Ep1171: Fernando Dominguez Pinuaga: Quantum Meets AI

If you understand, store now, decrypt later, which means that nation states are stealing our information. They are storing it. They can't read it now, but they are holding it until they have a quantum computer ready. And they will read it in three, four, five, six years when they have the quantum computer, which means our information today is in danger today, not tomorrow when the quantum computer is ready, but today.

So the governments and major CISOs and companies that understand this, They say, I need a plan to protect myself against that. But then they use the tool and they say, hold on, your tool is helping me not only with that, but it's also helping me understand my classical encryption today. Because infrastructure is huge. And there are many instances of your infrastructure that are across everywhere and globally.

If you've done a lot of M&As, you're acquiring infrastructure and have in cryptography that in many places that you don't know you have it. or you don't have cryptography, one in five breaches come because of a credential that was lost or not updated at the right time. So that tool is helping you monitor and modernize today and preventing you from the future. All right, Fernando, thank you so much for joining me here on 20 Minute Leaders.

How are you doing? I'm doing great. Thank you, Michael, for inviting me. No, thank you for taking the time to do this.

You guys are on a big mission and you're up to big things. And so taking time out of the day-to-day to make this happen is fully appreciated. Fernando, you're holding really interesting work today with Sandbox AQ, a company that I've been following closely for quite some time. In fact, my dad is an advisor within the company since its founding.

and Jack Hittery is a phenomenal being. And if the two of you are working together, then that means that you are a phenomenal being. And so really excited to have you on here. Thank you again for joining me.

Tell me a little bit about yourself and how did you even find your career trajectory ending up today with Sandbox AQ? Sure. Well, first, your dad, Yossi Matias, he's more than an advisor. I feel like he's an inspiration.

If we can do half of what he's done at Google, in the last decades with Soundbox, that will be a huge success. Wow. My journey is I started industrial engineering, and I studied in France. I ended up working in the energy industry, which today is becoming very relevant again with everything that is happening with the data centers and high-performance computing.

And then I ended up somehow at Google Latin America, where Google was expanding in Latin America, helping Google products expand. Some of those were Google Maps, crisis response, Google News, YouTube, advertisement. That was a time when digital advertising wasn't that known yet. So we had to teach everyone about why they had to be in the digital world.

When today it's so abused, back in the day it was like, this is why you have to be on it. And mobile phones were not that on yet. but I wanted to go back to my roots of engineering and I started helping Google X projects that were projects that were not aligned with the main business of Google like the self-driving cars, Project Loon that was connectivity with balloons. I joined Google X for an energy project and I spent six years at Google X working in many different projects from robotics, machine learning or AI when it was called machine learning and many different projects in robotics, logistics, climate change, agriculture, all projects at Big Thinking.

That was the, if I have to say, the thing that I learned the most at Google X was think big. You have a big impact. Let's go. And you know, Fernando, just to pause for a second, I actually just saw, I looked it up and I interviewed Mike Cassidy for 20 Minute Leaders five years ago, four years ago and 11 months.

So December of 2021. 2021 or 2020? Five years ago, wow. Yes, I never worked directly with him, but I crossed paths with him.

He was leaving when I was arriving. He was one of the ones that started and launched Project Loon as well. Exactly, yeah. So think big.

So you learn to think big and you are working on these huge projects that are, you know, all of them are huge. Each one that you mentioned. Yes, and in that cross path, I mean, it doesn't sound that obvious, but thinking being is one of the things that I see the most at Silicon Valley and also in the American culture. It's like, I'm going to build something that is really going to change the world, which is something that other nations, other cultures don't really have embedded into them.

And also that you can bring your idea and nothing is considered stupid. So you can put any idea on the table and you know you're not going to be considered as a fool because you're thinking too big about it. The best example is let's go to the moon, Kennedy, in the 50s and 60s. So that's a moonshot.

That's the main mantra of Google X. Let's think 10x. And Sergey Alnari, you probably know this, they always push you to the limits. When you say I'm going to do 20 of this, they're going to come back and say, why not 200, 10x?

Absolutely And so how did that shape the way that Sandbox AQ got started here And tell me a little bit about the story of Sandbox The story comes from Sergey hiring Jack Hidari our CEO as an entrepreneur in residence And he started bringing more science into software engineering, more physics, more professors from Stanford, from Berkeley, coming and teaching methods and techniques that were used in academia. And then started inviting all software engineers to listen to those techniques and try to use them out there.

Neural networks, CNNs, and tensor networks, all these techniques that today are like some of the principles of AI. Back then, this was 2015, maybe 2017. Attention is all you need. It's a paper of Transformers that was published in 2017.

So that's when, that's how Sandbox started, more as an educational platform to bring science into old Google products. And then we started converging AI with quantum physics. When you think about physics, you think about quantum physics, and you think about all the techniques that have been developed in academia to try to understand that field. Now, can we take them from academia and use them to solve real problems and application problems today?

And we started converging those two worlds. That's where the A and the Q from Sandbox AQ come from. It's converging AI and quantum. Today, it sounds more obvious.

You listen and more and more quantum AI or AI and quantum. Even Jensen from NVIDIA has gone from, in January at CES, saying that quantum computers were 15, 20 years away, to radically change, invest in three or four of the top companies in quantum and AI, including Sandbox AQ, and launching, last week, they launched the first hardware product for performing hypersummon computing with quantum computers and classical computers. So in nine months, NVIDIA has stepped up and said, these two technologies are going together.

That's how we started six, seven years ago. We worked very close with the quantum computing team at Google in Santa Barbara. And they were doing an excellent job. And we say, we're not going to be working on quantum computers.

Our quantum technologies are much more than just quantum computers. We're talking about quantum sensors. We're talking about simulation. And we're talking about cybersecurity and communications.

And that's all the areas that we discovered at Sandbox AQ. And it was so big, so relevant. I would say we need independence and autonomy to go faster and capture this web. Yeah.

And what are the different use cases where the convergence of AI and quantum computing really, you know, are, you know, I know that they're going to change almost everything in the way that we think through running computationally difficult problems. But if we actually look at these next couple of years and the world in which sandbox is playing at today, how do you see the big use cases shaping up to be right now? That's a good question. So let me show the first one, which includes the word quantum, but doesn't have that much of quantum, which is protection, cybersecurity, and a threat.

Quantum computers mean a new way of computing, a new, different way, not only faster, but different, which means that exposes encryption as we know it today. RSA, and in 1994, Peter Shor published an algorithm that said if one day we have a quantum computer, this algorithm will break encryption as we know today. Back in the day, we didn't have a roadmap for quantum computers. Today we do, and we have many players and a lot of investment on achieving that roadmap to achieve the 1 million qubits that we need.

So now we have the threat, and we have the solution to perform the threat. But to protect yourself against that threat, you don't need a quantum computer. You just need to change the algorithms that you're using to encrypt your data. And that's what we call post-quantum encryption, PQC.

And we discovered by trying to apply this that first we need to modernize encryption. Modernizing encryption means like what is being done manually today, looking for the encryption, the credentials, the PKIs that you have inside your entire infrastructure is being done manually everywhere because encryption has never been a priority in most organizations. so that's a lot of what we're doing with the department of defense and with many of the banks and many different players across the the globe it's modernizing encryption and right now that we have the inventory of all the encryption that you have in your infrastructure which ones are classicals which ones you have to update and create the crypto agility to protect yourself against the future threat of quantum computers now you're adding ai threads because all organizations are using AI many different ways.

We're exposing ourselves. We have AI in a company with 30, 40, 50,000 employees all using AI tools, having AI agents coming through your organization. It's a similar problem. So you have to create an inventory of what AI bots are coming to your infrastructure and who are they, what's the identity of those AI tools and are they allowed to access the information we're trying to give them access or not.

So that this new wave of AI is creating new vulnerabilities, and we need to protect that as well. It sounds like that even before, you know, we talk about where, how do quantum computers and AI, you know, help maybe catalyze or accelerate a bunch of these healthcare use cases, financial use cases, etc. You're even saying, wait, hold on. Even by the mere fact that we can now design and lay out a roadmap for the capability of achieving this million qubits, right?

Now we actually need to revisit the entire encryption infrastructure modernize it from the manual processes of identifying the PKIs etc And to upgrade to what you call crypto agility so that as quantum computers become a reality you're able to constantly evolve and adapt the cryptographic mechanisms. And so it's really interesting to hear, you know, when you're talking to governments like the Department of Defense and banks, etc. You know, how does it actually look like? I mean, because it's almost sounds like they have to pop open the hood and start, you know, looking through their entire infrastructure.

Well, they love it. They love the tool because they say this is not only helping me with a future threat, which is not that distant. If you understand store now, decrypt later, which means that nation states are stealing our information. They are storing it.

They can't reading now, but they are holding it until they have a quantum computer ready. And they will read it in three, four, five, six years when they have the quantum computer, which means our information today is in danger today, not tomorrow when the quantum computer is ready, but today. So the governments and major CISOs and companies that understand this, they say, I need a plan to protect myself against that. But then they use the tool and they say, hold on, Your tool is helping me not only with that, but it's also helping me understand my classical encryption today.

Because infrastructure is huge. And there are many instances of your infrastructure that are across everywhere and globally. If you've done a lot of M&As, you're acquiring infrastructure and have in cryptography that in many places that you don't know you have it. or you don't have cryptography, one in five breaches come because of a credential that was lost or not updated at the right time.

So that tool is helping you monitor and modernize today and preventing you from the future. Makes sense. What are you excited most about in the world of the post-quantum computing? That's a good question.

The simulation part is very interesting. So everybody's expecting the quantum computers to come and run those high-performance computing algorithms and problems that we can't tackle today. Pharma, it's a big part of it. Energy, that we talked about it at the beginning of this talk, is another one.

Most of what we do in the world is molecules. It's atoms talking to each other, interacting with each other. So when we think about high-performance computing, when we think about the AI revolution that we're in right now, most of what we're talking about and what people are focused on are Gen AI. Gen AI is great, but it works with words.

And words can only get you so far. And it's going to have a huge impact in programming for coding and in content creation. So marketing, customer support, et cetera. But it's not really impacting the real economy, the physical world that Genseng calls.

which is what impacts the GDP of the country. That's manufacturing. That's production. That's designing a car.

That's designing a new molecule, a new drug for a disease. That's chemistry. That's biology. That's mathematics.

That's physics. That's what we are using, which is large quantitative models, which are models of AI trained on equations that we know have been working for 100 years, not just on words, but also on numbers, quantitative. so that's what's really going to have an impact and that's the next big wave of AI that is coming right now because it will impact energy companies financial institutions when you talk about finance, banking that's numbers, that's mathematics, that's physics when you talk about a pharma company or a new chemical company that runs most of the combination of molecules that run our world in agriculture, in new materials refineries we create new open source data sets for catalysts, for example, for energy companies.

Catalysts are in the core business and the core elements of energy transition and hydrocarbon cycles. So that's what really runs our world and that's what's really coming right now. That's the AI applied to the physical world, not just the digital world. And what role long-term does Sandbox want to play in this world of post-quantum?

Does it really want to be this infrastructure layer of the encryption and the crypto agility? Does it have different visions? That's a good question. I mean, we are a cybersecurity platform and we started with the post-quantum approach.

We extended into all encryption. Now we are extending again to AI solutions. It's all about protecting our clients. So what the CISOs are needing instead of credential, access, cryptography, protecting your data, your infrastructure, and have a positive impact at scale.

That's what we're trying to help our partners with. And see, it must be pretty exciting, you know, being at the forefront of this for a few years now already, even well before Gen.ai even started, right? Yes, in fact, we started using Tensor Networks which is a technique that comes from quantum physics with transformers back in the day when we were at Google.

And that was one of the first proofs of concept that we had when the team was very small to say, wait, these techniques that come from academia and that have never been used in an application can really impact everything we're doing. So transformers come from a long time ago, like from 2017. That was a famous paper that we talked about. Not everybody has seen it through OpenAI but it has a lot of implications Yeah You know it funny how you saying long time and I thinking yeah well eight years seems like a very very long time these days Yeah, everything that's happened in the last three, five years, it's crazy, and more than it's happening, right?

Everything goes so fast. In the next two, three years, everything is going to change a lot. Yeah, and I think, you know, I'm looking at the industry or even from a VC perspective, You're seeing new venture capital firms forming just for quantum. You're seeing companies that are not only raising their C, Series A, but they're raising Series C, Series D in quantum.

You're seeing different layers of quantum from the hardware and what Jensen is talking about all the way to how do you write the new programming languages for it? And how do you write the relevant software to engage with it? So it really feels like we're, you know, we're like this, there's like this whole world that's about to evolve that. And it seems also that most of the world is not really talking about it yet.

Right. It seems like people are still very occupied with Gen AI and what we have. You don't hear about quantum still that often, at least not as often as I think we should be compared to where we are. Is that the right assessment?

I agree more and more often. I'll give you two examples. When we started five, six years ago, most cybersecurity teams had never heard of post-quantum cryptography. Now, every major company and government has a plan for it.

They might not already have the solution or acting on it, but they have a plan and they are working on it. That's a big, big difference. The other one is understanding that these two technologies, AI and quantum technologies, go together. It's not I need an AI team and I need a quantum team.

It's the new talents that we are hiring are computational chemists, computational biologists, computational physicists. These are the new generations of talent that understand that the digital world and the physical world go together. And they're going to look at the problem and then converge all the solutions, all the tools they have. And AI and quantum technologies are tools they're going to use to solve that problem.

They're not going to differentiate, oh, I'm just going to use AI, or I'm just going to use quantum. I'm going to use AI when I need it and quantum when I need it. I'm going to use them together. The other example is you see all the big companies, even NVIDIA you mentioned, or Microsoft.

They're calling the team, even Google, the Google team is calling it quantum plus AI. It used to be called quantum, now it's quantum AI. Microsoft is quantum and AI. NVIDIA is AI and quantum.

You see everybody's realizing that quantum is not as far as it thought. and that AI and quantum go together. That's very important. What is one thing that you think all of us should keep in mind as we enter this post-quantum world, either as entrepreneurs, those working on the frontier of new technologies, whether it's in healthcare or agriculture or finance?

What is something that we should be keeping in mind from today? That's a good question. I can think of learning and adapting it's nobody knew the best example is Transformers was created at Google Google was the king of search and of the technology now GPT is there all these LLMs and now AI for the physical world it's having an impact The most important part, I think, is what kind of impact are you having with the technology you're using? At Sandbox, it's all about positive impact at scale.

So we bring the scale mentality from Google X, the positive part of it, because we want to make good in the world. But it's about like, what are you using? Are words enough to have an impact? Words are really good, but maybe not enough.

We need to bring chemistry. We need to bring physics. We need to bring biology, which is what governs our world. And that's what we're going to impact.

We're going to solve the energy crisis that we are in. We have a huge energy demand. We're going to solve the climate change. We're going to solve new materials.

We need new materials, more resistant, lighter, more sustainable. That's only going to come if we use these technologies the right way. And that's going to be very, very positive and very impactful for the next generation. Fernando, thank you so much for spending this time with me.

I'll let you go back to doing amazing things and really appreciate it. It's been inspiring to watch the journey of Sandbox from the side and very interesting and eye-opening to hear your optimism and your perspective moving forward. And obviously, you're one of the few that are getting a front row seat to this transformation in the real day-to-day.