Loading summary
Transcript56 lines
- [00:00]
Yonatan Adiri
Foreign.
- [00:04]
Michal Ivram
You are listening to an art media podcast. So, Yonatan, what's your number this week?
- [00:11]
Yonatan Adiri
It's in percentages. 99.9%. That's one in a thousand Iranian drones launched from Iran. Only one of them made it to Israeli borders and luckily not causing any injuries or casualties. And we don't know if that was the work of Jewish space lasers, but I guess we'll find out in the next couple of months.
- [00:32]
Michal Ivram
Well, mine is 66.70. That's 66.70 and that's the price of crude oil per barrel. This is similar to what it was June 12, the day before Israel attacked Iran. And we've paid a lot of attention to oil prices. There's been a lot of speculation. In the end, this is where it nets out. So both of our numbers are very war related. Not surprising. But who wins here?
- [00:59]
Yonatan Adiri
Well, I'm back in Israel. Had the pleasure of having a few of the latest sirens here in the shelter. To me, this 0.01% making it here is a very kind of, you know, close to heart number.
- [01:17]
Michal Ivram
Welcome to what's yous Number from Ark Media, where we focus on Israel's economy through a global lens. I'm Michal Ivram Yonatan. Adiri yonatan, it is 8:30am here in.
- [01:29]
Yonatan Adiri
Palo Alto, 6:30pm in ceasefire Israel.
- [01:33]
Michal Ivram
I can imagine how good it feels to be home and to say that you're home. So welcome back.
- [01:39]
Yonatan Adiri
It is, it's great to be home. All credit to my incredible wife and four kids for, you know, managing through. I was here only for two out of the 10 days. The last ones, pretty significant hit here, 300 yards from our home. Pretty scary outcomes. I think people were very much unified knowing that we're suffering this, we're in the shelters, but it's for a good cause. That was pretty much in consensus. That's very, very intimately felt. And you know, much like I was sitting at the shelter when I was, I think 9 years old when Saddam Hussein fired his 45 missiles, our kids, you know, are dealing with a similar dynamic. It's not nice, it's not fun, but it gives them an understanding that this all for a purpose.
- [02:23]
Michal Ivram
Well, and speaking of that resilience, the war obviously is still ceasefire or not very much in the backdrop, but today we're going to focus on the AI arms race, actually another kind of battle and what Israel's role is in this. We're going to talk to Yoav Shoham, who is an AI expert, longtime professor at Stanford University's computer science department. He is now co founder and co CEO of AI21 Labs. And this is one of Israel's most prominent AI companies today.
- [02:58]
Yonatan Adiri
Yoav is also, it's an incredible privilege to hear from him. He has a perspective, he's built quite a number of companies, been to Stanford as a professor in the golden days of the 90s, inspired some of the most famous digital entrepreneurs of Silicon Valley. It's a great opportunity to speak to him, to understand how he thinks about AI in general and stick around. It's going to be interesting.
- [03:17]
Michal Ivram
Absolutely. Well first as usual we're going to take a look at some of the week's pressing news AKA our big shorts, including some key numbers that should be on your radar like the Windex. Stay with us on to our big shorts. We're going to kick things off with this week's what's your number index or Windex. This is our very own index of Israeli based and or Israeli founded companies. Yonatan, how is this week's Windex looking?
- [03:48]
Yonatan Adiri
So I should say last week, that is June 16th through the 22nd, both NASDAQ and S and P were on the rise in the green. Windex was in the green as well, but I would say underperforming by half a percent or so. So growing by 1.07 whereas S&P and Nasdaq grew by 1.67 and 1.6 respectively. This was a good week for the markets. Windex included an interesting move in the green this week on the Windex. It's a combination of two autonomous companies from Israel. Always great to see that we're still critical in the infrastructure of that growing industry. Mobileye, the pioneer of autonomous car sens until recently the biggest acquisition out of Israel by Intel and by the way.
- [04:34]
Michal Ivram
Co founder of Mobileye, also co founder of AI21. So nice connection there.
- [04:39]
Yonatan Adiri
Yep. We're going to hear some interesting feedback from Yoav Sharamon Amnon Shashua, that founder, really a legendary Israeli entrepreneur. Mobileye has been growing at 16.3% last week grew 16.3%. Innovis, a smaller company and upstart IPO'd early on its journey on its path. That makes a solid state lidar there's a big competition on how do cars actually actually see both companies rose with Innoviz growing 45% 45.74 on the news that Volkswagen's robo taxi going to be launched next year incorporates both companies technologies. This is a big win for both.
- [05:17]
Michal Ivram
So let's See, not surprising to see these numbers, honestly. I mean there's so much going on on the Robotaxi side. Waymo announcing that they're going to be launching more markets. I think they just launched Atlanta, they're about to in D.C. and other locations. They're all over the place in Palo Alto by the. And then of course Elon Musk, you know, with his Robotaxi launch. So a lot going on in that space. I would not be surprised if we keep seeing these kind of positive numbers.
- [05:42]
Yonatan Adiri
I agree. And I think we'll kind of look at the subset on the Windex in a few weeks and kind of try to see companies related to that feel and how they're doing. Very much like we did the defense tech space. On the negative on the red solaredge, which we've been following, kind of going back more or less to its original figures. Pretty hard time in the solar energy space to me. Another indicator to watch is the primary victory today. What seems to be a primary victory for a Marxist BDS supporter, New York mayor candidate that is Zoran Mandani, after doubling down just yesterday on his will to arrest Prime Minister Netanyahu upon his next visit to New York City. I think a win for Mohani casts a really dark cloud on Israelis doing business in New York. I mean the whole rationale behind the Windex is to reflect this tie between Israeli founders and growth technologies in New York. This would have a major negative impact on Israeli related economic activity in New York. This is a major one.
- [06:37]
Michal Ivram
Yeah. I mean maybe the bright spot here is everybody speculating that real estate prices in Florida are going to experience a boom now. But very good point to bring up and I think worth a deeper dive in a future episode on more broadly speaking on BDS and sort of what traction we're seeing and what are some actual real potential impacts.
- [06:59]
Yonatan Adiri
Yeah, overall I think, you know, we saw Michael Eisenberg this morning, you know, one of the titans of Israeli tech industry publishing his thoughts and perspective on the Mamdani candidacy. And it's the Florida piece, but it's also the Israel and the Tel Aviv piece.
- [07:12]
Michal Ivram
Well, let's move on to some other things we're seeing happening both in Israel and elsewhere. For starters, seven year old digital payment company Melio was acquired by SMB giant Zero. This is following a reported negotiation for acquisition by Bill.com, an American company for 1.95 billion cash and share deal which fell apart in 2023. Yonatan, what do you make of this?
- [07:39]
Yonatan Adiri
I think first, you know we're not even halfway through the year. Israel is at 42, $43 billion of acquisitions, IPOs overall, you know, kind of realized economic value of its tech industry. That's almost 8% of the GDP. These are unprecedented numbers. Milio's acquisition was negotiated from the shelter, from the mamad. From the mamad. The founders kind of. There's been a nice conversation online talking about how the last couple of weeks they had to navigate and when I was in that seat, there's a lot of T's to cross I's to dot when it comes to these acquisition agreements. When we acquired an American competitor, it was. The last few days were always a nightmare. I can't even imagine what it was like for the founders with sirens and with the kids and with hits in the general Tel Aviv area to negotiate and finalize a deal with a New Zealand based acquirer. So kudos to them for the grit and the persistence. A seven year old company reaching this type of valuation is important. I think the third part is how Israeli fintech is not just on the top layer, right? The application layer. This is kind of like the plumbing. So Milio is in that space. I would value the total sort of enterprise value at this point outside of Meliot, I would say 12 to 15 billion dollars of market cap in that space. That's great to see Israeli companies, as opposed to where we are in AI, this is similar to where we are in cyber, right? We're at the very bottom of the stack, which is really good because wherever the industry goes, these companies have the grit and the wherewithal to sustain this type of challenge. I think we're in a great position there on the app stack, but also at the sewage, at the plumbing. And big kudos to the team at Melio for negotiating and finalizing an acquisition agreement. Really under, under fire.
- [09:30]
Michal Ivram
All right, let's get to our long play. This is our deep dive into one significant issue affecting the Israeli economy. And as we said earlier, today's long play is a conversation with Yoav Shoham, longtime entrepreneur in the AI space and professor at Stanford University. Yoav has started and sold several AI companies well before ChatGPT and this whole current explosion of Gen AI. And his newish company, AI21 is one of the most prominent AI companies in Israel today. So we're going to hear more about his company. We're also going to talk to Yoav about how he thinks Israel is positioned in the AI arms race. Yoav, welcome to what's yous Number? First question for you is just. What's your number? Did you bring a number to share with us?
- [10:18]
Yoav Shoham
It's gotta be four.
- [10:20]
Michal Ivram
Four, okay, meaning.
- [10:22]
Yoav Shoham
Well, it's, you know, more than three and less than five. But start with, I have four kids, the most important thing in my life. And I was associated with four higher education institutions, starting with the Technion in Israel, ending with Stanford in California. And I've had four companies, so it seems like a destiny.
- [10:45]
Michal Ivram
So, Yoav, you recently came back from Paris and I think you were stuck abroad like Yonatan was as well, and many other Israelis. But I saw that you wrote about this dinner hosted by Emmanuel Macron, where there was a discussion on AI strategy, and you wrote that his mobilizing of the ecosystem to support French AI should be studied by every national leader. So as we're talking about what Israel's strategy is currently, which you play a role in, and what it could or should be. Tell us a little bit about just what you meant with those words. What can be gleaned from what France is doing?
- [11:22]
Yoav Shoham
France has emerged as probably the preeminent AI power in Europe, eclipsing the uk, which is also substantial. And I think Macron really was pivotal in this. What he did was a combination of things, of drawing attention, coercing the very substantial French industry to support the French AI, both in terms of booking orders and investing. And so Mistral is a poster child. But there are other companies who are kind of drawing a lot of attention in France, usually in Paris and the environment. There are a few such cloud companies there also. And it would not have happened without his. And he's organizing high profile conferences. This meeting was a very kind of intimate meeting convened by him. And Maurice Levy from Publicis was, I think, the driving force behind it. He's also the driving force behind VivaTech, which has become Europe's largest tech conference. I'm not aware of any other leader who has done as much for their own industry AI. There's also a general European ethos behind this effort. Not just France, but certainly it's France and then Europe.
- [12:37]
Michal Ivram
What do you think about the term AI arms race overall, just to begin with?
- [12:42]
Yoav Shoham
It's perhaps a little militaristic, but I think it captures reality. Everybody's rushing to build capacity. Capacity can be in raw compute or talent or whatever, anywhere up the stack. I think some of it is fomo, to be honest. But nonetheless, I think every country owes it to itself to develop that muscle. If I take Israel as an example, by the way, I actually wear another hat. I chair this scientific committee of the Israeli AI project, national AI project. So I have some visibility into it. Israel as an example, will never compete in the quantity of compute that US has or China has, or for that matter Saudi or the uae. But it nonetheless needs to have its own AI infrastructure. And having means multiple things can mean actually having it physically on within its geographical borders, or at least having access to control over that capacity elsewhere.
- [13:42]
Yonatan Adiri
Quick question on that, Yoav. Do you think it is plausible that at some point we will see cloud infrastructure being closed for certain countries because of kind of a split, be it China, US and allies? And in that context, do you think at some point in the foreseeable future we're going to somehow ameliorate the energy kind of better? AI means exponentially more energy or sort of how do we fix that as part of this arms race?
- [14:07]
Yoav Shoham
So first of all, I think people don't realize that AI doesn't start and end with just very expensive high end energy compute of LLMs. In fact, we're sort of reaching a plateau of just throwing more raw power at the models. And so I don't know that energy which right now consumes the system, consuming a ridiculous amount of energy that'll continue, but I don't think it's going to grow unboundedly. It seems to me that currently the energy supply in different parts of the world is correlated with the amount of compute that reasonable for those areas. So for example, UAE has a ton of compute and as much energy as you'd like. Europe, I think is aided by it having invested in nuclear. That's going to help us, certainly in France. Israel also seems to have enough energy to support the amount of compute it will build in the near future. Right now we're talking about not many thousands of GPUs, and even if we grow to tens of thousands of GPUs, I think that's within the capacity of the grid supply.
- [15:19]
Michal Ivram
Give us your take on what is Israel's AI strategy currently and let's talk about where it could be going.
- [15:25]
Yoav Shoham
Israel unfortunately is distracted by all kind of stuff, internal, external, and you know this as well as I do. Israel being Israel, it had two AI plans. There was one and then the second one was adopted. This was headed by Dr. Orna Berry, whom you may know. It was adopted as a government decision. It was budgeted not at the level they require. I think they requested 5 billion shekels. The first tranche was half a billion and then the second tranche of another half billion. So 1 billion shekel is not nothing, but it's very little. It happens under the auspices of an interesting forum called Telan, which is a. It's hosted in the Ministry of Education, but Ministry of Finance and Ministry of Defense and the Innovation Authority. Everybody's involved there and he's doing good stuff. But more needs to happen. And there have been various noises from the Prime Minister's office about efforts to catalyze a more substantial effort in this area. And we'll see what happens.
- [16:34]
Michal Ivram
So if you've got obviously US and China, the big kind of superpowers here in the AI space. So let's talk a little bit more about the role that some of the smaller countries can play here. I mean, if you extrapolate. Yes, yes, there's a focus elsewhere right now, but in the near future, what role do you think Israel can play and should play? What are the competitive advantages that the country has when it comes to AI?
- [16:59]
Yoav Shoham
So while it's true that if we are trying to play the popular chatbot game, that's a very expensive game, by the way, there's no mystery here. What needs to happen is just a lot of work and a lot of money. Money and a lot of compute. That's a game that few will play. But nonetheless, there will be multiple core technology provider from other places for a number of reasons. One is independence, the other is needs that aren't general purpose chatbot needs. So if you go to France, you have Mistral and a couple of others go to Canada, you have Cohere and Israel you have AI21 as a core technology provider initially building LLMs. And our LLMs are somewhat underbranded and working to fix that. But we're actually doing very well because of the unique architecture we have. I think the biggest game now is moving beyond LLMs to AI systems because, you know, in the enterprise there's a sharp drop from pilot to deployment. And there are many studies that show it. One AWS study shows sort of 6% of the Project CO2 deployment. And the main reason is there are two reasons. They're kind of vanilla reasons. One is new technology, need to understand the use case, privacy, security and that, you know, compliance. But the two kind of things that loom large, one is accuracy, reliability, and like I said, if you're garbage, 5% of the time, you're dead in the water. And the other is cost. The LLMs are costly to deploy, they're hard to predict and they're hard to predict the cost of. And they can be expensive CIOs and now CIOs are looking, looking very carefully about this and world is moving to AI systems. We just announced Maestro, which is really the first such sort of orchestration system that's industrial grade and that's. And I haven't forgotten your question, that's no longer an issue of just throwing money at things. It's much more an issue of brain versus brawn. This is very much a blue ocean and with no particular advantage to people, to companies or countries that have a lot of money and compute.
- [19:16]
Yonatan Adiri
Maybe a replay or rerun of previous Israeli dynamics where B2C was too far and so we ended up with B2B and enterprises. Are you seeing the same? Is it also the type of talent that we have that drives us in that direction? Or it's more of a strategic choice that AI21 took, but also maybe other companies where you see the blue ocean.
- [19:35]
Yoav Shoham
I think it is a strategic choice we made because, you know, we actually had a very successful consumer product called wordtune, a writing application and it's been just amazingly successful. But we didn't see it taking us to be where we want to be. From the very start. We wanted to be a B2B company, but that just wasn't a market. And now there is a market. Israel has become, as you Jonathan know, and in fact Michal too has become actually quite good at consumer application. There are a bunch of, you know, application builders who are using AI. So I don't think it's indicative of the Israeli market in general.
- [20:14]
Yonatan Adiri
And last episode we discussed the acquisition of base 44 by Wix. Not so much in the context of the degree of AI that was deployed, but more from the context of this North Star of potentially like few people, billion dollar companies. Right, Like a small group of people with a killer app. What's your thought process on that? Over the next seven years are we going to see sort of smaller teams relying on unique architecture to your point, Maestro orchestration and then generating outsized value or at some point at scale, when you deploy a B2B in the billions, you're going to need the classic infrastructure of human beings still to build that company.
- [20:51]
Yoav Shoham
I think realistically you're going to need boots on the ground. We started with small number of people, but to scale and accelerate you need people and certainly for go to market and customer support and all of that. But even for product development, at some point you need. There's probably economy to be had. Certain aspects can be covered or augmented with technology. And so I think people will look carefully before they hire Big armies. But the one person billion dollar company I think is a catchy slogan, but probably not realistic.
- [21:31]
Michal Ivram
Speaking of having people of boots on the ground, I think you've always had co founders in your companies. At least that's my recollection. I know you've talked about your co founders. I've heard you say, I think Amnon Shishuay was. He's got twice your brain. How do you characterize your co founder selection here?
- [21:50]
Yoav Shoham
No. So actually Amnon is maybe an exception here, but my rule of thumb is somebody who's half my age and twice my brains.
- [21:57]
Michal Ivram
That's what it was.
- [21:58]
Yoav Shoham
And I also said that it gets easier and easier to find them. So I have two co founders, Ori Goshin, who very much fits the bill, and Amnon, who's just twice my brains, but not half my.
- [22:10]
Michal Ivram
You have last question for you. I know you've spent a lot of time over the years thinking about some of the deeper questions that AI brings up for humanity. And there's this recent MIT study that's been making, at least in the media, in the headlines, been making a lot of waves that seems to suggest that the use of ChatGPT among students is eroding critical thinking skills. What do you think? Where do we net out here? Does AI overall help the way we think or does it erode it?
- [22:43]
Yoav Shoham
There's no question that technology in general changes us. And if I look at my kids, they're, you know, I'm sure it's true for you. Look at the kids. They're growing up in an environment that's very different from ours. And it's actually hard for us to understand how they're thinking and, and even before AI. So we bemoan their addiction to screens, their short attention span, the fact that they don't do long read, and it's all true. Does that mean they're worse thinkers? I'm not sure. They're different. Does that mean that we're not going to have the scientific say, breakthroughs that call for somebody just banging their head against a problem for days on end and having the stamina and the perseverance? I don't know the answer to that. People make the analogy to gps. You know, we don't navigate as well as we did before. It's true. But I don't think I'm a worse spatial orienter because of that. I just have a tool and it changes how. So I think it's early to say what ChatGPT does to us. It's early to say that I'm not cavalier about it because it does do so much, these AI systems that it can make us lazy. And I think we'll need to develop skills to control for that. And maybe at the end of the day we'll now it'll be very hard to function productively without this very smart assistant. But like with human assistants, we'll become extremely good and managing the assistants, coaching them, checking them. But it's a good question. And I still don't know what MIT stands for.
- [24:36]
Michal Ivram
Well, we miss you here in Palo Alto, where Stanford University is located, but excited for what's to come with AI21 and appreciate you speaking with us.
- [24:47]
Yonatan Adiri
Thank you.
- [24:48]
Yoav Shoham
Thanks for having me. It's good to see you both again.
- [24:55]
Michal Ivram
Well, we began our show with the numbers of the week, and we're going to end with the words of the week as usual. China can now continue to purchase oil from Iran. Hopefully they will be purchasing plenty from the US Also. And that, as I think most of us know, is President Trump's recent social media post. I, by the way, I was gonna go with the like, these two countries don't know what the fuck they're doing, but I decided to go for this more tame one. Yeah. Are you surprised by this, Yonatan?
- [25:24]
Yonatan Adiri
Yeah, I actually think it reveals something that sort of was on the table throughout the war that got me thinking. And you put it forward with your number at the beginning, right. The price of oil. We saw Russia and China standing down. We didn't see pressure on Israel. We didn't see pressure on the US Neither from China nor from Russia. There might have been a deal. Right. So I think this is indicative of sort of maybe President Trump saying, saying you've been good. Right. And so now I can kind of lift the veil on what everybody knew was going on, which China was procuring oil from Iran. There's nothing new there. But in lifting that veil, President Trump actually sort of did a favor to the Chinese in a way, kind of legitimizing it, although it's against the sanction regime that the US and the Security Council have imposed on Iran. So I think overall, what I take from that is, you know, kind of in hindsight, 12 days of international support, or at least no stopping of what was a crucial play for Israel's future.
- [26:22]
Michal Ivram
All right, that is it for today's show. Thank you for tuning in to Ark Media's what's yous Number. We hope you found it interesting. And if you did, please be sure to, like, subscribe rate review. You know, the drill, but most importantly, please share it with others who you think will find it interesting. If you want to make suggestions or share your feedback feedback, please reach out to us@whatsyournumberkmedia.org.
- [26:53]
Yonatan Adiri
What'S yous Number is produced by Ilan Benatar and edited by Michalev Ram. Sound and video editing is by Martin Huergo. Our theme music is by Midnight Generation. I'm Yolata Nadiri.
- [27:03]
Michal Ivram
And I'm Michal Avram. See you back here next week.
- [27:06]
Yonatan Adiri
See you next week.
- [27:07]
Yoav Shoham
Week.
- [27:11]
Michal Ivram
This podcast offers general business and economic information and is not a comprehensive summary for investment decisions. It does not recommend or solicit any investment strategy or security.