How to Win the Global AI Race

0:00You are now listening to the EU Start-ups Podcast. Hello and welcome to the special episode of the EU Start-ups Podcast. We're here at the A5X conference in Singapore by Mutation, the National University of Singapore. Throughout this next few days we're going to be chatting about innovation, funding, and the developing bridge between Southeast Asia and Europe. Not to mention, they sing in port defense tech hackathon, whose winner is going to show up at our very end of the series in a special episode. So I hope you're looking forward to hear us talking about the challenges, the innovation, and how we're all comparing when it comes to localized markets such as Asia, Southeast Asia, and Europe, and even the global challenges. Not to mention, the lessons we can learn from each other. Well, Orna, thank you very much for joining me on this very special and first episode

1:01of this series that we're going to be doing here for the EU Start-ups Podcast. It's a pleasure to have you on. Thank you very much for having me. Great. So before we get into the questions, which I know you have a tight schedule, so I do want to get started. I was just hoping you could give us a little bit of background as to who you are and what your experiences. So, as you say, then I'm Orna Berry. I have a PhD in computer science. I have two lives. One life is related to my professional career and the other life is related to my personal, my family and my personal affiliation with a lot of volunteering, which covers normally where neither the government nor a organization cover. And I find very important. So from a professional perspective, very early on, I moved from research to industry.

2:05I was an entrepreneur, I was a chief scientist of the Israeli government. Actually, it is to cover the industry research and development funded by the government. There's three major funding agencies for research and development, the academia defense and industry. I was in the sleep. Then I went back to work for 10 years with my investors. My company was sold to a German conglomerate to Siemens. And then I was hired to become a co-poid vice president and the general manager for the Israel Central Excellence for EMC later acquired by Adele.

3:10In this interim period, I also called cancer, which was considered to be terminal and an interesting work done with my doctors. I came to a resolution that I'm talking to almost 11 years later. And then worked for two years, almost three years with the research and development authorities in Israel to form national programs in quantum science and technology and artificial intelligence and data science. And then I joined the Google Cloud. So that's where I am today in the office of the CEO of Google Cloud. No, I even came across a Wikipedia page of you. So you have quite a lot of experience and it comes to all this. It's actually in common for the guests that we have to really have such a storied pedigree,

4:14especially because many of the conversations that I have are usually with people that are either getting started or have just found success. But you bring so much kind of experience and wisdom throughout all these experiences I've had that I'm hopefully going to give to a very interesting conversation. So with that in mind, I think the first natural question that comes to me when you're telling me about all this is how you've seen the sector, the innovation sector. And I know that's a wide topic, kind of progress throughout the years. And now we live in this AI age, right? And I kind of want to hear your thoughts on how you see it standing right now. It can be when we're talking at US or Israeli, but I think a better look would be at a global innovation landscape. So AI is pretty new. In the sense that it has several waves where it started, I think, with the Japanese fifth

5:18generation that the machine we're supposed to act on their own. And thinking about robots and about otherwise. But really in the last few years with the emergence of the infrastructure on one hand and the ability to create the algorithms that they exploited in order to come closer to human intelligence, that's I think where it stands today. And on the way, there are many, many issues because the more you count on machines, the more you want the machines to do what you intended, you want them to be secure. We have seen in my early days of the career networking evolve and networking critical components,

6:19nobody talks a lot about networking per se. That's the performance of the cloud, that's the reliability of interconnection. You go into a place and you don't have an internet connection, it gets you anxious. You need connectivity of all kinds. So you look at everything from the processing to the common use of applications of agents, of what not. And this is an ample development that took quite a bit of time from the time that we started connecting devices to the time that we started using them as a transparent infrastructure. And how do you see the challenge, and you just mentioned here, the infrastructure, the challenge of infrastructure when it comes to things like data centers. That's a big challenge for us in Europe right now.

7:21And the European Union is putting all this work into establishing more in data centers. Of course, there's certain public backlash even to it. But do you find that the infrastructure question is a big bottleneck for AI? Infrastructure is always bottleneck. You always, when you want to know, when you know what is the next generation that you want to establish, you build it first in terms of infrastructure because you cannot do without I think a great example of building a fairly clean infrastructure is Finland. And the Hamina data centers that the first generation was built with significant power based on wind. And the second generation is also built with the ability to distribute the heat to be used

8:23in the surrounding. So there are discussions about fusion and about atomic energy and variety discussions of where you get the sources of the energy and then how you translate the energy into the data centers. And many places are using solar energy are using variety. Solar, of course, is only for the daytime and when sun is out. And from your work at Google Cloud, is that something that's also kind of impacting the business? I'm sure that what I do is not related to the infrastructure. I assume that I have the infrastructure that they need. So I can deal with education, I can deal with health or research, I can deal with topics

9:27on top without a... But the only one that is maybe related to dealing or layers is security. That's it. But the assumption has been correct, right? You've assumed that the infrastructure is in place and that there's availability and it is. Yeah, but coming from the point where we shifted between mainframes to network connected the computers back in the early 80s, living through that, I don't assume everything. And yes, I get distracted when I don't have an internet connection. It works. Okay. And I want to pull back a little bit because you talked about Finland as a good example of well infrastructure application. So that makes me kind of curious about your thoughts, because you've worked on national programs as well when it comes to these initiatives. What makes or breaks a nation when it comes to their attitude towards AI, cloud infrastructure?

10:39I think that we know that the regulatory discussions, because of the fact that you put regulations, we talked about it, regulation and innovation do not go very well together. And the other end when you scale and it is a massive intuition into all layers of life, then you must be protective such that you use it with at least responsibility and explainability. That's a big question, isn't it? Especially for us Europeans. I can't tell you how many founders I talk to that talk about regulation as being the main stoppage for them and the big reason why they might go to Silicon Valley over staying in Europe. So first of all, people can go wherever they want.

11:42They can get visa, they can get investments, they can any talented person has high mobility. And one of the issues that the country needs to consider is the fact that talented people have high mobility in this world. So I think that the topic of guardrails is not necessarily only in Europe. You hear about China, I haven't been to China in a while, but I'm sure that there are topics in China that manufacturers have to open up to the government. And when you buy products outside of China, people may want to have more privacy and less

12:43exposure of the loan. So I think that wherever entrepreneurs work and invent, first of all, the stage that is deep science is separated from the regulation. And so you come to develop a drug, it needs to go through their poor world process because you cannot be sure that it's safe, that it's efficient, that you know exactly who to give the drug, who not to give the drug. You can kill people if you give them the wrong drug. So I think the same holds for software, because software manages our infrastructure. So you want to build software that manages the infrastructure in a way that you can maintain

13:46it and control it. So I think that it doesn't need to limit the imagination. It requires certain protocols, paramal get segment that you need to put in in order to use it for good purposes. And that's what I think the issue is. The discussion of it's very difficult to understand the head of time. But a medication does. Viox was invented for one thing, used for another thing, and caused the damage in the original discipline. And eventually people stopped using it, even if it was very efficient for orthopedic pains.

14:46So you need to know where you live, you make inventions, but you want to use them responsibly. Using responsibly is also understanding what they do. And you've mentioned here the mobility of talent. And I think that is another side of the argument here because yes, founders, entrepreneurs are mobile and we live in such a globalized world. It's a lot easier, right? And then a European can come to Singapore, Singaporean can go to the US. This is all made easier. And it's recommended that everybody does that because if you want to go to a market and you don't even have any connections, they are not to talk about the fact that knowing the pricing and the need and the variety of other issues, it's much harder to establish the ground level that if you go to a market that you have been there before.

15:49But then what about the source of this supply chain? Because there has to be an interest nationally speaking to make sure the residents and citizens of these countries are the ones promoting their own local economies. There has to be an argument there about kind of exporting talent and importing talent and how that can lead to problems. So I think that many times and for many entrepreneurs, their roots are very important. I spent 10 years in the States and I came back to Israel. And when I came back to Israel, I didn't only come as a professional back to Israel. I came with a social responsibility and I knew that I'm going to make sure that what I'm doing is making a contribution to the state. And indeed I spent time in public sector, I spent time volunteering and in addition to

16:54having a professional career in order to put food on the table and to keep my brain working. And I hear that a lot actually and there's a sense of national pride that comes to successful entrepreneurs. Even if they've gone abroad and even if they've gone to Silicon Valley or wherever it is in Asia, many of them will return to their home countries in an effort to promote their own local economies and all that. So from what you're saying, then I'm understanding that that is an important aspect of this talent. Supply chain is maybe the exportation of talent but then returning to their countries to improve. Yes? Yeah. And sometimes you go for three years and sometimes you go for ten years and sometimes you go for 25 years and you come back with a massive experience and they potentially also go to the economical night to make others successful as well because that's a commitment and

17:59that's fun to see that they're going. Okay. And to shift the conversation here a little bit because I do want to tap a little bit more into all your background when it comes to AI, when it comes to deep tech, all that development. I'm curious to hear your thoughts on if there is something right now that worries you or makes you question about the way AI is developing and the speed at which it's developing. Because we've talked about regulation and is that an ecosystem? No, so you get into another ecosystem that is not only the technology. The technology in the hands of the users is the question. So you want guardrails in place when you count the users are not sufficiently knowledgeable or responsible to use the technology responsibly.

19:05So that is where you care a lot about guardrails and part of guardrails regulations. And regulations are part of a government but also part of verticals like the financial sector. So I personally think that AI is phenomenal. I think I look a lot at education and for the first time this is a mechanism that allows to build something very, very complex because when you use AI to just accelerate a well-defined problem, it's one thing. When you use AI in order to be adaptive to different personas or different personalities

20:08or different capabilities, it is an explosion in the positive sense that you're actually allowing suddenly people with different capabilities to have pretty similar opportunities because of the fact that the AI is not only converging, unsolving and summarizing something. It's also opening up and has the multiple channels to actually create the dialogue of what is best for the opportunity or for the student in this case such that you can match and make more people much better professionally, much happier with the ham on moment of their own

21:09that is different than others. Okay, well I think that's all we have time for today but I want to thank you very much Hornett. It's been a pleasure and a real insight to talk to you. Thank you. Thank you.