hour. Now, you know, at the cell level, it's well. below 100, it's close to $50. That's 100x drop. But if you follow that learning curve, well, it went from 5,000 to 3,000, so someone had to pay for that. So that was the camcorder business. Recall those technologies, you know, so that paid the high price. Then it was laptops that provided, okay, that's maybe the $1,000 level, then came mobile,

phones. Then came smart devices, tablets. Then those markets were increasingly larger, but required a lower price point. And then it came the premium electric vehicles. Now it's almost more than 15 years ago. And Tesla was the market leader there with their very expensive cars. And that provided learning. And then came medium price, mass market electric vehicles one by one. Each of those product and those customers could pay an increasingly lower price, but at a larger volume. So battery is one of those examples of just having demand at every cost point. So it isn't just about the beachhead market. It's about multiple increasingly larger market that it can embrace the price points

as the technology come down. I think this is absolutely critical. Semiconductor is the same thing, right? You needed someone to pay for that very low yield initially before it became a mass market. And I think in connecting the technologies and the solutions to this stack of markets, because we talk about stack of capital all the time,

but we also need stack of markets to provide this driver. Somebody's got to take a view, right? And somebody's, And it only makes sense that a developed economy has got to take a view for purposes of everybody on the planet. Now we're beneficiaries of what the Germans did 20, 25 years ago with solar panels. We're beneficiaries of what the Taiwanese did, what the Japanese did, what the Americans did. Now, if you stack up or if you put on stage all these available technologies for energizing here, humanity, which one would you pick that would have the highest acceleration as to be able to

be applicable to large-scale developing economies, whether it's nuclear, hydro, solar, or you're in a camp that believes that for every bit of these choices, they run the same chance of being able to be attached to large-scale developing economies. I think there are many opinions on this. So let me share my opinion. Sure. You know, when it comes to energy, you know, we have electricity, we have fuels, we have heat, sort of as the three main forms of energy carriers. And my opinion is that that's focus on the one in which the demand is growing.

because when the demand is growing, things are easier, right? Yeah. It makes it easier to invest in the area. And if you look at these three sectors, the one that's growing the fastest today is electricity, the demand for electricity, thanks to AI and other aspect to drive up demand. I mean, we're seeing a renaissance in terms of the demand for electricity, not only in this country, but globally everywhere. So the pie is getting bigger.

That's an opportunity. for additional solutions to come in. There are folks because of demand supply mismatch. Now there are companies, countries, which are able to afford or pay more for energy in order to access it for electricity. So I think that represents an opportunity. Let me take a moment to share with you

some of the vision we have here at Stanford and our Energy Institute on electricity. So we have a vision that electricity ultimately should be clean 24.7. That's the goal. There are many solutions that can get us there. They include renewable plus storage, so solar wind batteries. That's one round. Another route would be advanced nuclear electricity. Yet another round will be natural gas plus carbon capture solutions. And the final one is geothermal as another form of base load electricity. So we have a lot of options. And depending on where you are in the world, depending on your

natural resources, you will compose your 24-7 electricity system based on your local resources and constraints. I think our responsibility are to make sure that we innovate, the lowest cost pathway for each one of these options, and to have as many of the partners in the world demonstrate and deploy them to drive down the cost. So I think we're in this unprecedented time where demand is skyrocketing. And that's a good thing. And I think we ought to piggyback off that and use that to finance the learning curve, use the appetite of our large companies that are in the,

this area in the hypers, the data centers, the compute-heavy companies, they are willing to finance much of this. And that's take advantage of this growing demand. What's your take on nuclear? I think nuclear, as I mentioned, is one of the four solutions that we can embrace for 24-7 clean electricity. Now, nuclear is not without challenges. We have to think about nuclear waste.

We have to think about in terms of safety, not necessarily from operation, but from a proliferation perspective. And these are important questions. And if we are able to develop solutions to address these two core issues, I think the acceptance of nuclear will increase. Let me give one example. For instance, can we decrease substantially the cost? of reprocessing to enable a circular economy for nuclear. That would be incredible. If you do nuclear fusion, you automatically get rid of that problem.

Yes, it's very hard. Yes, the probability of success is unknown at best, but I think we should give it a try. So there are pathways for nuclear to be without its safety in terms of proliferation and also in terms of waste. So I think that's a huge opportunity for innovation. And I think developed countries like the United States should very much be investing for this because this will benefit the world in terms of global security as well. And by way of the modularization for data center purposes, it just seems intuitive that it

can be managed from a risk standpoint a lot better than the past. Yes. And I think it's a complex topic, because when you have many, many small modular reactors, this can also make the security aspect a little bit more challenging as well in terms of providing physical security for the sites when they're highly distributed. So I think it's not a clear-cut answer on if SMRs would improve security of nuclear. What's the realism of fusion being in place, in many places around? on the world? I think many are trying. I think many are trying because the opportunity, if successful, and the impact is extraordinary. So I think Fusion is one of those truly high-risk,

high-reward things. And I think we ought to try it. Now, this is not to say we don't do the high-reward low-risk. Those are things we must do. There are many solutions that we already have today. Solar, wind batteries are all high reward, low-risk options. Fusion is high-risk, high-world. So I think balancing our portfolio of solutions in terms of our research and development to tackle many of those, I think makes a lot of sense. You know, I, you know, to come back to my childhood,

I worked on nuclear fusion as a high school student. I was an intern at General Atomics in San Diego as a high school junior. And because I was interested in physics, I worked with a scientist there to develop sensors for measuring the temperature and the velocity of plasmas in a tokomac reactor. So I have a tiny bit of understanding of nuclear fusion in my formative years. I think the challenge with Fusion is while, oh, let me put it this way, there are clear signposts for Fusion. I would look out for them, right? So there are clear milestones. If we achieve them, we know that's a big deal.

And I will be closely watching them, entrepreneurs, startups, large corporations, which are building these things out. But as I mentioned, if successful, I think Fusion can be. tremendously, it will change the equation completely. You know, with the way to geopolitics of things changing the way it is as we see it, how will that impact proliferation? Or you think it's actually easier now because it seems to look like multipolarity on a grit where you can identify who's likely to be your friends, who's likely to be your rivals,

who's likely to be in a swing state, so that you can better manage the geopolitical risks and the proliferation risk when it comes to nuclear. I'm not an expert on the topic, so I would just offer a high-level thought here, which is this security component to the energy transition could not be overstated. I think it's probably as important as economics.

As I mentioned in my home country of Taiwan, security is the dominant consideration. Taiwan is willing to go back to coal for the benefits of security, for example. So the value of security is very high. But with that said, there are also solutions that can offer security and sustainability and affordability simultaneously, solar is one of them, right? It can be extraordinarily inexpensive. It can offer security by having domestic energy production,

and it can offer this S-scale today. So I think the challenge is this option is not available for everyone, depending on the natural resources with the sun being one of them. them in each region. And this also speaks to the importance of a Jensen technology like batteries. So batteries can really complement solar by making it available 24-7, but right now batteries is too expensive. So for that purpose.

So I think there are a lot of levers that we can pool to simultaneously increase security, affordability, and sustainability at the same time. But we need more of those. So I would say technologies and solutions that can hit all three at the same time, we should give them as much attention and capital as much. And then also recognize where are the gaps? Where do we need to invest to make sure that more of these solutions are available to the world in regions which say don't have as much solar resources?

You know, if you use Chad GPT, simple search. requires 10 to 50 times more energy as compared to a simple search on Google or any search engine. Then if you try to create a sophisticated AI generated image on SORA, it requires up to or at least 10,000 times the amount of energy compared to a simple search on a Google platform. Then if you take a look at the queue as to enter the grid in the US, We're looking at about 2,600 gigawatts, right? On top of pre-existing 1,300 gigawatts, you aptly pointed out AI. Even in a developed economy, like the U.S., there's seemingly not enough energy.

It's going to get amplified in a less developed economy, right? Doesn't it seem that energy is the structural impediment to the degree to which humanity wants to AI itself? Energy is fundamental to prosperity. So is information. So I completely agree with your viewpoint, Gita, that the two are completely intertwined. If you want more information, you need more energy. If you want prosperity, you can need both. So they're all connected deeply.

Let me take a moment to comment on energy efficiency. It's one of the underappreciated levers in energy. So we have the supply side. So we talked mostly about supply side, nuclear, solar, wind, batteries, those are supply side. But the demand side can also exhibit fast learning. If you look at AI, if you look at how far energy efficiency has increased, sort of the watt hour per compute. It's dramatic. I have no doubt that efficiency of AI algorithms and hardware will improve dramatically over even just the next few years.

And the reason is simple. It's driven by market dynamics. Energy cost is the highest operating cost of AI. So corporations are very eager to drive down cost. So cost of energy is going up. So the only way to really get there is to look the demand side is how do I use this energy per product I sell? Because for chatyPD, they are basically selling electrons. So I am extremely optimistic.

Research at places like Stanford and in companies are all tackling this energy efficiency computing as a core problem. And when the energy efficiency goes up, you're also going to see the profilitation of AI in everyday life, because then it will be cheaper and cheaper and cheaper to use it. And that's a great thing for human prosperity. So, you know, I encourage us to look at both the supply and the demand side at the same time. So that's increased supply, decreased demand. And I think that's the recipe to having the energy transition.

So you think the data intensity is going to be caught up by the energy intensity and will be caught up by the energy efficiency? I think that's to level supply. I think they're all levers to stabilize the system. Right. And we need to invest our resources to pull on each one of those levers. And I would say, at least in my limited view as an academic, I think there is a greater focus on the supply side than that. the demand side, but we're changing this. Because the demand side, the opportunities sometimes can be extraordinarily large. And often on the demand side, it's drop in. It sometimes can be as

easy as behavioral changes or modifications of how you design buildings, how you design factories, without using any new technologies. So I think this is a great lever that we need to pull even more so. I want to get to batteries, which is your area of expertise. But before we get there, talk a little bit about the parity between internal combustible engines and electricity. How is that playing out now? Well, you know, we can look at parity in different and several different aspects. Yeah.

In terms of performance, right? So fuel, say, octane has incredibly high energy density. This is why it's so challenging to replace, say, aviation fuel with batteries. Because the energy density there is just simply not. There is a more than 20x difference between fuel and oxygen. This is a fundamental challenge because with the fuel, you just have to carry the fuel. But the batteries have to carry a lot of other things as well. That makes it heavier.

That makes it more voluminous. I think that's a key difference in attributes. We have to recognize it. That's the strength of fuel is that it is very high energy density. Of course, on the sustainability side, batteries, in terms of its operation, is by virtue of the chemistry, it's carbon neutral, right? Because it's a closed system. But that's not always to forget.

There are carbon emission. in the creation and the manufacturing of the batteries. And now studies have shown that the payback time for things like batteries in terms of its carbon footprint can be on the order of just a few years. So this is already making strides. The final part is economics. So this is probably the biggest differentiation. So if you look at the cost of the fuel infrastructure plus the internal combustion engine,

it is considerably more competitive than battery technology today. So I always say this is a driving force. So we need to innovate in order to make batteries less expensive in order to reach that parity. But right now, we are not at parity yet. Finally, let me also mention in terms of economics, it's not just the economics of batteries. It's the total cost of ownership at the end of the day. This is where electric vehicle has seen some great successes. With EVs, because it's based on motors, then the maintenance costs is lower.

And this is something that the consumers can feel. So I also encourage folks to think about it, not just in terms of internal combustion engine versus a motor plus batteries, but that's view at the level of a car. So over the lifetime of the car, what's the expenditure? But that's not only go in the car itself, let's look at all the mining we have to do, life cycle of the whole car in terms of environmental footprint. And so far, the analysis are showing, this is highly promising, that when you start integrating over the lifetime from cradle to grave, that we are indeed getting closer to

parity. So I think these are where things stand in terms of internal combustion engine versus batteries. It's getting very close if you look broad enough, but there's still need for further improvement in terms of economics. I think in terms of energy density, there will always be applications in which it is required. And it's going to be very challenging to overcome that intrinsic advantage of fuels with

this very high energy density. Where's the level of parity? Is it like $100 per kilo an hour? Again, it's challenging to think parity in that respect because it's not necessarily a system level comparison. As in not all in cost? Yes. So one way to think about it is the cost per mile.

What's the levelized cost per mile on electricity versus internal combustion when you consider the cradle to grave cost? So that parity is complex. I don't have the numbers, but I think that's how we should be thinking about it. And are you optimistic? I'm optimistic. Within like a few years or decades.

As a scientist, I always say, it's 10 years away. So I'll refrain from making predictions. Okay. Well, that's, that's reassuring, I mean, encouraging in the sense that, you know, there's hope. Batteries, right? You've been teaching batteries and all that. Talk about the evolution from nickel, lithium, sodium, so that the lay people like me could understand, you know. know, how things are going to pan out in a near foreseeable future or even distant future.

Yeah. I think the evolution of batteries has been very similar to the evolution of semiconductors. So, semiconductor has always been about the density of computation. And likewise, for battery, it has been a density of energy. So you can follow the trajectory of batteries over the past 100 years or even more on how the energy density has progressively increased. And that's what enables all these new applications of batteries. You know, if you remember our big phones of the early 90s,

and that uses a technology known as nickel-metal hydro batteries. And a step change was made when we went to lithium, mountain batteries in the late 90s, when Sony commercialized lithium mountain batteries, then they got a lot smaller. And that's what it enabled camp quarters, the handicams from Sony. Because the battery was no longer huge, the battery was small, and so on and so forth. And that's what we have the iPhones we have today. That's what we have electric vehicles that we have today. So the evolution can be more or less mapped as an

improvement of energy density as the primary attribute that tells you whether or not you can use a battery for a particular application. Now, going forward, I would say energy density is one consideration, because now batteries has reached all corners of markets and application. We also have to think about cost. Because when you're looking at these smaller markets, maybe cost is not so dominant. You know, for a smartphone, the battery costs only a few percent of the total device buffer in an electric vehicle is many tens of percent.

So the sensitivity of the total system is much higher the batteries going forward. So the second curve to look at will be the cost curve. And then it's going to be a very interesting trade-off between costs coming down and energy density going in. In semiconductor, we basically saw both of them as the density went up for compute the cost came down through innovation. And I think battery will go in the very same direction as well, in which we have improvement in both.

Some technology will improve costs more so than energy density. Some of them might trade energy density for costs. But fundamentally, what is a breakthrough? A breakthrough will be you do both at the same time. You have higher energy density and you have lower cost. At the end of the day, that is the holy grail that we're working. looking for, that you can do both. Most can do only one out of the two.

It's expensive, but very high energy density. Or it's not expensive, but it's low energy density. So while it's good to have different technology for different market segments, the holy grail would be one that can do both. How is sodium better? So sodium is a very interesting case study. So today's battery uses lithium as the mobile ion. sodium is much more abundant, much more inexpensive compared to lithium.

But, generally speaking, based on where we are today, it is substantially lower in energy density for sodium. So with sodium, the task is clear. Increase the energy density while keeping the cost. Now, let me also highlight another higher order consideration. when you start building big batteries, for example, battery at the level of the electricity grid, actually energy density feeds into the cost. They're not decoupled.

So if you have a low energy density chemistry, that means your site for a big grid scale storage has to be bigger, which means your engineering, procurement, and construction costs would be higher. So there's also coupling between the two. So it's actually a quite complex interplay of performance in terms of energy density and economics. But, you know, we're very excited to be developing chemistries here at Stanford that could potentially do both at the same time. Low cost and high energy density. In a photosynthesis context, right, you talked about energy being created by the light, energy being created by the light, energy being created by the heat. Are we now able to optimize on those two? Or we're still far from where

you think it could be? You know, photosynthesis has the attribute of very high energy density, because you're making glucose, for example. The challenge with photosynthesis speed or power. So the power of photosynthesis is very low. The energy is very high, but the power is low. It takes a long time for plants to convert sunlight to glucose. So it's a power problem for photosynthesis. And engineers and scientists being working very hard on creating artificial photosynthesis, but they also confront the same challenges. Energy is very high, but the power is very low. So the opportunity there is, how do we speed it up? And how do we greatly accelerate and go beyond the rates of traditional photosynthesis in biology and create artificial enelot that can do so

at a much faster rate in terms of power. And that remains at frontier topics in chemistry and physics. Intuitively, it just seems that solar is the cool way to democratize energy so that you can put solar panels on tops of huts and villages, houses, anywhere, right? Do you see that as the real democratizer of energy for most parts of the world? I think it's a key component. So, you know, let's think about solar. Thanks to riding the learning curves, solar costs have come down tremendously, more than 100x. And that is leading to endless application for solar.

Now, the question is, to further increase penetration, what needs to be true? Well, number one, it needs to be even less expensive than today. But now we're hitting a bottleneck. Solar sales today is a very small fraction of the total cost of solar systems, right? In some cases, it's 10%. You know, what is making solar remain at its point? price it is actually in places like United States, it's insulation. It's the power electronics. That balance of system, that non-solar cell component is overwhelming.

So that is an opportunity. You know, scientists like myself, we think, you know, as chemists and physicists, we want to make better solar cells. But if you talk to people in the solar business, they will tell you, just make insulation less expensive. And that would help as well. So I think there's huge opportunities to innovate solar as a system. And by the way, not only limit it at the technology level, but also to include business level. I learned recently that the origination cost for a residential solar in the United States is on average $10,000 per household. 10,000. Not cheap.

Not cheap. And that's just a business cost of origination. So I can bet that there are many startups, and I know a few, trying to bring down the cost of virgination. Let's begin to think at a systems level, right? At the end, it's not just the component. Solar cell is just a component of a solar system. And by the way, if you want solar to be available 24-7, you need to include batteries as part of that. So I think we can begin to start to look at the system and find the biggest levers to drive down,

the cost. I think when we succeed in doing that, that will help democratize energy, because then we're able to deploy solar. And speaking to that point, there may be specific or region-specific levers in developing countries to perhaps make use of local labor and other constraints, so then that we can use that to pull down the cost at the system level faster. And this is maybe where policy can really come in locally in order to drive that. But to say that, you know, solar has reached its floor is not true. You know, solar cells may be getting close, but solar systems long way to go, right, in terms of how that system can come down and cost.

Same thing for batteries. You know, battery cell today, if you look at the bill of materials, you know, lithium ion batteries probably can get to maybe $35 a kilowatt hour from today's 55 to 60 if manufacturing was free, right? That's the best case. So that room is finite. But today, the battery system cost is not $50 a kilowatt hour. It's more like 250 or even 300. So that means the battery sale costs is also just like solar, a small fraction of that. So innovating on balance of systems, power electronic, safety, those are all things that adds to the bill of material at the system level, the construction cost, the permitting cost. Those are components.

So I'm very excited to see how these stack of solutions can then work together to deliver the economics. It's not just the deep tech itself. I think that's one component of it, but there are many things that need to be added on. And as history have shown us, at the end, it's not typically one thing that seceded. It's a combination of things that stack up together to drive the economics down. So I think that's a huge opportunity. To the extent that you take into account the all systems cost for solar, that it's still way up.

I didn't know it was $10,000 for origination purposes. In the U.S. For residential. Yeah. still a lot of money for the American economy, right? And to the extent that it actually becomes a democratizer, what do you think would happen to the pre-existing grid? I mean, if you don't need the pipes anymore,

you can just latch onto, you know, the roofs of houses or even big buildings, what do you think will happen to the pre-existing grid? Will it be of still utility? Because there are other sources of energy? Edith, this is a good question. The penetration of renewables will continue to grow because of the economics.

I think this is inevitable. And this is a great thing. As renewable penetration increases, it places much more stress on the green. grid because the grid wasn't designed with bi-directional energy flow. The grid was designed for unidirection flow. You generate the electricity and you ship it to where it is needed. So that is one challenge for grid operators today is they're dealing with the so-called duck curve

in terms of turning up and down their generation assets because of the renewables like solar and wind. So that's a challenge for the grid. Now, will everything just become what people call behind the meter? So everybody have their own power system. Certainly, data centers are thinking of this. I think we are seeing some evidence of that. But remember, all the resources being disconnected from one another also prevent the flow of energy from one asset to another. and in the presence of increasing extreme weather conditions,

I think the ability to move energy quickly is also very critical. Let me give sort of a pie-in-the-sky-type idea. What is the complete opposite to solar plus battery as a distributed energy resource? What would really sort of oppose that? It would be a lossless grid. If you have a lossless grid, then you don't need batteries anymore because then you can move energy very quickly across the world, right? From the Sahara Desert to wherever you need to be.

So, of course, that's a very long shot. And I do think that's something we should bet on as a lossless transmission grid. But that is the opportunity of the grid, because it allows us to shift the energy very quickly to mitigate breakdown of the local electricity system. So I think in the future, if there was no grid, you're going to have substantial reliability challenges. Because then everybody needs to count on their own generation assets locally to provide it.

But there is huge opportunity there for innovation as well at the intersection of the grid and these renewable generation assets and everything in between. How much would it cost to amend the pre-existing grid as to make it bi-directional? The grid is, by definition, can be bidirectional. It's just not optimized for it. I would say the total sort of cost predicted for renovating the U.S. grid is about $3 trillion. So it's a substantial amount. But the time is actually quite appropriate now to think about with the increased penetration of renewables, with the increasing demand of energy,

our grid is being pushed to the limits. Stanford just published a report with six big ideas on how do we build the grid of the future. It's going to have tremendous benefits. is going to create many, many local jobs because this is building infrastructure. Think of this as building roads and highways. It's one of the same. It's a highway for electrons. So I think there's a huge amount of economic benefits.

And the reason why I think now is the time to do it is we're seeing the first time major increase in demand in a long time. So that's harness this increase in demand. That's used this opportunity to invest in the grid to make. to make it more amenable for renewables, to make it easier to move electrons, to make it easier not to have congestion, just like highways, you can have constructions on the grid. And actually, these are leading to tremendous inefficiency in the system now.

So I think over the next few decades, we're going to really see a lot of very interesting technologies to transform our grid to the grid of the future. For developing countries, I think it's a huge opportunity because much of the grid, for example, in India, has not been built. So there's an opportunity to incorporate the latest power electronics, the latest control systems.

So then we have a grid that's highly flexible, highly efficient, highly reliable. So I think this is indeed an opportunity for the next few decades. I skipped a question earlier. The Japanese were quite visionary in the 90s, right? They came out with the idea of electrifying cars. but to the state, they've shied away.

They seem pretty dogged with hybrid and hydrogen capabilities. What explains that posture or behavior? In terms of hybrid, the thesis is really simple for hybrid, which is if you only drive a short distance, why carry a very heavy battery because it's not being used most of the time. So for many markets, it actually makes perfect sense. I'll give you a personal example. I have, in my opinion, a highly sustainable energy system for my life.

I have a fully electric home that has a 20 kilowatt solar system. I operate only on heat pumps for my heating and cooling means. induction stoves. I have no natural gas use in my house. However, for my mobility needs, I have two cars. One is a battery electric vehicle, so only batteries, and the other one is a hybrid, which has a 30-mile battery and a 400-mile tank. So again, this is just my own experience in my own life, why do I have those two cars? Why didn't I go with two electric cars?

That's because of my lifestyle. I drive very short distances. I only live a few miles from here. And a hybrid is perfect for that purpose. I don't need to carry a big battery. I carry a small battery. And then this car I can also drive to see my family in Los Angeles without having to worry about charging.

So that's why I have a hybrid. So I think it depends on your use cases. And the market, it is not at all an ore. It's more of an ant. So depending on your needs, which is region-specific, you will have different needs. So I think there is a place for both battery electric vehicles and hybrids as well, depending on use cases. And I know many of my friends and family who have both, as I do, to enjoy the benefits of those. two solutions. And it depends on different countries, too, a different region, different temperature,

all will induce in different solutions making sense for the region. So I think at a higher level, when it comes to technologies for energy, I very much like this concept of having multiple solutions because it's not a one-size-fits-all. And it certainly is not the same solution in Southeast Asia and in the U.S. is a different solution. It will not be the same. Would this warrant a long-term coexistence between fossil and renewables? I think fuels will be a very critical part of our energy system in perpetuity. Because as I mentioned, the fundamental advantage of fuel is its energy density.

I always say the challenge, the challenge for scientists, engineers, entrepreneurs, corporation is, how do we enable a net zero future with fuels as an energy carrier? Because you need it. So I see several pathways. One pathway would be think about how to dramatically decrease the greenhouse gas emission of fuel, both in terms of making it and using it. So for instance, we can be thinking about carbon capture or carbon capture and utilization. or in storage solutions. We can have a circular economy of carbon dioxide, for example. We can store them underground in a doable fashion.

So I think that's one possibility. The other possibility is to develop carbon-negative solutions. For example, through biology. Plants are carbon-negative. So there may be also opportunities to combine them because at the end of the day, we want a net-zero energy system or a Nessero system as a whole,

but there can still be emissions. I don't see any way around it. For long-distance ocean freight, long-distance aviation, it's very challenging to see that being electrified, for example. So you will have to use fuel, and you will emit it to the atmosphere. So there is where carbon-negative technologies or developing sustainable fubel could be important.

Now, this is highly challenging, because economics matter a great deal, because the incumbent fossil fuel is extremely inexpensive. So when we talk about things like sustainable aviation fuel, we often miss the most critical part of the conversation. How much will it be? Because if you talk to companies, they will be very hesitant to pay even a small premium,

economically speaking. So I think this really speaks to the opportunity and challenges, but fuels will be here in perpetuity. We need to be thinking about how do we have fuels in a net zero system at the end of the day. I think that's the right approach of thinking about it and much of our efforts here as Stanford is pursuing that goal. Very clear. Well, I'm going to ask you the last question.

I know you've got to go. we've talked about how abundance of capital hasn't really translated into proper distribution of capital. And we're on the verge of seeing commoditization of intelligence on the back of artificial intelligence, commoditization of labor, on the back of robotics. to the extent these two get commoditized, we're likely to see abundance, right? In these two commodities, what do we need to do to prevent this commoditization from not being distributed as a public good? I mean, capital is a public good. It hasn't been distributed fairly as much as it's abundant.

So I think risk managers and capital allocators need to figure out a way to better reallocate capital, right? Then if we're going to end up with massive amounts of artificial intelligence, massive amounts of labor on the back of AI and robotics, how do we make sure that this abundance is translated into better distribution for a bit? humanity's sake. You know, I think fundamentally, probably, you know, the single most equitably distributed thing right now, and it's not perfect, but I think it's the best, is information.

I think information through the advent of computing, social media, reaches all corners of the globe. Now, some regions better than others. I think that's a good example to look at. So I don't have a solution, or I don't have a good way to think about it, but I think having energy, robotics, AI, reach a similar level of equitable distribution as information is the goal. And by the way, they're all tied together anyway. So I'm quite optimistic in that.

And the reason why I think information has been disseminated so easily is because the cost of access information has dramatically decreased through, you know, the single biggest innovation of all times, the Internet, right? So that enabled it. And of course, the Internet thrives today because it makes economical sense. So information flow is a very economically effective medium and it helps the economy thrive. So I think if we can bring that same mindset to energy, to AI, to robotics, I think that's going to be great for prosperity of the entire planet. I think this is a hugely important goal. And Gita, I'm so grateful for your leadership in driving this important conversations. Thank you so much, Will. Thank you, Gita. It was a great pleasure to be here.

That was Wuchew from Stanford University. Thank you.