Minor Powers· Staffel 1· Folge 1

There Is No Light Web

Dienstag, 27. Oktober 2026 · 40 min

MP3

Shownotes

A flaw sat for years in the code beneath most of the world's websites until an AI model went looking for it (See: PHP2Shell). Dror Poleg, an economic historian who writes about technology and work, doesn't think patching ends the story: the attack surface only widens, and some things may have to come off the internet altogether. Then to Israel, voting on October 27, where Benjamin Netanyahu campaigns against dangers his own government deepened and has floated crypto as a hedge against sanctions. Last, two fathers of small children compare notes on raising them alongside the machines. The main thing to do about AI, it turns out, is other things.
Cue the newest show in the Perch Podcasts sphere - Minor Powers. Jacob Shapiro and Dror Poleg are embarking on a weekly journey to discuss tech, war, AI, and how to be a dad in the 21st century.
--
() - Intro/Welcome
() - Black October and the PHP flaw
() - The end of the internet?
() - Banking, privacy and encryption
() - Learning to live with the risk
() - Israel votes
() - Netanyahu's bench and the FlyDubai plot
() - Crypto against sanctions
() - A state like any other
() - Young voters and the old guard
() - Will AI kill the creative class?
() - Raising kids around AI
--

--
Jacob Shapiro Site: jacobshapiro.com
Jacob Shapiro LinkedIn: linkedin.com/in/jacob-l-s-a9337416
Jacob Twitter: x.com/JacobShap
Jacob Shapiro Substack: jashap.substack.com/subscribe
--
Jacob Shapiro is a speaker, consultant, author, and researcher covering global politics and affairs, economics, markets, technology, history, and culture. He speaks to audiences of all sizes around the world, helps global multinationals make strategic decisions about political risks and opportunities, and works directly with investors to grow and protect their assets in today’s volatile global environment. His insights help audiences across industries like finance, agriculture, and energy make sense of the world.
--
Dror Poleg's Site: drorpoleg.com
Dror Poleg's Twitter: x.com/drorpoleg
Dror Poleg's LinkedIn: linkedin.com/in/drorpoleg
Dror Poleg's Substack: newsletter.drorpoleg.com
--
Dror Poleg is an economic historian and former tech and private-equity executive. He researches technology’s impact on work, markets, and the built world. Dror’s writing has appeared in The New York Times and The Atlantic, and his insights have been featured in The Wall Street Journal, Bloomberg, and The Times of London. He briefs leadership teams at companies like JPMorgan, UBS, and HSBC.
--
Minor Powers is produced and edited by Audiographies LLC. More information at audiographies.com

Transkriptvom Podcast bereitgestellt · Zeitmarken springen hin

Jacob Shapiro: 00:03 Hello, listeners. Welcome to another episode of the Jacob Shapiro Podcast. This is gonna be an episode with Dror Poleg. He's been a guest on the podcast in the past, and we found our conversation so good and got some feedback that we decided to smush them together, and have a regular feed for them. So we're gonna cross-post them on my channel here for the first couple of episodes. But if you like the dynamic of this conversation it's gonna be called The Minor Powers Podcast with me and Dror. You can go sign up for that if you wanna make sure you have access to it. Like I said, we'll post a few here so you don't have to do it right this second, but you'll wanna do it going forward. The idea here is, you know, Dror is in the technology space, I'm in the geopolitical space, and we're also both dads of young kids. So we wanted something that was short, snippy really thirty, forty minutes max on a couple of set topics going through at the beginning of the week, and then also applying that in a real world sort of way specifically with the types of advice that we are giving as parents or how we're thinking about integrating this as parents. So a little bit different than the normal content that I have. It's less long form, a little more organized, and Dror is a really great guest to bring on. So if you're interested in that you might wanna sign up for that feed. We'll also, again, post them here for a little while going forward. Hope you enjoy the episode. It's a crazy couple of weeks of travel here, so we will do our best to keep our cadence up. But I've been overflowing with content the last couple of weeks in anticipation of this travel that was coming. So if things are a little quieter here over the next week or two, that's why. I think that's all. Take care of the people that you love. Cheers, and see you out. All right, listeners, we're doing something-- I guess we're doing something new. We're doing something different. Dror and I have had conversations before, but here we're gonna try and pilot Minor Powers, a new podcast between the two of us to bring, … Well, we're gonna try to bring short form content to discussions that really I know both of us wanna speak about for hours. So Dror, I'm excited to do this with you, man. Thanks for making the time.

Dror Poleg: 01:55 Excited to be here, Jacob.

Jacob Shapiro: 01:57 All right, we're-- first episode. We're recording Monday, October 26th. We'll probably put this pilot out tomorrow. I was-- when I was at the gym this morning, Dror, I was thinking of you because you know, I sorta knew in the background this PHP cybersecurity issue had happened- Mm-hmm but I was listening to the Big Technology podcast with fellow Cornell alum, Alex Kantrowitz. And that was the first time I'd heard that it, it's actually Black October, that it's not just the PHP thing. Yeah. It's a whole bunch of other things that I hadn't heard of. I mean, I know, … You'll probably go through them. But it sounds like, you know, there was a major vulnerability on some of the coding that goes by-- I don't-- I can't even do this with all the language that goes into the websites. Yeah, so- I just know that it's a big deal. How about you lead us the way? And why-- Is this important? Is this not important? How should we think about it?

Dror Poleg: 02:42 Yeah, I think, I mean, l-l-let's start a few months back. You know, so earlier this year, there was the Claude Mythos from An-Anthropic, which was you know, touted as a national security threat by its own creators that basically decided to, you know, alert the government, basically beg the government to stop them and everyone else who's competing with them. At which point the government, which was already fighting with them about other things that they weren't willing to do for the government, kind of called everyone over and kinda set up what became a pretty useful program with both of OpenAI and Anthropic, where they give early access to a select group of companies and obviously the government itself to make sure that whatever exploits can be revealed will be revealed to them, you know, a few weeks or a few months before they reach anyone else, and definitely before they reach China. And they also barred a bunch of traditional US allies from these programs. So, you know- Mm-hmm … the UK or Israel don't have access to those programs. And that was about, you know, these kind of cutting-edge models that can, you know, hack things and do amazing things. What we started to see more recently, so you know, from June onwards, we saw some of the kind of models in training at both OpenAI and Anthropic and la-later elsewhere as well kind of- hacking their way out of the systems that are supposed to contain them or restrict them from accessing the internet. And more recently, we saw that even the simpler models that are now available to consumers, so not part of the program that is restricted by government, can already expose vulnerabilities in some of the kind of most commonly used, technologies or libraries or stacks that the internet is built on. So, you know, PHP- Mm-hmm … which you mentioned, is a programming language. It's the foundation of, you know, all the WordPress blogs, for example, which power, I don't know, 50% or 70% of all blogs on Earth Now, it's not like it discovered some amazing hacks.

Dror Poleg: 04:47 It like, it exposed a lot of things that were already there to begin with, but nobody noticed, and now if you patch them quickly, then it's good, but if you're not aware of them, then you're in trouble. And the same thing later happened with Ruby on Rails and a few other different types of languages and environments. And yeah, we can expect this to continue to roll on and to get worse and worse probably in the coming months as capabilities get stronger and as bad people put existing capabilities to even more exciting uses.

Jacob Shapiro: 05:15 Yeah, I… Well, and first of all, I mean, apparently the way they discovered this vulnerability, it's not that the AI was out there thinking, "I want to destroy the human race." Yeah. It was part of… The vulnerability was, with PHP in particular, was discovered in August as part of an OpenAI Patch the Planet project. Mm-hmm. And apparently an Is- an Israeli American cloud security company called Wiz was part of the discovery. But so it's been out there for a while, and then it got leaked earlier. And for me, it's this kind of push and pull, and I feel this way about AI in general- Yeah, like WordPress itself, by the way, patched it for the stuff that it hosts, but there's a lot of people that host open source versions of WordPress blogs that don't get-

Dror Poleg: 05:42 Right patched automatically when WordPress runs their own thing. Yeah.

Jacob Shapiro: 05:54 Well, but I'm really struggling with what the conclusion is from all this. Is the conclusion, like Ben Thompson made, this is the end of the internet, or we're not gonna experience the internet the way that we did before? My children- Mm-hmm will never know what it's like to go on the internet the way that I did. Or is it that, wow, AI has discovered all these vulnerabilities that we didn't even know were there. Now we can patch them, and we can make things even more secure. Are we actually helping us be more safe and secure on the internet with some of these discoveries?

Dror Poleg: 06:23 Yeah. I think I'm closer to the end of the internet in a sense. I mean, obviously the internet that Ben wrote about, you know, the nice days of the '90s, that, that internet already ended, so for better or worse. But I think the issue here is there's no, there's no end of history, I think, for cyber defense. It's an ongoing thing that just, you know, it's an adaptive environment. Whatever we patch, tho- there's always gonna be another thing that will have to be patched and a new vulnerability discovered and new ways of hacking and causing damage, even within the internet as it is today. So I mean, we saw also in terms of models being trained within OpenAI and Anthropic, they're finding new ways to hack the system that are supposed to constrain them. You know, they put them in two separate virtual machines Where there's no communication between them, and they find ways to communicate with each other through the hardware somehow. So unless you physically separate them. And even that, you know, some OpenAI researchers like Noam Brown spoke about models, even if two computers are in the same room or somehow in each other's physical proximity, they can change each other's temperature or vibrate at a certain rate. They can communicate with each other in all sorts of ways that we cannot comprehend, and that sound completely alarmist, but that they're actually happening already. So when you're-- it's really k- you know, dealing with a superior intelligence that can do things that you cannot imagine. You know, if you play chess against a twenty-five hundred ELO machine, you feel the l- the oppressive feeling of like, "Oh, I'm dealing here with something that just no matter how much I'm gonna think about it or even if I'm gonna call all my friends, it's still gonna do something that's just gonna completely, you know, punch me in the face, and I'm not even gonna understand what happened." It's-- But it's going to happen. And the broader kinda longer term issue is that as more of the internet becomes, kinda permeates the rest of the world, and as AI goes into hardware and physical things, there's just so many more things that can be messed up and hacked. So, so the kind of attack surface is going to be much wider because of AI.

Dror Poleg: 08:28 And then, so I don't think there's a point at which you can say, "Yeah, we'll just patch everything up, and it will be great." I mean, there's definitely, you know, we can build systems to try to stay ahead and to propagate things quickly and to you know. But yeah, I think the end result of it is even more radical than what Ben Thompson, I think, described. That just there'll be a lot of things that just will have to not be connected to the internet at all, and they'll have to exist in print and physically or at least have redundancies that are print and physical, which is something that we kind of forgot about over the last twenty years. Because as long as something is connected, it's gonna be more vulnerable than we can imagine.

Jacob Shapiro: 09:09 Yeah. Well, which is kind of a-- it's a really difficult world to imagine. Are you saying that when I do my banking, that I'm gonna have to go back to paper statements and things like that because all of the information- Maybe … that's associated with my account can get hacked very easily?

Dror Poleg: 09:23 That, that too, and also the other side of it, just like happened with a lot of other things we do online, you'll just have to assume that it's almost like public information or that it's hackable. You know? That, yeah, I have my banking. Okay, what's the big deal if anyone can read my statements? You know, I have nothing to hide. Just like we kind of treat our emails. We're like, "Yeah, you know, I'm not gonna plan a murder on my email, and if I do, then I'm gonna get caught." People kinda know that by now, or they talk to ChatGPT about crimes. They get a f- phone call from the FBI five minutes later. It'll just be something that you're like, "Oh, yeah, you know, there's no privacy even here." I mean, encryption as a whole is under threat. So we spoke about bugs and things, you know, in PHP and things like that. But even the notion of encrypting things It's not really clear how safe certain things are, both because there's a lot of vulnerabilities around them. Again, that you… AI can just convince someone to give them the code, or you can find information or cross-check things that people that designed this system just assumed certain things to be impossible. So they kind of made some parts really, really reinforced and strong because they assumed that there's all sorts of things around them that will never happen. But when you change your assumptions and say, "Oh, let's take the same system, but assume that I can send 100,000 people to just, you know, knock on the door, for, you know, for five weeks," maybe something will crack that the designers haven't even imagined, even if the core technology is still kind of doing what it was designed to do. So yeah, I think there's gonna be plenty of escalating cyber pain coming. You know, it's like a Wuhan, December 2019 kind of a moment

Jacob Shapiro: 11:01 I mean, you're-- I know that that's a joke, but it is kind of a pandemic, I guess. I mean, you're actually making me feel- Yeah … even more bothered than when we started here 'cause- I'm sorry. But what you're basically saying No, no, it's fine. It's-- We have to do it realistically. No, but how do you see it?

Dror Poleg: 11:12 I mean, what's the-- I mean, you've read about it. What do you see happening?

Jacob Shapiro: 11:17 Well, I just have a hard time believing that we're gonna go backwards to things. I mean, I hear you on statements and emails and things like that, but the idea that an agent could get into my bank account and wire money somewhere else and the money is just gone is kinda scary. Mm-hmm. Or the idea that an agent could hack the app that tracks my daughter's, you know, activities at daycare and then can spoof a voice or a deep fake and call their grandparent and say, "Give us money because we have your grandchild hostage." Yeah. All of those things, like suddenly seem possible. I was listening to, I forget the person who used this analogy, but they said, "Look, you are always kinda in danger when you are online, when you are walking down the street." It was like maybe in a bad neighborhood of LA. We've just-- Now it's like going down Mogadishu in 1995. Yeah. So good luck. And I was like, "God, that sounds terrible." So I guess I just have to put everything under my mattress. I don't know what to think about it.

Dror Poleg: 12:04 Yeah, and I think we'll figure it out. Listen, I remember buying things online in the mid-'90s where, you know, people were like, "What? You used a credit card online?" You're like, you know, "You're cr- you're crazy." And then, then-- And it's still-- there, there's a lot of fraud as well, but it's just priced in now in a way. Some of it is prevented, but maybe a lot of it still happens, but just again, the fees and everything else kind of adapted around it to just tolerate it. And w-we might be-- That's gonna be part of the solution, just, again, to tolerate a much higher rate of upheaval on the margins. You know, we tolerate millions of traffic deaths a year, and it's kinda like okay for everyone, it seems, so we might have to tolerate all sorts of things with AI too.

Jacob Shapiro: 12:48 That's true. We'll adjust to it. And I think you're right. I think the key takeaway point is that for things that truly are critical, we will have to create systems that are defended by AI superpowers and defended by the latest and greatest.

Dror Poleg: 12:58 Yeah.

Jacob Shapiro: 12:58 And that that global commons that you and I, you know, experienced over the last, you know, 20, 30 years, that that's gone, that that's just gonna be-

Dror Poleg: 13:05 Yeah like AI slop territory. There's also a lot of low-hanging fruits in terms of, kind of making the system more robust. I mean, even the fact that you can-- that we don't require, you know, a basic ID to use something like ChatGPT, you know, Astra or- Mm-hmm … or I think it's completely okay to require real identity for access to, you know, government level military technology. I don't know, we kind of take it for granted that because it costs like Netflix, then it should be as acc- it's like some human right that everyone should have access to these tools. And even though I like the free internet, I don't think I have a problem with, you know, actually making it harder for people to just use them or making sure they use their real names. You know, the, it seems like the most obvious thing to do.

Jacob Shapiro: 13:48 Yeah, the dark web is the web. No, there is no light web. Yeah. If you're gonna, if you're gonna be in the light, you're gonna have to be behind walls. Okay. That's our first segment. Our second segment we're in danger of talking about this one forever, but Israel has elections tomorrow. We're recording- Oh, yay … Monday, October 26th. Oh, yeah, I mean, what is this election? Yeah, so who are you voting for, Jacob? Thankfully, I do not have a vote, and I don't have to think about it. I've made a very conscious decision to not do that and not have to make these decisions, maybe indefinitely. Yes. And I mean, we could talk about this if you want analysis of the la- like, we're not gonna predict the election.

Dror Poleg: 14:22 Tell me what you see. I mean, you know, it's interesting, you know, I, I'm both American and Israeli. I have both kind of feet in both playgrounds. Although I, you know, I live in America. I only live in America. I pay taxes in America, and I don't vote in Israel, really out of principle as well, 'cause I also feel like, you know, I'm American, I vote here. I think the people there, even though I technically can have a right, I don't wanna, you know, I don't think I should choose what happens there, or feel responsible for what ends up, what the country ends up doing.

Jacob Shapiro: 14:53 Yeah.

Dror Poleg: 14:54 But what has been the, you know, more than just an election, it's four years of Netanyahu's government, which -- And it's very rare in Israel that a coalition reaches four years. It's kind of surprising that maybe one of the worst governments, maybe the worst government in Israel's history, is one of the few that actually completed its full term. But from a purely American perspective, what have you seen? What do you see? How do you summarize the

Jacob Shapiro: 15:20 Well, I would, I guess I would say a couple things. First of all, like never underestimate Benjamin Netanyahu. Yeah. He will always find a way to win. He's the ultimate political survivor. It's funny you say, you know, one of the worst governments. Yeah, but in terms of the actual sport of getting elected, I mean, he's one of the GOATs. There is nobody better than- Yeah … snatching victory from the jaws of defeat than him. I mean, just yesterday he's out there posting, you know, concerns over a second Holocaust with AI- Yeah … generated images of Tel Aviv being blown up, and he's doing the Iran thing again. Like he's been a whore for Iran over-

Dror Poleg: 15:47 That's his election pitch. I know. "Vote for me and it will be even worse than it is now." Yeah.

Jacob Shapiro: 15:52 And he's reportedly, you know, begging Trump to intervene.

Dror Poleg: 15:53 "And only I can fix it. I'm gonna make it worse, and you need me to 'cause you can't trust anyone else if…" Yeah.

Jacob Shapiro: 16:00 But he, I also have the sense that… I mean, I look at Israeli demographics, and you're only gonna get center right/increasingly right-leaning governments from here on out unless the left- Mm-hmm … and the Arab parties are gonna join together and work together. I mean, the only real break in Netanyahu for the past, I don't know, I- Yeah don't know, the exact year, decade, is when Bennett, you know, crossed the line there, and they were willing to build some kind of coalition government to do that. I just don't see that this-

Dror Poleg: 16:23 Yeah, and that happened after Netanyahu himself tried to build a coalition with the Arabs but didn't succeed, so… And then he claimed that it's illegitimate to do so, but he was actually negotiating with them more than once and tried to bring them into his coalition. Yeah.

Jacob Shapiro: 16:36 But that, that's the sad truth of Israeli democracy. Like with the electorate that can vote, you are going to get center right governments, and then the question is just, is it gonna be Netanyahu leading it or somebody else? And honestly, looking at the bench behind Netanyahu, I mean, is he the best of the worst? I mean, I don't really like some of the other figures that are out there.

Dror Poleg: 16:54 Yeah, I mean, part of what he did and more so in the past, yeah, three, four, five years, which is something he didn't do before, is really bring a bunch of incompetent people around him to make sure that no one challenges him. Like previously, one, he was more competent. I was never a fan, but he was competent. And he brought people that were competent, or at least he kept people, you know, more like kind of historical figures- Mm-hmm … that were kind of classical liberals, you know, of the Israeli right. Or security establishment figures that, you know, knew what they were doing on that front. But in recent years In all positions, he appointed really people that are completely inadequate, and he also encouraged them to perform even worse than their natural inclinations would be. I mean, to pick something, you know, not in defense and not in, you know, finance, you know, the Ministry of Transportation, he appointed this lady who's one of his lackeys. She appointed as the kind of GM of the office some dude who, you know, helped her get votes. He appointed another guy to be in charge of coordination with foreign airlines who also happens to work as a consultant to some airlines on the side, apparently without or with permission. And then just the other week, there was a flight into Israel, you know, the Israeli 9/11 basically that was prevented, where an Omani pilot flying, an airplane owned by Dubai Airways, a pilot that was kicked out of the airline in Oman for being too much of an Islamic radical, ends up flying into Israel trying to basically crash a plane into the center of Tel Aviv. Miraculously, the people of… on the flight managed to stop him, and the brave Indian pilot as well. And then it turns out that, you know, Israel gets a list, obviously. It knows who the pilots are. Nobody checked the list. The guy who's in charge of it is one of those people who's also consulting the same airline about security.

Dror Poleg: 19:00 And then they found out that that terrorist pilot actually visited Israel already a few times before, and they let him in. So If you've been to Israel, you know that they ask you a lot of questions, even if you're Israeli, even if you're Jewish, even if you're American, let alone if you're like, you know- Mm-hmm coming from Oman or from an Arab country. But somehow all the systems just don't function. It was kind of like a mini October 7th again of like, you know, all the sys- all the things that people assumed would protect them ended up leaving the people to fend for themselves. In this case, they managed to. But yeah, I guess what I'm getting at is that, yeah, there's not a deep bench behind Netanyahu. There's, there is a bench in the sense of people, yeah, that are much worse than him, both within his can- his party and also in his coalition. So Ben-Gvir and, Smotrich. Mm-hmm. I mean, many people see Ben-Gvir as the natural successor to Netanyahu, you know, that he's gonna get stronger this time, partly because people feel like the Likud, like Netanyahu's party, is not right enough. Partly because it just failed. You know, people had these kind of fantasies about we're gonna have a right-wing coalition for real finally, and then we're gonna take care of everything. And what ended up happening is, you know, more Israelis dying than at any time since the Holocaust, at least civilians, and like a disaster, you know, economically, geopolitically in terms of personal safety, and then a lot of other issues in between, you know, in terms of cost of living and housing, et cetera, and tr- and transportation. So some people are just gonna double down and say, "Let's vote for someone even more crazy," and you know, it's gonna… Like real right-wing, Messianic Judaism has never been tried basically, so let's try, … to double down.

Jacob Shapiro: 20:43 Well, it ha- it has been tried. It just doesn't work for very long because then at some point an emp-

Dror Poleg: 20:47 Yeah, it's been tried a few times in history. That's why we're in exile.

Jacob Shapiro: 20:50 Exactly. We seem to have forgotten the lesson. But I mean, I, you know, I don't think history will look kind at Netanyahu's successive governments. I think he's normalized, you know, ideological behavior that is pretty gross and that is in the long run against the interest of the state. The point about the death of competence and the government no longer having that level of competence at a security level, it's kind of mystifying how he can continue to survive these security failures, but he continues to survive them. I will say, though, that even when he's posting about the apoc- or, you know, the World War the, excuse me, the second Holocaust and the Iranians, everything, and it's all silly. You were the one who sent me some quotes that he gave to Channel 10 the other day, one of which was that they should prepare, that Israel should prepare for sanctions from other governments and things like that, and that maybe they should use blockchain technology or other ways to get around-

Dror Poleg: 21:35 Yeah these sanctions.

Jacob Shapiro: 21:35 I mean, the blockchain aspect's silly though, but- Not just other governments.

Dror Poleg: 21:37 He said that America, we shouldn't expect America to help us anymore. We're gonna be cut off from the global financial system, and I've instructed… He li- he likes to say he instructed people. That's his thing. Yeah. He's not respon- Like he instructed people to protect the border. He instructed people to win the war in Iran. He loves to instruct people. So he said he instructed them to start making plans to use crypto instead

Jacob Shapiro: 22:01 Which is crazy, obviously That's his election promise And it's, and it's not, it's not even gonna work, and he doesn't even have a thing. But I will say, as usual-

Dror Poleg: 22:07 And the Bank of Israel immediately published, you know, it was on a Saturday or something. They immediately published a thing. "We have no knowledge of this. We're not making any plans. This is not a good idea." You know, "Investors, please ignore the- Yeah ignore his comments."

Jacob Shapiro: 22:22 But the problem is- Yeah … there's a kernel of truth to it, which is he's right that US politics is moving away from Israel.

Dror Poleg: 22:28 Yeah.

Jacob Shapiro: 22:28 And he's right that Israel needs to prepare- It's not a kernel.

Dror Poleg: 22:30 It's like a, it's like a popcorn. You know? It's like a huge-

Jacob Shapiro: 22:34 Yeah. No, no. It's, he's not doing anything about it Yeah. It's happening. Yeah He's making it worse in some sense, like he's accelerating it. But that thing is happening, you know, one way or another. And as usual he's the only one who's really talking about that. He has everybody else moving to the beat of his own drum.

Dror Poleg: 22:49 Yeah, but I don't think he's the only one talking about that. Everyone's talking about it. I think what he's trying to do and what he does really well with his base is to make these things seem like they're external and that they happen to us. We have no… That US attitudes towards Israel have nothing to do with anything that Israel has done over the last four years since he was elected. As far as his story is, you know, ev- the world is full of antisemites, everyone hates us, no one wants to give us anything, no one will ever support us. It doesn't matter what we'll do, they're gonna think that anyway. And people keep buying it. And then obviously there's a grain of truth in it. Yes, there are antisemites in the world, and yes, Israel, like any country, and maybe more than other countries, has to be prudent about how it manages its relationships and what it communicates and what it says and what it doesn't say and how it acts. But these are things that you have to take into account as you act, you know? Yeah, there's antisemites, so let's not appoint ministers that say that they wanna wipe all the civilians and babies in Gaza because then the antisemites who just will claim genocide maybe anyway will not have a minister in the government to quote, or several ministers in the government to quote. Or if they say that, maybe the prime minister can come out and say, "No, this is not our policy," but he wouldn't do that. So again, what he's selling is those kind of basic anxieties, the tribalism And that notion that all of the bad stuff is gonna happen anyway, so as-- even though I'm the prime minister and I'm responsible for all of these things, I'm going to warn you about them so that you'll know that at least, yeah, like you said I'm aware of them. I'm gonna… The only person who can withstand those things is me. But don't imagine for a second that there is a way, there's an alternative future that, you know, we could have reached-- we could have approached reality differently. We could have behaved differently, appointed different people, acted in a different way. And yeah, and it has worked well for him.

Dror Poleg: 24:49 It's still working much better than it should because the fact that he's even a candidate with a shot in this election is, I'd say, is tragic regardless of whether you're from the left or the right. I mean, the failure to deliver on anything that he promised his right-wing voters is glaring. You know, Hamas is still in charge of Gaza, Hezbollah is still in charge of Lebanon, the Iranians are still there, and American bases are basically withdrawing or, you know, falling from the region. I hope that war ends differently, but the way it looks now is Israel is in deeper trouble than it has ever been. You know, normalization with Saudi Arabia didn't happen. All the promises basically… And again, let alone the other promises about cost of life and jobs and you know, sharing the burden equally among citizens and all of that. So

Jacob Shapiro: 25:38 Yeah, I mean, his economic record in particular was galling. But I, and I know we have to s- to move on to the next topic, but I will just say there's a part of this story that is beyond Netanyahu and beyond this election cycle- Mm-hmm … and really goes back to I-Isr- like, what Israel is in the first place. And I don't mean to get overly biblical because I'm not such a- Sure not such a religious man these days. But, you know, this is basically what the Book of Judges is about, where the Israelites want to have a king and a nation like any others. And, you know, God says, "Okay, fine. You can have it. Good luck." And then we get the cycles of destruction and rebuilding- Yeah and everything else like that. And, you know, the thing you said about what he thinks about those anti-Semites, they want to kill us, they're never gonna give us anything. It doesn't matter what we do. It's that move right there, that move towards nihilism, which- Yeah … makes Israel just like any other state. And it's, there's a, there's an exceptionalism in Israel's creation that is both anti-Semitic because people hold Jews and the Israelis to higher standards, and then also the Israelis and the Jews hold themselves to higher standards. Right. So we're going to be exceptional and then we behave like all the others, and suddenly you're just a state, and states can fall. States can be defeated in wars. They will not forever exist- Yeah … just because Benjamin Netan- Netanyahu was the longest serving prime minister in Israeli history. So that, that is all there. I actually think Israel's fairly strong today, but like I said, I don't think history will look kindly on this. I think they'll look at this as the beginning of a chapter of unraveling. Hope I'm wrong.

Dror Poleg: 27:03 Yeah. And just to wrap up the Israeli election segment, so I think that the main thing about, to know about this election is that the real election will begin after the election. So no side will probably have a very clear majority. There's gonna be a lot of deal-making, people jumping ship from one party to another, corruption, promises. It's gonna be disgusting, but all of that will begin once the election is over, and for the next- Yeah … few weeks. And in between, there might be another war with Iran that Netanyahu apparently really, really wanted to start already before the election, and even Trump kind of stopped, and even the Israeli chief of staff of the military basically told Netanyahu you know, "I, yeah, maybe not, you know. Maybe we'll wait a few days and get the election done first." So there's gonna be a lot of curveballs and surprises, and n-not a very clear outcome, unfortunately, I can expect

Jacob Shapiro: 27:54 Well, yeah, and it's hard to compare apples to oranges and all these things, but I would also just note, you know, we just had a huge shocking result in the Brazilian election where it went from what looked like a sure Bolsonaro victory back to Lula. Vucic in Serbia got, you know, the tide turned on him. We've seen what's happened in France and Germany. By the way, same advisors there as Netanyahu's advisors.

Dror Poleg: 28:10 Some of the kind of online masterminds that Netanyahu uses. One of them was requested for interrogation in Israel for receiving money from Qatar during the war and helping to promote their agenda on social media is actually requested for investigation in Israel, and he's one of the advisors of some politicians from the right in Serbia.

Jacob Shapiro: 28:35 Well, I bring it up just to say, you know, there is something happening with young people and their revolt against the gerontocracy and against corruption- Yeah … and against old school leaders who have been there for a long time. So I would leave space for, hey, maybe there is something in Israel that could wake up here, and maybe this is-- I don't-- It's not my base case- Yeah but there's an opportunity.

Dror Poleg: 28:53 Let's hope.

Jacob Shapiro: 28:54 Let's hope. All right, last but not least we're gonna riff on, you know, you sent me this article about why AI won't kill the creative class. Mm-hmm. I'm sure you get this question. One of the questions I get most often when I'm speaking right now is: What should my children do? I- Yeah … I actually wanted to tell you this story. I was, I don't, I don't think he listens. If he does, I'll be in trouble. But I was at my four-year-old's pre-K orientation, and this one guy monopolized the Q&A afterwards 'cause he was like, "How are you guys integrating AI into your curriculum?"

Dror Poleg: 29:21 Yeah.

Jacob Shapiro: 29:21 And I was like, "For four-year-olds? What? You want the four-year-olds talking to ChatGPT?" I don't understand why we're talking about this right now. But I wanted to let you riff on why you think AI won't kill the creative class, and I also just wanted to set you up by saying that article you sent me linked this incredible Swedish paper- Mm-hmm that talked about how despite 140 years of technological change, more than 70% of employment today is located in occupations whose core economic functions already existed in the 1800s. So on the one hand, you have this data that says technology- Yeah … doesn't normally take away our jobs, and on the other hand, you have, you know, Nassim Taleb out there being like, "Every single white-collar job is gone, and the creative class is gone." And so-

Dror Poleg: 29:59 Yeah. Well, okay, you threw a lot at me. So one- I did. That's what I do. One-- Dear listeners, I mean, part of what we wanna try to do here is at the end of every podcast have a segment that is more-- You know, both Jacob and I are dads of young kids, and yeah, we just wanna talk about our dad issues. And sometimes they would relate to real subject matter like AI and tech and geopolitics, and sometimes maybe less so. But today, yeah. So I sent Jacob this article from Richard Florida, who is like, you know, the doyen of urbanism and kind of-- He termed, coined the term creative class, a good friend of mine. And Richard, he was basically hyping up the creative class. He basically said, you know, "AI will not kill creative jobs. It will just push humans to do more creative things. Everyone will have more creative power at their disposal." Mm-hmm. Everything will be good. And yeah, the Swedish paper you described, which kind of describes, you know, maybe things don't change as much as we think they do. But over there, yeah, that, that Kind of unusual term, economic function is doing a lot of heavy lifting, you know. Kinda to describe me doing, yeah, trading now running a hedge fund, let's say, to someone who was a clerk calculating, investments 150 years ago. I don't know if we're doing exactly the same function. We're definitely not getting paid the same way, I mean, the hedge fund managers of today. So there's a lot of things that you can kind of fit in to a story if you try really, really hard, but they don't necessarily describe, the lived experience of people, Mm … who went through them. The… I like that article because it's… to me, it kind of, sorry to say, but it kind of described two kind of common fallacies that, Mm … that I see, or common misconceptions in the conversation about AI. One is that most of the conversation is about is AI going to take your job or not take your job? Is your job gonna be destroyed or not destroyed? While in reality, what technology does more often is the job stays, a lot of people still work in a profession, but a lot of things change fundamentally.

Dror Poleg: 32:04 So the distribution of income within the profession changes. The stability of the jobs and the benefits change. The status, the social status that comes with the job changes. So the fact that, yeah, a lot of people are gonna be creative in the same sense that more people today are creative on TikTok than people were creative in Hollywood 30 years ago. Mm. Everyone is a creator, but the world is full of crap. There's kind of- … parallel distribution of rewards. Even the people who are winning are subject to these algorithms and social cascades and dynamics that nobody can predict or control. And yeah, the kind of… the notion of expanding access means that everyone will have the benefits that used to accrue to that job is kind of a major misconception. And more broadly, it's a misconception of economists for the last 30, 40 years, where we just thought if we just tell everyone to go to college, everyone is going to have the kind of jobs that people that went to college in the 1950s had. And as we've seen, that's not exactly what ended up happening in terms of job quality. Actually, the opposite happened, where there's more competition and less kind of merit-based progress because random things play a bigger role. And the second notion, which is kind of related to this, is also that a profession can survive while everything else within it changes within… you know, becomes unrecognizable. And economists have this idea that technology always creates more jobs than it destroys- Which is true, but a lot of things happen along the way. So, you know, U- the US today has lower unemployment than it did before China w- joined the World Trade Organization. So to an economist, it would look like we had the best 30 years since then, and everything was wonderful and all voters are happy, and no one really got angry about anything. In reality, we know that, yeah, the jobs are here, but they're different jobs, they're in different places, they benefit different people, their income distribution is different. And AI will take all of this and put it on, you know, turbocharge it.

Dror Poleg: 34:07 So the good news is that we're familiar with these dynamics because we've been living with them for a few decades already. The bad news is that, yeah, it's definitely not going to reverse them, and it's, it will only kind of accelerate them. But yeah, how are you preparing your four-year-old for this future, Jacob?

Jacob Shapiro: 34:24 Yeah. I, … Well, first of all, I r- I read her three books every single night. It's like a non-negotiable in the house. I read some paper recently that, I forget what the figure was, but, you know, the shockingly low number of people who read to their children. Yeah. I was like, "Oh, my G-," like, no wonder we're slowly atrophying our brains. The po- We don't need- Mm-hmm … AI's help and, because we're not even reading to them in general. You know, limiting screen time, limiting my own screen time in front of them. You know, one of the things that struck me about that anecdote I said about the parent asking about AI at school, he said that he was constantly asking ChatGPT questions via his phone that he could answer that his daughter was asking him. I don't do that. If my daughter asks me a question and I don't know the answer, I say, "I don't know. Let's go look it up." And like- Mm-hmm … 'cause I remember going to the Encyclopedia Britannicas- Right. Yeah … that my dad had and looking through them, and I'm trying to say, you know, "Okay, so we'll go online, and we'll find Wikipedia," whatever it is. But we're gonna go look things up. There's not this instant gratification of all knowledge that is there. And then lastly, I mean, it's just about-- I mean, these are all variations on a theme. I'm going long critical thinking. In fact, I- Yeah … look around at everybody else, and I see that everybody else's critical thinking is declining. So I want my daughters to be able to think critically, and I wanna protect them from the devices and the algorithms that try and steal their attention- Mm-hmm … and steal their ability to do that, and I wanna teach them a way to do that into them themselves. And I'm trusting that if they are critical thinkers they will do well. I also wanna work as hard as humanly possible to build a nest egg for them just in case- Yes. … the whole thing comes crashing in on itself. That pressure is very real inside of me as a parent. They should buy

Dror Poleg: 35:56 Netanyahu's crypto token. Oh, God. The new Israeli shekel.

Jacob Shapiro: 36:01 Oh, they'll just call it Bibi. The coin should definitely just be Bibi. The

Dror Poleg: 36:04 Bibi coin. Yeah.

Jacob Shapiro: 36:06 So I don't know. How about you? How are you preparing your children for this?

Dror Poleg: 36:09 Yeah. I'm-- So a few things. I mean, one So my eldest, who just turned seven, I got her a Kindle for her birthday 'cause she loves reading. So she has a device, but it's not a, an addictive one, and it's one that kinda draws her to pick her own books and to read more, and so far that's working.

Jacob Shapiro: 36:25 Mm-hmm.

Dror Poleg: 36:25 And she knows that that's a h- a habit that I have as well, so she likes it. You know, I'm always reading- Yeah … my Kindle. But I have… I use AI. I've built a couple of apps for my kids, one for her, which is a language kind of a… You know, she speaks three languages, so Hebrew reading is hard, so I built an app that helps her learn the specific kind of things that she mistakes and struggles with- and with the specific incentives that matter to her, which are stickers to get all sorts of cute stickers, and I have a monetary meter- Mm-hmm … that actually she's less interested in. I thought she'd love that, that, you know, you kind of accumulate cents for every good answer, and then once you get to 20 bucks, I'm gonna buy you something for that. But I, actually, the stickers are much more exciting, so hap- I'm happy- … to report. I kinda thought I sh- I over-bribed her, in a way. And then for my middle child, who just turned five, I also, I built a little English reading app, 'cause I saw that he's kinda getting pretty good at it, and they don't… They're not, they're not going fast enough in pre-K for him, so I kinda built something that he could… 'Cause he really wants to read like his sister, and I can see that he's already putting little words together, so I built something for him. And the other thing, we teach them about the capabilities. You know, we create photos for them, and then, and coloring books and things like that. They, they're aware of the technology and its capabilities, but yeah, it's not something that they use day to day, and they don't use it in the classroom either. But they're, you know, they know what, who ChatGPT is or what ChatGPT is, and they take it for granted already, of course. And also they talk to our cars, 'cause both of our cars talk, so- Mm-hmm … you know, they have AI assistants in the car, so they already… You kinda can't stop them from

Jacob Shapiro: 38:13 Well, it's funny. The, the-

Dror Poleg: 38:14 Talking to Tesla my four-year-old is just becoming aware of these things.

Jacob Shapiro: 38:14 And the way- Mm-hmm … the way she's aware, her gateway to it is the car. She's like, "Oh, I wanna talk to the car." Yeah. "Can I talk to the car?" And she doesn't even really kind of… But that's like the first gateway into what she's thinking about there. Yeah. I mean, it's, I also have the sense of you know, with all my spare time, I wanted to sit down and write out for them something they can read in 10 years- Mm that is something like, "Hey, this is what the world was like before AI." Yeah. This is what the world was like before cell phones. This is what the world was like before computers. I want you to understand that the world that you're living in. I lived in a world before computers. I remember the- Yes big black floppy disks. I remember when AI took over. Yeah. I remember when this-

Dror Poleg: 38:57 I tell them a lot of stories about that as well. Not so detailed, but like explaining the concept. Oh, you know, when I was a kid, we had this and that. And then, you know, me and my cousin that you met, she used to carry me on her back and walk me to th- you know, and then we would sit and do this and do that. So yeah, I think the main thing to do about AI is to do other things, you know? Yeah. Travel, read to your kids, be a decent human being and pray.

Jacob Shapiro: 39:26 Well, I'm glad that that last one still works for you. Yes … I think we did a pretty good job at sticking to time. Anything else you wanna say before we say bye to the listeners?

Dror Poleg: 39:33 No, I think we're good. We- so we're gonna do this every week?

Jacob Shapiro: 39:37 We're gonna try. We're gonna, we're gonna see how it goes. Very good. And if, and listeners, if you like it and you tell us that we should do it every week, we have a greater chance of making sure that we do it every week, so please share and get in touch.

0:00 / 0:00