I've Got Questions with Sinead Bovell - The AI Fight You Need To Be Watching.. (& What Everyone Is Missing) | IGQ Live

Episode Date: March 5, 2026

In our first ever IGQ live, I break down two of the biggest stories in AI right now, and take live questions from the audience. The episode starts with the growing standoff between Anthropic and the United States Department of War, and the pressure government is placing on the AI company to leverage its technology in ways that violate the company’s ethics, and what this moment means for all of us. I also unpack the viral Great Intelligence Crisis scenario from Citrini Teport. A scenario that explores a near future where AI rapidly automates white-collar jobs and has a cascading impact on the economy. I breakdown what it gets right, what it gets wrong, and the real risk we need to be paying attention to. I also take questions directly from the I’ve Got Questions audience, before closing with my perspective on what all of this means for us today, and how individuals and leaders should be thinking about the moment we’re in. If you want to attend the next IGQ LIVE, make sure to subscribe to my newsletter and socials to stay up to date. Follow my work here: Substack: ⁠⁠⁠⁠⁠https://sineadbovell.substack.com⁠⁠⁠⁠⁠ Website: ⁠⁠⁠⁠⁠https://www.sineadbovell.com⁠⁠⁠⁠⁠ Instagram: ⁠⁠⁠⁠⁠https://www.instagram.com/sineadbovell⁠⁠⁠⁠⁠ LinkedIn: ⁠⁠⁠⁠⁠https://www.linkedin.com/in/sineadbovell⁠⁠⁠⁠⁠ Twitter / X: ⁠⁠⁠⁠⁠https://twitter.com/SineadBovell⁠⁠⁠⁠⁠ YouTube: ⁠⁠⁠⁠⁠https://www.youtube.com/Sineadbovell⁠⁠⁠⁠⁠ TikTok: ⁠⁠⁠⁠⁠https://www.tiktok.com/@sineadbovell

Anthropic is drawn a line in the sand that their AI system cannot be used for mass surveillance of American people, and it also cannot be used for autonomous weapons. The Pentagon had given Anthropic until Friday to either draw back their guidelines, so the Department of War could use their AI systems as they saw fit, or Anthropic would be blacklisted from the US government. This is really significant in geopolitics and in technology more broadly. This could be the moment where American tech companies are fully seen as one with the US government. What does that signal to the rest of the world?

This is our first IGQ live. We thought it was time to answer some questions, to share things of some more breaking news, because, I mean, there is a lot going on in this moment. Some of the topics that I want to get covered tonight are Anthropic versus the Pentagon and that showdown. We're going to talk about this in Trini Report, the Great Intelligence Crisis,

and their take, their fictional take on the future of white-collar jobs or the lack of future of white-colored jobs. So those are some of the topics, but I, but if you have questions, throw them in the chat, I can see them. So we will be sure to dive into them. And so the first topic that I wanted to to dive into today is the Anthropic versus the Pentagon. And if you're not familiar with what's actually happening here, so Anthropic is one of the leading frontier AI lab. So when it comes to aspects like coding, language, Anthropics Claude model is starting to take the lead. In fact, last summer, Anthropics signed about a $200 million deal with the U.S. government.

And they are actually the only AI company that their models can be used in classified tasks. However, at that point in time, when Anthropic did sign with the U.S. government and the Department of Defense for use of their AI model clawed, There were some stipulations, like it can't be used for things like mass surveillance of American people, or it can't be used for things like autonomous weaponry. So Claude has since become quite entrenched in a lot of the Department of Defense's activities. And some of this is actually things like administration. You can think about some of the tasks and activities that they do are on computers from the 80s or such old software. When you bring a technology like artificial intelligence, you're going to streamline things. You're going to make things a lot faster. You're going to make things a lot more efficient depending on how the technology is used, right, and making sure that it's used within the ethical guidelines. And historically, Anthropic has positioned themselves as the AI company that does care significantly about ethics. They are the company that has a constitution that their AI model, Claude, has to follow. So these are the policies like it should be harmless. It should be polite.

all of the different guiding principles of their AI model. And they actually are one of the AI companies that advocates for regulation. And meaningfully, science petitions and favor of regulations when their competitors do the exact opposite. So long behold, it's been over six months. Anthropics, Clod, their AI system, has been deeply entrenched now in the Department of Defense's systems. And apparently they really like using Claude. And Claude has been on a tear.

Everybody has been posting about how good this model is and just how much better it keeps getting. Anthropic finds out maybe a few weeks ago now that Claude was actually used in some of the activities in Venezuela that the U.S. had undergone when they had seized the president of Venezuela and captured him, brought him to the U.S., Anthropics' Claude model was a part of those operations. So Anthropic, of course, raised a few questions about how their model was used. They had had some stipulations about how it couldn't be used. And this is the beginning of the tensions about a month ago, starting between the Pentagon and Anthropic. Fast forward to today. Anthropic is drawn a line in the sand that their AI system cannot be used for master gallons of American people. And it also cannot be used for autonomous weapons.

So it can't be used to fire a drone at a target without anybody in the loop. right, so it can't be used to execute command without any human in that loop, which is pretty reasonable. And it's not just reasonable from state of ethics, from the debate around autonomous weapons, which is a whole other thing. LLMs, right, to the large language models, which is the type of AI that underpins a chat bot like Claude, these are prediction machines. They're not necessarily world simulators. They don't have live context. situational awareness. So they are not actually equipped to be making a final fire decisions without a human in the loop. And the CEO of Anthropic came out and said just that,

nor models aren't even in a place where they can do that. The Pentagon has had given Anthropic until Friday to either draw back their guidelines and remove their guidelines so the Department of War slash Department of Defense could use their AI systems as they saw fit, or Anthropic would be blacklisted from the U.S. government. Anthropic refused to back down, and now they have since been threatened with this blacklist. But the blacklist is actually really serious because the Pentagon is seen that they would be actually a supplier risk to the U.S. government, which means any company that the government works with can't work with Anthropic.

And this gets really tricky, right, because Amazon works with the government, for example, works with the Department of Defense. And that's the cloud infrastructure that Anthropic uses. So this could essentially crush Anthropic, depending on how far the U.S. government takes this. And this is really significant for a multitude of reasons. And some people may say, okay, well, anthropics should just hold their ground. And, you know, this is ridiculous. You can't just, I mean, it's a free market.

How can the government in America come in and say, you have to give us your AI model? And the U.S. government actually can do that because they have the Defense Production Act, which is like a Cold War era industrial policy that was used in times of war to ensure that the U.S. had enough goods and services to support their national security and defense posturing. So that defense production act, and I want to read some of it, that's a bit more specific, it would give the president broad authority to direct the private industry and the name of national defense to produce. produce a good or a product that the government deems necessary. And the U.S. government is threatening to use the Defense Production Act on Anthropic and allow Anthropic to give them their models without their clauses of safety and security attached. And it doesn't stop there. It's also possible that they ask for a version of the product that is retrained without the safety guardrails that are baked in to the AI system as it stands.

today. So this puts Anthropic in a really bad position. I mean, the U.S. government is a huge client for Anthropic, right? It was, I think, a $200 million contract last summer. So to lose them would be really significant to their revenue channel. Perhaps not even, you know, they wouldn't even be able to sustain after that. It's hard to say. Over 500,000 people did download the app over the weekend when they heard that. I think people are now starting to see Anthropica as the Robin Hood. for the American people or just for the people more broadly. But what's really interesting about a move like this, so wherever you stand on national security,

that companies should be willing to support their government and make sure that they're patriotic. And if the frontier of defense is moving towards a world in which AI is going to be a big part of it, which is true, domestic companies should want to support their government. So even if that is what you're posturing is, putting a U.S. company in this position could actually significantly backfire against the U.S. and their competitive position. So point one, if I am,

even an ally of the U.S., and I see that the United States is saying, okay, we are going to unilaterally position ourselves at the top in this AI race, we're going to do what's unilaterally best for us, even if it means subjecting an AI company, a free company in our market, to go against their will. if I'm an ally, certainly if I'm a competitor, that's going to make me think if they're willing to do this with their own company. What happens if I installed that AI company into our government or into our domestic companies? And the U.S. government turns on us as a country. Are they going to also take their models with them? And does that mean that we are vulnerable if we depend on an American AI company for our country's AI?

for our government's AI, for our private sector's AI systems, this could end up turning other countries, even allies, away from U.S. companies, because they deem it as risky, unreliable. And then it's like, okay, well, the U.S. models still are the best, so it doesn't really matter. Smaller regional powers are going to be forced to use the U.S. because they're the best or they're just going to have to suffer from not as sufficient or not as capable models.

That could be true in the short run. But you have to also remember that talent goes with it where they're motivated to work. Jeff Hinton, one of the godfathers of artificial intelligence, moved to Canada to build his AI dreams specifically because of the U.S.'s relationship with nuclear weapons. And he wanted to make sure that the AI that he was building wasn't going to be used in unethical ways.

So he went to Canada. researchers will follow if they or will move if they don't feel the work that they're doing is in line with their ethical value. So there's flight risk of people. Also, necessity is the mother of invention. So if all other countries feel like, okay, the U.S. is going to pursue what's best for them unilaterally, we have to really get together and figure something out with our own resources. So eventually it could make other countries more competitive and actually increase the competition against the U.S. And it could also lead to a lot more unethical use of the technology globally.

Because now if you are a competitor, an ally, you know that all bets are off, that people are throwing ethics out the window when it comes to how they're procuring these AI systems. So if you're another country, you're thinking, well, I better do that too, because if they're willing to put their own companies in a state of panic about how they're going to use the technology, imagine what's going to happen in the field. So this could also kickstart a more fierce AI rate. with other countries, ones that were considering ethics and now may have second thoughts about it. And ones that already are like, yeah, ethics aren't really at the top of our list.

And now after this, they're like, oh, for sure not. Forget ethics. Moving on to the next slide. So this could cascade in really dangerous ways. So these are some things to think about that I think are not a part of the general conversation right now. Where Anthropic goes with this, I genuinely don't know. You can be threatened with jail if you oppose that the first. fence production act. You don't really have a choice. So this puts the company in a really tight

ship. I can tell you that watching online, people respond to Anthropics' strength in this moment has been really something. It does seem like OpenAI has signed on with the Pentagon and followed the will of the Pentagon and so has XAI, probably unsurprisingly. So what Anthropic decides to do, it's actually going to be quite significant. And this moment, yeah, it's much bigger than Anthropic versus the Pentagon because other countries are watching this moment. And this isn't against the backdrop of global peace. I mean, look at the moment we're in. Right.

So what is this going to inspire other countries, other armies to do? And maybe this is what actually kickstarts innovation elsewhere when other countries start to combine their regional power to build something. And this could be the ticket that shows, you know, we really can't, you know, Most countries couldn't really depend on each other in this moment anyways. But tech companies were seen as, even though, of course, they have to follow national priorities, they were still seen as global. This could be the moment that that stops, where the technology is now seen, American tech companies are fully seen as coupled as one with the U.S. government. And they are threatened by the government if they don't follow orders. What does that signal to the rest of the world, right?

So I think that this is going to be a pretty big turning point in geopolitics and techno politics and in technology more broadly. Okay, I'm going to look at some questions to see if we have any questions about this. Yeah, it's a full-time job to try to keep up with this. I mean, yeah, it is. It happens to actually be my job and I feel like I'm spinning. Love Claude. Yeah, Claude, I mean, and that was a whole other thing, right? The frontier of artificial intelligence, it's really hard at this point to say.

say, is there a company that's going to win? The models are, like, one goes ahead for three months and then the next one goes ahead for three more months. It's really hard to say who's where. But Anthropics Claude was starting to pull away in a lot of significant functions, like coding in particular, but also people liked it for writing. The fact that the U.S. military, it's one of their favorite AI models, also says something that they don't want to pull away. And there was actually a funny, well, not funny, but given the times, quote that the CEO of Anthropic had stated that, because he says, there's two threats the U.S. government is placing on us.

One labels us as a security risk. So if the U.S. government says that Anthropic is a security risk and blocks them from any contact with any other company that works with the government, the government's claiming them to be this massive security risk, and that hasn't happened to an American company before in that way. On the other side, they're saying that it would threaten the U.S.'s national security if they couldn't work with Claude. So either way, it's like, how are they a security risk if they're so essential that it threatens the national security?

So the whole thing obviously isn't making a lot of sense. And then, of course, from the free market argument, not a big look. But, yes, so if there are questions on Anthropic, the issue, issue of dominance by the U.S. and China. So, yeah, what could this mean for the U.S. and China? I mean, China is obviously watching this moment. China, though, is interestingly taking a very different approach to AI dominance. So United States tech companies are pursuing AI dominance through the lens of, we have to make the smartest, most capable models. That is the U.S. company, the private sector strategy, and therefore the U.S. as a nation, that's their strategy.

We'll get the most intelligent, most capable models, and that's how we are going to dominate the AI future. Then you have China that's doing a very different strategy. It's like one country's playing chess, the other country is playing poker. China's models are quite capable. They, Bidant has some very capable models. Baidu has very capable models. Deepseek.

But they're not necessarily at the frontier. But they are pursuing a strategy of diffusion, of deployment, of ensuring that the rest of the world deploys and depends on Chinese AI models. So they're giving their models away for free, what's called open source AI. And they're not only giving their models away for free, they are, especially when it comes to the global south and an emerging market economies, they are sending computer scientists there. They are saying, okay, we'll give you cloud, we'll give you compute for free, which is really expensive. We will give you training for you and your government and your family and your friends. We will give you the whole package for free. And all you have to do is use Chinese AI models.

That is a very different strategy. Dominance in terms of intelligence or dominant in terms of the whole world depends on your AI system. And once that AI is that foundational layer, that infrastructural layer, it's really hard to switch. Your developers in your country start to get used to it. It starts to get embedded in different entrepreneurs. workflows, different company workflows, different government workflows, to China's pursuing a very different strategy. This could also end up making Chinese open-first models stronger because countries will feel like, okay, this is less of a risk. You give an open-source model away. You don't really

own what happens to it after. It doesn't mean that it's not a national security risk, right? You can, you embed your country's values in that AI model. The model is a reflection of the priors, whoever claimed history in that country, all of that gets embedded into the model. It's like a soft power asset moving into a different country. But countries may feel like, okay, we're going to go with open source models because we don't know what's going on over there. And this country, you know, America's pursuing their unilateral interest. We're going to have to go with something else that's free and quick and that we have control over. And that could be open source. And right now, China does lead in open source. So this could end up strengthening China's position. You can't

look at technology in a silo. You can't look at national security in a silo. You can't look at geopolitics in a silo or energy. These things are deeply interlinked. And they have the system cascading impact. So if you only take this kind of one unilateral lens to it, you end up doing things that could be against your best interest. Okay. So we have a question here. Jessica Brown, do you think AI tech companies acting as one with the government sets a dangerous precedent around technological warfare? So it depends on how you look at it, right? because governments have always procured contracts from the private sector. And that's how the arrangement has historically worked in democratic countries.

Governments aren't as efficient at technology and science, although they fund most of it. So most of the tech giants and a lot of the technology that we see in the United States in particular was born off at the back of American research. So government research, government grants, DARPA, universities, academic labs, and then they essentially give this for free to the private sector. But the private sector is more innovative. They can move faster. They can get loans and raise capital at better rates.

So it does make sense in some ways that innovation comes from the private sector. And the government isn't like just trying to build science experiments, you know, at the White House. It doesn't necessarily make sense. But you come up against this ethical wall when you realize what tasks and services civilians share with technology, with science and technology equipment that could also be deployed for war. So it's one thing if you are, you make drones for defense contractors. People seem to, no one's really too worried about that. You do a deal with the Department of Defense. Nobody's too concerned about it. But if we're also, if we're using your technology for shopping and for

sending photos about our families with and then we also realize, oh, we share the cloud with an AI system that's also at the very same time conducting war operations that really intersects with our ethics and with what we think is right. So where companies and countries stand, I mean, it's actually a really tricky one. And I actually put that back on people, right? Where do you think the line should be drawn? Do you think it's on governments to invest themselves in an innovative, And to make sure, because at the end of the day, too, defense and national security is deeply coupled to the technology of the moment, right? So when the horse strap was invented, it's actually really funny.

When you think back into the medieval times, when that horse saddle was invented, it was actually one of the biggest technological leap forward at that time, because it allowed soldiers to be sturdy on the horse and therefore much more fierce in war. So that was a huge technological leap forward in warfare. And the same with cannons and gunpowder and all of these things. So the frontier of war technology continues to move forward. So then the question becomes, okay, in this moment, should the government actually be developing these technologies themselves and just leave the private sector out of it and not have to drag them in? Or is it better for the private sector, which is faster and more innovative to build technologies that could be actually safer and cheaper and send those to the government?

And that's actually part of what a democracy is for, right? How do the people think that those types of problems and issues should be solved? On the one hand, you could raise taxes a little bit more and fund science more, and that's actually what used to happen decades ago, and that's when NASA was huge, and not that NASA's in the name of defense, but as tax dollars and got reallocated and as taxes came down, different tradeoffs and decisions have to be made. And so that's what I think, you know, the people should vote on that. Where do you think who should be building these technologies of the moment that countries believe are needed to keep them safe?

Should that be in the private sector or is that, you know, government, you've got to figure that out? So that's one thing. Yeah, I think I'd put that back to the people. Okay, so this question is from education, entertainment. Do you feel anthropic overreacted as I thought the military already used AI systems and making military decisions? Okay, this is a great question. AI is everywhere in the military, in all militaries, in every country. And it has been this way for close to probably more than a decade.

And not necessarily AI as we use it today in the chat bot form, but machine learning and deep learning. I mean, a lot of what happens on the ground for militaries is based on what they see in satellite imagery. And there isn't somebody in an office in a clerk somewhere, looking at 7 million satellite snapshots. That's artificial intelligence, analyzing high risk areas or targets and that sort of thing. So AI has been a part of the world of war in a terrifying way for a really long time. So yes, you're right. That part isn't new. What is new is one, like this kind of chatbot style interface in war.

Two, LLMs and generative models being a part of planning operations, being a part of identifying targets, the type of data and the amount of data that they can analyze in a multimodal way. that is new. Three, it being so intertwined now with commercial companies having these capabilities, even though some commercial companies have been working with a lot of governments for a while. So did Anthropic overreact? I mean, I think not wanting your model to be used for mass surveillance of the citizens of the country you belong to, I think is a pretty fair red line.

And the line that they drew about not having clawed or their AI model clawed, be a part of autonomous weapon decisions, be in that essentially that kill chain. It's not just because Anthropic had a moral stance against autonomous weapons. It's also about the technical limits of the technology in this moment. Large language models and generative AI systems are statistical prediction machines. They're going to be prone to error. They don't, you know, next token prediction isn't the next necessarily. the best strategy for a wartime, real-time situation that involves a lot of situational awareness

that AI systems don't have, a lot of context and judgment that these types of AI systems don't have. I mean, if you look at a lot of the academic research on using large language models in war games, large language models consistently escalate in the games to nuclear as the option. So this has been done repeatedly, different academic labs, sometimes the AI labs themselves, but sometimes different universities, and they use AI in war gaming. And war gaming is obviously just the games that they use in war to understand different scenarios and to make plans. And when they started to use LERD's language models in these games, the LARD's language models would mostly always go to the nuclear auction, go nuclear. And to the question

becomes, well, why does this L, why does this language model want to just detonate? And that seems like the best, the best slide to go to. It's because when you think about what these AI systems are trained on. They're trained on tons of fictional stories, fictional and nonfiction stories about the Cold War, about nuclear arms races, about different escalation dynamics in war. And so all of those movies, all of those scripts, all of those Reddit threads, all of those blog posts, all of those books are embedded in these AI systems. And that's what they're making their decisions on. So not just real-life situations. Poems. And nonsense. And that's a part of the war strategy. So these systems actually aren't ready to be in a literal kill chain. It's a really terrifying thing. So Anthropic isn't wrong for drawing that red line there. So I don't think that they overreacted. And I also think even as a private company to be able to say, I think we should be able to draw the line of how our products are used. And of course, you know, in times of war or in times of emerging.

like in COVID, they did enact the Defense Production Act for things like face for masks and vaccines and that sort of thing. So if it's a dire time of emergency, things change, but that's not the moment that we're in. So yeah, so that's how I would answer that question. Okay. If the government doesn't want to face ethical pushback like with Anthropic, then it has to build its own AI and leave private companies out of it. That will only leave the citizens as the last line.

Yeah, I mean, it's a really tricky tradeoff, right? Because on the one hand, of course, the private sector is more innovative and they move faster and they get capital at much cheaper rates. And they're just tend to be a lot better with science and technology and with markets. And government's best role is guiding the private sector in those ways. But yeah, when you come up against a red line of ethics and how citizens end up sharing that technology or sharing that service or when you think, oh, well, my substableness. description revenue is actually going to fund this thing, it creates huge questions to be answered. We have a few different questions in. I'm a big fan of perplexity.

So perplexity is another one of the AI engine. So this is from Brie O.B. I'm a big fan of perplexity, but I fear I've not even scratched the surface yet. I think most people haven't scratched the surface of what these AI tools can do. I mean, we're still only three years into this game. And this is going to be a decades-long horizon with this technology. This is a long horizon technology. So we are at the very beginning of our journey with this.

And you know what that makes me think about? And we're going to pivot away from geopolitics and AI in the war in a second. But the final thing that I want to say about this, where we are in this moment with generative AI and the impact that it's already had and the situation between Anthropic and the Pentagon, this is at a point in time in AI's journey where the technology still isn't, it's amazing, but then at the same time,

it can still be catastrophic. It can make ridiculous errors and make a lot of things up and completely lose its sense. Imagine when the technology actually always works as planned. Imagine if and when we do get to that scenario of, let's say, artificial general intelligence or whatever it is that we want to call it,

where AI is more capable than the average person at everything is far superior than most people when it comes to intellectual domains as we currently see them at everything. And what happens then? What are the pressures private companies will face in these types of moments?

Imagine when AI actually lives up to the hype and it's not making these mistakes and the temptation to use this technology in war and how that could get, not even just in times of war, even in times of peace to continue to front that that kind of threatening posture, what could be possible and what that power is going to motivate or incentivize people to do. I mean, we're so early on. And that's why I think it is good that people are asserting themselves in this moment and kind of raising their voice and even anthropic to say, we have to have a conversation about this. Because this technology is moving

much faster than I think we realize, even though it's making a lot of mistakes, what it can do today versus what it could do three years ago is quite stunning. And what it's going to be able to do in three years from now is going to be more stunning. So imagine what it could tempt bad actors or actors that feel like they don't exercise power to do. Okay. AGI.I. in 2028, according to Elon. So this is mastermind. And if you're not familiar with what AGI is, artificial general intelligence,

this is kind of like this invisible line that nobody can actually see, nobody can define, nobody knows if it's actually like northeast or northwest, but everyone's moving towards it somehow, backwards and blindfolded. But hypothetically, if we reach what is artificial general intelligence, it's AI systems that are at,

equal to or smarter than humans at virtually everything that we would do in terms of intellectual activities or tasks. And that is technically the holy grail that most AI labs are moving towards, particularly the American ones. Some people say that we'll reach it in the next couple years. I tend to, I do like Demis Asobis is the CEO of Google Deep Mind. And he seems a bit more rational with his AI forecasts, and he puts AGI at around 2030 or onward. And I would say that that's fair. But what we also don't know is we can't predict a breakthrough. And one of our 2026 predictions for IGQ was that there will be an architectural breakthrough at some point on the horizon that changes the course of artificial intelligence development. The same way there is the

deep learning breakthrough, and then there was what we consider now generative AI, but the transformer breakthrough, the architecture breakthrough with the paper about six years ago now, that there will be a new breakthrough that pushes the frontier of artificial intelligence. And we actually talked about this in an episode on AI in biology, quite stunning. You know, AI and silicone isn't where the story stops. Scientists are training artificial intelligence systems on neural cells. from actual human brain cells. And so AI is going to start to learn in a way that actually mimics the human brain, not just in a way that is aspirationally towards the human brain.

And that's going to accelerate AI. And I'll let you see that episode because the scientist, the brain scientist, is quite convinced that that leads us closer to the world of consciousness and all sorts of other things. But yes, it's possible that there are different breakthroughs and artificial intelligence that then accelerate how fast we get to the next frontier of AI. And because I don't think generative AI is where it stops at all. So this was topic one that we wanted to get through AI Anthropic versus the Pentagon.

The second area that I wanted to discuss, and I'm going to go through actually some of the comments, and then we can jump into, I don't know how many of you read that report, the great intelligence crisis about, it was a fictional scenario about what AI will be able to do and what that means for jobs. But I'm going to go through some of these questions here.

You should never assume that AI will never be able to do anything. Yeah, that is a comment that I do say quite frequently. Every time we say that AI will never be able to do something, draw, sing, write. We end up being wrong about that prediction. So my philosophy is it's better to think what are the things AI shouldn't do than to just randomly assume it will never be able to do that thing. Because there are two different things. One category is this is something that we actually feel like is offside for an AI. The other is, oh, we don't even need to worry about it because AI won't be able to do it. And the next thing,

you go to bed and you wake up the next morning and it's live to air and AI has never been better at it. So yes, I have retired the same. You know, don't worry, I can't do that because it's just we always end up on the wrong side of history on that one. I'm a teacher at university. How can I help my students? I ask them to use AI and to teach me. I'm asking them to use AI. I think that's a good place to start. I do do a lot of work with teachers and universities and we're in a really tricky moment because we are literally building the plane and flying it at the same time. AI is totally here. So it makes no sense for a school or a class to pretend it's not and then assign homework as if AI just hasn't been invented yet. So my

My advice in the short run is to assume that everything that happens after 3 p.m. with your students is going to be powered by a supercomputer. And so you have to assign assignments and homework to accommodate that, to account for that. Because if we deny the fact that AI exists at home and on people's laptops and on their smartphones, we do end up short-circuiting the thinking when we ask people just to not use it. Because then they do, and they're going to offload. I mean, we are humans. We are biologically wired to take the path of least resistance.

to expel the least amount of energy. So people are going to go home and offload their essay to AI. So instead you have to design homework or assignment in a way that either the deep thinking happens in class or whatever does get assigned at home. It doesn't matter if an AI system, it's irrelevant that AI is involved because of the nature of that assignment. And so then that deep thinking, maybe orality, debate, that sort of thing happens in the classroom.

That's what happens in real time. But part two of that is the curriculum redesign, which is kind of like the broader systems change that needs to happen. And this shouldn't just be falling on, this shouldn't fall on teachers. I mean, they have so much on their plate. But this is the curriculum. So this is like ministries of education, department of education. This is that level, zooming out. And whoever is designing curricula at various universities and seeing, you know, where are we in this moment?

Right. Like, where are we actually in this moment? where is our curriculum in this moment and what future or past is it preparing people for? And the challenge that a lot of teachers will come up against or curriculum designers will come up against is, well, I don't know what the future is going to be. AI keeps changing every day. And then there's also this quantum thing over here and everyone's going to space over there. So I actually don't know what to teach.

And that is actually an answer, right? So if we don't know where the future is going, and all we know is that it's going to be underscored by artificial intelligence and quantum and synthetic biology and space and all of these other things, is that we have to prepare kids for anything, right? So that's a very different type of curricula, where it doesn't matter which way the future goes because you've equipped kids with the foundational skills that they're going to need to thrive, deep thinking, problem solving, strong judgment skills. So if the curriculum's designed in that way, it actually doesn't really matter what happens

in the future because kids. can thrive in it. And that's going to take a longer adjustment. Universities and schools are starting to work on it, and I know because I'm in some of those rooms. But for the meantime, it is adjusting the kind of short-term shift is what happens in the home and what happens in the classroom. That has to change to prevent the circuiting of thinking. Ed tech technologists here, in your opinion, which company will win the AI arms race in education? Oh, in education. I've never heard it worded AI arms race in education. That's really interesting.

I didn't know that it was an arms race in education. I'm not actually sure. It's actually, I'm not entirely even sure that the way AI is being implemented into schools today is the best way or going to be the way that it stays. Because here's what tends to happen. We tend to look at the best way. future as just this extension of the present. And so all we do is we take what we're doing in the present and say, okay, how can we slot AI in? And that's not actually how the future works, because the entire system changes. It's like not changing our roads that were once designed

for horses. We had to rebuild them for cars. So what tends to be happening in education right now is slotting AI on top of a system that was designed in the Industrial Revolution. And that's only going to get everybody so far. It's going to have to be that system-wide redesign to account for the fact that the world is changing and what education, how and when and what education was built for, that world is come and gone. So it has to actually look different. So that's a long way of saying which company is going to, which AI company will win the ed tech race. I don't even know how well the ed tech

what you know how well the ed tech race is going to be honest I think that that whole thing is going to have to look quite different yeah I mean if because if I was to say most kids in school today are going to work in 64 industries in their lifetime and they're going to work on average 20 hours a week and 84% of them are going to work in space is that is what we have today sufficient for that much of a non-linear shift.

Because that's what the future is. It's a non-linear shift, right? Going from agriculture to an assembly line is completely non-linear, right? The skills would have been entirely irrelevant from the agricultural age to the assembly line. Like, that was an entirely non-linear shift. That's what's about to happen again. We're in another industrial revolution. So so much has to change outside of just putting ed tech on top of what, what,

exist today. Okay, so the Great Intelligence Report that I mentioned. Okay, so yeah, let's talk about the Great Intelligence Report because this will, this matters for, it could matter for a lot of us or maybe it doesn't. The Great Intelligence Report, it was essentially a blog post released on a sub sub-sadfac a few weeks ago. And it sent markets spiraling. So it was a fictional scenario about in 2028, AI gets to a point where it is quite capable. So companies start to deploy it more and more in the workforce and start to actually lay off a lot more white-collar workers.

And that's because most AI is, and this is kind of the reality of where the tools are today, they're much more equipped to come after white-collar jobs. And I say that even with my own job, jobs that are on a computer. jobs that are considered more knowledge workers. That's what these AI tools are currently capable of. That's the type of work that they're starting to learn how to do, things that happen on a computer. So how the fictional story and the great intelligence crisis goes

is that companies start to be able to replace some workers with artificial intelligence, particularly in software companies, especially in things like code, in roles like coding, because that is one area AI is pretty good at, but other areas in software companies or in companies, they specifically called out companies like Visa and DoorDash. And so as these companies

start to deploy more and more AI, they start to fire more people. Now, the economy looks like it's thriving because GDP is going up because the national account looks really good because these companies are getting more and more and more productive. But on the other side, they're firing a lot of people. So the economy is,

actually not functioning as it should because people can't afford to buy things. So the more people that get, the more people that lose their jobs, that changes their consumption patterns. So they don't buy as much goods and services as they did because they no longer are employed. That spiral starts to cascade. So the more firms deploy artificial intelligence and the more people they fire, the weaker consumption gets in the market because the more unemployed people, the fewer things that they can buy. And then you get what they call in the report, this ghost GDP. So the economy looks really good from one vantage point, but there's nobody actually in it buying anything, because unemployment goes, I think, up to 10% in the report. And eventually it just gets to this

inflection point, and it collapses, and it crumbles. I mean, it knocked IBM stock double digits this report. And it also comes from the idea that AI can do a lot of what software companies currently do. So moving towards a future where instead of me saying, okay, I want to, and this would be from the perspective of the report, I need to go on DoorDash and I need to pull out my order and pick the restaurant. I can just soon tell an AI to design me that app if I, that's what I need in real time. Everyone can just get software on demand. Nobody really needs to stream from a software company or software as a service company. That business model starts to make less and less sense because those companies, they're essentially, they store a lot of data and then kind of market

that back out, but they don't, the services they provide in terms of software, AI is increasingly getting better at doing. AI can create software on demand for an audience of one. And I guess it's part two of that point is as unemployment starts to rise and people start to buy less things, companies have to use more artificial intelligence to cut costs. And then that's how the whole things spirals even more. So that's not a new thought experiment. Economists have been thinking about that with AI for close to actually a decade. I was at an economics conference two years ago in California and this scenario was modeled. What happens if AI starts to get better and better so companies are able to replace more people with it? But then those people stop buying goods and

services and so who's even in the economy? The challenge with that perspective is that as soon as that everything else about the economy and markets and people and policy stay fixed. And the only thing that changes is this one variable. AI replaces people. It doesn't account for the fact that policy can be flexible. Workers can have bargaining rights. Consumers change buying patterns. Changes in surplus revenue can spur new types.

of innovations and new types of products and services. So that part, me, markets are really complex. They're this emergent organism that moves in very nonlinear ways. And they treated it all as linear. We're also missing the fact that even if AI, let's say AI can do a lot of jobs that we would consider to be knowledge work today. And that's not actually entirely hypothetical. As AI starts to improve, some of the tasks, more and more of the tasks that we see in the workforce today, AI will increasingly probably be able to take over.

What that timeline is, and we can get to that in a second. But markets and jobs and work and innovation and goods and services aren't static. We start to produce new things because our needs and desires change because we're in a different. economy. So when we went from, it's like going from the agricultural era to the industrial revolution and then everything just stopping. No, we continued to innovate and now we have the knowledge economy. So the real question is, what does the economy become when intelligence as an input goes to zero in terms of cost? What does the economy, how does the economy reconfigure when intelligence as an input is abundant? What is the shape of? What is the shape of

that economy. And it will be something. It could be very unrecognizable. It's going to reconfigure. It could be reconfigure in very strange ways. But it will reconfigure itself the same way it would have been impossible to imagine the idea of a social media manager during the Industrial Revolution. Like, oh, well, don't worry. Things are going to eventually continue to change, and eventually people will be swiping and managing influencers on Instagram. I mean, it's just impossible to predict the nonlinear ways markets reconfigure themselves as technology continues to evolve, automate, and push the frontier, and expand what becomes possible, but new things will become possible. So I think that that's what that analysis misses. It doesn't treat intelligence as

as an abundant input what the economy reconfigures when that becomes the future. What the report, though I think does get, does point, this is call attention to that I think is important. Is the speed at which this change and this unraveling and this replacement could happen, it is possible that jobs disappear faster than new ones get created. So the Industrial Revolution, it was a nightmare, but they had, they also had 90 years. Now I'm looking back on the Industrial Revolution, I'm like, guys, you had 90 years to figure this out. why was it so chaotic? We have about, you know, nine minutes. But truly, imagine that change, right? About 75 to 80% of people did something different over the course of 100 years. What happens

if that happens in 10 years? So not an industrial scale change, but over a decade. So we eventually do new things and we reconfigure work in the economy in entirely unrecognizable ways. Great, we have been there before. But if we, if that compression is 10 years, that's what I'm actually concerned about. The disappearing faster than the economy can reconfigure itself. And that's the question that I think we need to hear answers from from policymakers. And if you've been listening to me on IGQ or if you've been following me on substack, you've heard me call this out over and over again. We've heard enough about AI, sovereign AI. We're going to bring it all home. We've heard you amazing. What is the plan for the people in the economy?

me whose jobs AI may automate. And I know AI and automation doesn't make for the sexiest campaign pitch. Like, oh, I'm the person that's going to talk to you when you have no job because of artificial, like obviously that's not the best slogan for somebody's campaign, but someone somewhere at some point is going to have to talk about it because that transition is coming and we can see it. And one of the things that I also bring up quite frequently is even something like insurance, if your health insurance is tied to your job, we are moving towards a more independent contractor-based future. Because as AI continues to learn new tricks over time, and that requires different skill sets to pair with it, company is going to be less likely to

hire full-time rules because that rule is going to look, who knows what it's going to look like in a year. So we want to see the rise of more contract work. Great, for the contractors. What does that mean for health insurance, right? And dental insurance and social security and all these things that we can see. And that's going to be, that's happening like time now. And we don't hear much from policymakers about that. So I think the report calling attention to the lack of safety nets we have, we are starting to see some policymakers talk a little bit more about jobs and transition and skills, but it's time to really get in the game with that conversation because that part could unravel over the next couple of years faster than it seems like anybody is preparing for. So that is

the great intelligence crisis report from the Centrini. Sintrini is a research organization, I think, the research finance analytics company. So they published that on SubSacker a couple weeks ago, and it just sent markets into a spiral. But that also shows the strange, probably unhealthy feedback loops between social media and markets. Did it warrant that much of a sell-off because of a fictional blog post in 2028? Really well-written. A great thought experiment.

I hope some policymakers woke up and smelled. the AI potential automation coffee, but did it warrant that sell-off? No, I think some of that has to do with social media and things trending. And I think we're going to have to think beyond that if we're going to get through anything in the next couple years. Okay, time for some questions from Jordan. Health insurance isn't really U.S. strong suit. Yeah, that's okay, that's another thing. And I would actually rather just talk about it. Health insurance, and I know it's even, it's a privilege for many countries who have it. And I know some countries are struggling with just a, with basic GDP.

And so the idea of free health care is just that people are just trying to survive. So I totally understand that. But in economies where health care could be an option, they're going to have to have that conversation. Because let's say 2028, we'll use a different area just because we all have PTSD from that report, 2029. And let's say 67% of the workforce is independent workers and contract workers in the U.S. So no health insurance that's tied to an employer. What would be the plan? And it actually would be really tricky to say, okay, everybody self-employed, get your health insurance together in four years. The infrastructure just isn't there to support that. So countries that aren't offering or even

considering free health care, they're going to actually be forced into that conversation. as a social safety net. And this is where AI can come into play. This isn't about people always, oh my gosh, running into the socialism field. That's not what this is actually about. AI companies, if they're going to be winning economically in this future, based off of our data,

we should all get a ride on that elevator up, right? So everybody should get a part in AI's success story because it couldn't have been made without all of us and our data. So that's where different taxation models start to come in and no UBI can suck out of all the air out of the room because then we think, oh my gosh, I'm going to be floating in my pool and get a check from an open AI. That sounds like a nightmare. But that's not even what I'm talking about. I'm talking about just at a bare minimum. There should be changing tax structures to ensure that we're all getting a ride on this AI story because we're all a part of it.

And something like that could support a new health care model. So it doesn't have to be this kind of like physics-bending equation. It just can be quite basic, but we can see that coming. So something we're going to have to think about. For the future of work and AI, what are your top five resources to stay informed? Well, of course, I've got questions. But if I was to say, what do I use to stay informed about it? So I read a lot of research reports, a lot of academic papers.

I'm following a lot of patents and acquisitions. So my data and research diet is a little bit different. Bloomberg, I think, is actually pretty great, but they're kind of expensive. But Bloomberg's a pretty good, has a good eye on all of this stuff. I do like MIT Technology Review. They don't necessarily get into the future of work, though, but they do get into artificial intelligence. They track it quite well. The Atlantic has some good stories.

Oh, it's tricky, right, because it's all so disaggregated. What else do I follow? I do actually like substack. There's so many people that I follow on substack are so much content that I see on substack about AI that's really helpful. So I think if, yeah, if you can even just get on substack and start just searching

some of the AI content, I think you'll also slot there. And of course, we'll continue to keep you in the loop on AI. And I actually write a lot about this stuff on substack too. Yes, so we have a comment here. This reminds me of an article I read recently about the 1973 MIT computer's prediction

of the end of civilization, right? How did Y2K go? Did everyone survive? Are we still looking for people? I know. It's funny, right? We have these forecasts about it being the end. Even the idea.

And this might be a hot take. I don't think it is. But a lot of people in the AI industry think that AI is going to be humanity's last invention. Because we are inventing something that also invents. And so it may be humanity's last invention. that's not with an AI. But I think we will, AI will just be a big input

to the things that we invent going forward. Because AI, it doesn't have desires, right? This is not a system that's alive that has wants, needs, in the ways that, you know, humans do. They have goals set by, that's a whole other thing. But sure, they have programmed goals and they get reward if they're doing reinforcement learning and that sort of thing.

But they don't have actual desires. and feel competition and scarcity and all of the things that be humans are motivated by for better or for worse. So we're still going to continue to want and to expand and to compete and to win. And that's going to continue to push the frontier forward. So this may be humanity's last invention without AI, but I think we're just going to step into a new era where it's AI-powered invention, which is an entirely new thing. AI will also invent on its own.

it will probably spot gaps in data and come up with an idea for something that nobody had suggested. And it just kind of, you know, drop something in a suggestion box that nobody had thought about. That will also happen. But I don't know. I don't think that this is our last invention. Everything was the last of everything. And here we are now still today, literally live streaming. Do you recommend any books or thought leaders to follow?

It's hard to know where, who to look to. Oh, books, I have a huge library of books. It depends on where you're getting started. Let me have a look. Okay, so if you want to actually a great book and a great person to follow on the future of work, and thank you for this question because it reminded me. Ethan Mollick, and I have his book here called Co-intelligence.

Ethan Mollick, he posts on LinkedIn or on Twitter. And I think he's also on threads. So wherever you are, he posts there. and he's great. He constantly talks about where the frontier of artificial intelligence is. So this is what it can do today. It can do this in radiology. It can't do this in law. So he's actually a great person to track. And he posts a lot of different studies about artificial intelligence. So if somebody said one person, I mean, that's a lot of pressure. But I would say he would be a great person to follow. And you could also read his book, Co-Intelligence. I do love Joy Boulinini, Dr. Joy Boulamweeney, her book on masking AI. If you want to understand how AI impact. are our rights, our privacy, the ethical implications of this technology. Her book on Masking AI was amazing. If you want to understand where could this possibly go? How advanced could this technology really get?

And what would that mean for society? You could read Genesis. This is by Eric Schmidt, Henry Kissinger, and I think it's Craig Mundy. So that would be another great book recommendation if you're trying to think, okay, how far could this technology get? And just to kind of put in context the types of ideas. So Eric Schmidt, he was the former CEO of Google, and he goes viral all the time for his different takes on AI. And he's right in, he's in the lab rooms. He's in the U.S. security rooms.

He's at the frontier where all these developments are happening. And so some of the ideas are, you know, in two years, all of us would be able to say to an AI, system, make me a competitor to TikTok. And if it's not beating TikTok in two hours, let's rebuild. And you have an AI system that can build a TikTok competitor like that. And that's going from voice to software. So from a prompt to software. And that's the pipeline that we're on. And that's actually going to be possible, truly. That's actually not a fictional scenario. We will get to a point where the cost of software generation falls and collapses to near zero, close to just generated a website. Oh, I built an app. But here's what people miss like a part

two. Because everyone's like, oh my gosh, five code. I have 500 apps. But when everyone can build a TikTok, what is the business strategy, the marketing strategy and the competitive advantage that you have implemented that brings people to your app over the other competitor, over the 500 others. So all that intelligence does is raise the floor. So more people now are in the game, but the borrower is actually more challenging to compete. So that is an entirely different type of skill, right? So then that's much more strategy thinking. That's much more judgment. That's understanding the fundamentals of business dynamics. So it gets more interesting. So it's a lot more complicated. There's always a part two to some of the grand ideas. But those would be some of the

books and places I would start. Someone said they're currently reading on masking AI. Great. I love to hear it. Create a list of Amazon affiliate-linked books on AI that you would recommend. Would people appreciate if I did that? I guess I could do that. I could have considered myself an organizer on Amazon, but that could actually be very

helpful for folks. I'm trying to think, are there any other books in here that I really enjoyed? If you want a great cybersecurity book, not to go back to the wars with AI, but this is how they tell me the world ends. It's an amazing cybersecurity book. And it takes you right into how much of our infrastructure is actually cyber infrastructure and how that underpins everything about the modern world and how vulnerable it actually is. and the world of spies and hacking. It talks about geopolitics. So if somebody's really interested in cybersecurity

and how that underpins our modern world, this is how they tell me the world ends. It's fascinating. I'm trying to think if there's an optimistic take on the future. I have so many. Mustafa Sullyman's book, The Coming Wave, is also great. He talks about what we have to try to do, what we have to regulate, how capable AI is going to get,

what we can do, what we can't do. trying to think what else. Yeah, those I think would probably be my book recommendations for now. Okay, I'm going to get through some more questions because we've been chatting for an hour. And thank you all so much for coming to our first live. We hit over 100 people today, which has been amazing. I really wasn't sure how many people would show up because we did it quite last minute and I've never done live on YouTube.

And I was also hoping it would actually work. It's so funny, I'll do sometimes, you know, presentations on neural interfaces, and then the internet doesn't work. And it's like, if you'll excuse me for a second, I'll just call R2D2 and get the internet powered up. And just to also leave you with, you know, so what should you be doing in this moment, right? So we talked about Anthropic at the beginning.

We talked about the great, you know, intelligence crisis. What does all of this mean for you? And I would say, with respect to anthropic and, the Pentagon. It's not like everyone get in your car and drive to the Pentagon or drive to Anthropic and write your letter. That's obviously super not really helpful.

To continue to read if you can and to continue to see how these dynamics play out. Because it does actually help how we vote. It does actually help which products we choose to use and not use.

You know, our buying power is actually a pretty important one. And especially if live in America, it seems like, you know, the stock market, even though Anthropic is a private company, but markets in general, money in general, seems to be quite a big incentive. So if you can continue to just track some of this a little bit if you have the bandwidth, I know that there's so much going on. But it does help to, it does help even in the background of how you start to make decisions about this technology. Because this technology is a really big deal, right? It's going to impact all of us. We all deserve to have our voices heard.

heard in it. This is a foundational technology. All of our data went in to train it. So we should all get a say. And it's really hard to insert your voice if you don't necessarily know what's happening. So this is one of the stories that I would say just keep an eye on it if you can and start to see how it plays out. That would be my advice for Anthropic. With the great intelligence crisis, I mean, we kind of talked about there will be more coming into the economy. It just hasn't really been invented yet. But what do we do? And this is something we talk about a lot on IGQ. We're cultivating our skills now, right? Not not waiting to see how fast AI develops, what it does next, the next company announcement. Things are changing.

In some ways faster than people had assumed and in some ways slower and that kind of jagged edge is going to continue to go. Build your judgment skills. Start using this technology. It's not going away. It's going to be as pervasive as the internet. In fact, in the next year or two, it will be eventually just like a computer. When you get to the job, it's assumed you can do it. So start using this technology. Start getting good at giving it orders and evaluating what it tells you.

Is that actually good or is it kind of nonsense but sounded really intellectual? So start using this technology. Start to improve your communication skills. Start to read more. So you can start to strengthen your critical thinking. That's one thing that I've been doing a lot, actually. I mean, I do have to read a lot for work, but even I've been doing even more reading. I've been doing a lot of writing.

I'm going to show you even some of my, I write a lot. When I go to meetings, I bring a notebook and I am drawing, these are like business models, literally, and I am drawing them out and seeing how different patterns connect and really trying to exercise those critical thinking muscles. So I'm continuing to read deeply, if you can, continuing to play with artificial intelligence, use these tools, even if it's just making recipes and just what's a part of your daily workflow in some way, really important. Exercising good judgment skills. Breaks from technology actually help because I think we know that, especially social media

and what it's doing to some of our attention spans and so forth. So all of these are kind of like the soft power skills that are going to help you in a world with AI. And we actually have, and we can link it once we repost this episode, an episode on the seven skills for the AI age. And I go into each one very deeply. And so all in human skills, emotional intelligence, all that appreciates and values. So continuing to exercise these skills is really important.

And even, you know, even small things. Sometimes when I'm tempted to send a text, I actually pick up the phone. To continue to practice that emotional intelligence skill, a skill of compassion, of reading people, of understanding people. Because when you think about it, and here we go again, next spiral, when you think about it, if AI continues to improve at the rate it does, and I've just noticed chocolate, so whoever put that there sabotage for me, if AI continues to improve at the rate that it does.

And I have the choice between a decent AI system or a person with similar skills, but they're funny and they are emotive or they build the company culture or team culture or they're just fun and enjoyable to work with or just kind. I'll probably choose that person, right, over an AI that's maybe not as good. but cheaper. I'm going to go with the person. So it's going to start to matter in the workforce, right? How you show up around other people. So those are other things that we can practice every day. And you can literally practice that at the grocery store. We don't need to go get a degree for it. So little things like that. And I'll finish up with this one question that came in. Yes, human contact will

become the why. Human contact is going to become really interesting. There's probably going to be an entire new economy coming around that or just innovations around that. Thoughts on the future of creators? Will more individual creators be the owners of their IP, movie, studio, record label, etc. At scale like Disney, Marvel, Pokemon. Right. So you're asking, does the creator of the future become the enterprise? And this is really interesting because in some ways, yeah, if you think about what AI means as a technology,

you're going to be able to stream intelligence and stream the capability, some of the capabilities that some of these massive enterprise prizes, these massive studios have. In the coming years, AI stems will be able to replicate stem aspects of that, not all. And not, I mean, because creativity is much more than the technical execution skill. The idea behind that script to begin with, I mean, AI could write an amazing sounding script, but it actually doesn't make any sense or it doesn't actually move people in the same way, but it can write well. Those are two different things. But sure, there is a future world where the individual is just as leveraged as the modern day studio. But I actually don't think that's entirely where we go because I think we're going to be

doing different things in the future with this technology. So in theory, what you're asking is if everything stayed the same and we continue on this trajectory, and it will be for the next couple years, will we start to see creators, start to go head to head with some of these entertainment studios. One, because of the nature of social and people owning their own audiences and having much more engaged communities and two, because they're going to be much more leverage. They have a following. Then they have AI and they have all of these different tools and capabilities. theoretically, yes, they could start to compete with movie studios or entertainment studios and not because the content is better, but because the market wants different

things. So what the creator offers in combination with all these new superpowers is maybe where the market that fits the needs at the market at that time. I think we're going to start to do something new and there's going to be a new crop of creators that come up that do something kind of AI first that's going to be really interesting. But yes, I think I mean even the entertainment studios, Hollywood has been impacted and traditional entertainment has been impacted by social media, by YouTube, by streaming. So it's technological innovation has already started to impact the way traditional studios had set up and how those business models were structured.

So that's probably going to continue to go to continue to accelerate that trend. But yeah, creators will probably be highly leveraged over the next couple years until something new happens and something new comes along, right? Because the future it's nonlinear. So things are coming that are hard for us to see. Okay. So this was really fun. Thank you so much, guys, for coming out. I love this community so much. Your questions and I read your comments on the podcast on YouTube. I try to respond to as many of them as I can. So thank you so much for your support. And we will see you at the next live.