Top AI Researcher Reveals The Scary Future Of Employment

So this video is rather important because it discusses the future of employment. And it's not just a random person. This is from Avital Balwit. And this is someone who actually works at the Frontier Lab. Anthropic. And her insights into how the future of work and employment is going to change are very important for individuals like yourself and of course, me. So I'm going to do a deep dive on this entire thing because I truly believe that this is really important that you can stay ahead of the future. Now, if you enjoy videos that are specifically focusing on how you can use AI to better your life and of course, preparing for AGI, such as how to invest in AI and benefit from the trillions of dollars that are flowing into that industry, don't forget to check out my post AGI preparedness group. This is where videos like this actually get uploaded a few days earlier. But without wasting any more time, let's get straight back to the video because it's really important. So essentially, this is a post about my last five years of work. Now, if you aren't understanding the significance of this, trust me, at least watch the first 10 minutes because. Because it's truly, truly incredible. So essentially what we have here, ladies and gentlemen, is a post by Avital Balowit. And this is Anthropic's chief of staff. Okay? And this is kind of like a personal reflection, SL diary on how the next few years of, you know, the economy are going to go for the average person in regards to work and where they fit into the economy and how things are going to be working. Now, the reason I'm covering this is, is because I think this is probably one of the most remarkable insights that we are lucky to have to get because a lot of these, you know, people at Frontier Labs, they truly don't talk about how the future of work is going to be. So it is something that I think is very, very important for us to discuss because this is something that is truly, truly pivotal, especially for this monumental period of time. So there are four main things. I'm going to timestamp them in the below, but just take a look at this. So she says, I am 25, okay? And these next three years might be the last few years that I work. Okay? Remember, this is Anthropic's chief of staff, and she says that the next three years might be the last few years that I actually work, which is, number one, an incredible statement. But number two, it should show you the gravity of what we are truly dealing with here. It says I am not ill, nor am I becoming a stay at home mom, nor am I, nor have I been so financially fortunate to be on the brink of voluntary retirement. I stand at the age of a technological development that seems likely, should it arrive, to end employment as I know it. So this is not some kind of clickbait title where I'm trying to draw you in. This is the exact words from this person who is working actively at Anthropic. Now, Anthropic is one of the leading AI labs that made Claude 3, which is of course Claude Opus, which is one of the frontier models that is performing very effectively. Now what we can see here is something absolutely incredible. We can see them stating clearly that if it does arrive okay, we're standing at the edge of this, which seems likely that it's going to end employment. So you know how a lot of the time individuals in certain communities, they may discuss labor and work as something that of course is, you know, going to be there for many individuals like you and myself. But the prevailing trend from many of the industry leaders that I continue to see is the fact that there won't be any true work after a certain period of time due to the fact that we have a situation on our hands where the technological developments that are coming are most likely to end this paradigm. Hence the reason, post AGI preparedness so that you can navigate this now. Now, basically, she then states, I work at a frontier AI company. With every iteration of our model, I am confronted with something more capable and more general than before. That's just absolutely crazy. With every iteration of our model, I'm confronted with something more capable and more general. Before. At this stage, it can completely generate content on a wide range of topics. It can summarize and analyze text passably well. And as someone who once made money at one point in time as a freelance writer and prided myself on my ability to write large amounts of content quickly, a skill which, like cutting blocks of ice from a frozen pond, is arguably obsolete. I find it hard to not notice these advances. Okay, remember that one? I find it hard to not notice these advances. Freelance writing was always an oversubscribed skill set. And the introduction of language models has further intensified the competition. Partitioned. Now, remember, the whole point of these videos is to ensure that you are paying attention because a lot of people just say, ah, you know, we'll find new jobs, we'll be fine, yada, yada, yada, yada. And I think that's very, very ignorant because it's quite hard to ignore the kind of developments that are going, and eventually they will be, you know, impacting nearly every industry on the planet. So it says the general reaction, okay? And this is where, you know, this is exactly what I say, the general reaction to language models among knowledge workers is one of denial, okay? One of denial, okay? And this is why I'm making this video, because I want you all to understand that whilst, yes, in previous technological revolutions, like when we had the industrial revolution of the farming industry and many people were thinking, oh no, all of these computers are going to take our jobs and all of us that work in the farm industry come. The problem is, is that we're of course going to be having a situation where no longer are we going to be needed in the farming industry, so we might lose our place in society. But of course, in that stage we did have a lot more things to do. But the reason that this time is different, okay? And you shouldn't be in denial, okay, especially among knowledge workers, is that AI is the automation of automation. So it's a feedback loop because AI is something that can automate itself, which is different than previous generations. Remember that this time it's different. So it says they grasp at the ever diminishing number of places where such models still struggle rather than noticing the ever growing range of tasks where they still have reached or passed human level. This is something that I truly, truly think people need to understand. It's about how you view the AI problem. Okay? Many people literally just focus on the fact that LLMs hallucinate once or twice or in generative AI images, sometimes they make mistakes, but you kind of look at things and you're like, wait, overall, okay, like if we take the entire pie of AI and you into account, overall, there's like a small subsection that with every iteration is growing and growing and growing and growing. Whereas like most people, like, oh no, it can't do this small bit. Oh, it can still not do this small bit. Oh, it can still not do this small bit. And eventually what's going to happen is that it's all going to be able to be done by AI, which means that like, if you're at this stage right now where you can see that, okay, AI is slowly starting to surpass human cognitive capabilities, it's best to, at this point to be like, okay, I'm going to pay attention here and at least position myself rather than being, as she states, in a place of denial. So she says. Many will point out that AI systems are not yet writing award winning books, let alone Patenting inventions. But, and this is the most important thing, most of us also don't do these things, okay? And this is what I'm stating that a lot of times people like, oh, AI can't write an entire novel from scratch. And AI can't, you know, paint this and it can't make this poster, you know, coherently with the text in five seconds. Like I could. But you couldn't do that. Most people couldn't do that. AI is able to do a lot more than the average person, which is the point, okay? Most people don't do these things. And this is something that most people are failing to acknowledge, okay? The economically and politically relevant comparison on most tasks is not whether the language model is better than the best human, it's whether they are better than the human that would otherwise do the task. And this is the most important distinction. I'm going to say it again. The economically and politically relevant comparison on most tasks is not whether the language model is better than the best human, it's whether they are better than the human that would otherwise do the tasks. Basically, this question is the best question, okay? Because it shows you at what point things get automated. And this is what I've spoken about before. Imagine this, for example, right? You have someone who works in a store, okay, let's say that person previously they did, you know, checkout at a grocery store, okay, Right now we have AI self checkouts or whatever. We've got all these kind of crazy things going on. Of course the self checkout is just a piece of technology. But the point is, guys, is that, is it sufficient to have an AI system be better than the average person that would do the task? In any scenario where that is true, you are going to get automation of that labor. It doesn't need to be better than the best person. It just needs to be better than the average task that it would take. And that is the important distinction because a lot of people think, oh, this thing can't create Hollywood blockbuster movies. It's not going to, you know, affect the movie industry, but it will affect other industries like the film industry, the photography industry, all of these other ones. Especially when, you know, we look at future models, how much smarter they're going to be. Just because they're not going to be as smart as Einstein, it doesn't mean they're going to be smarter, not smarter than the average researcher. Which is of course very fascinating. And then it says this makes the objection that AI systems are not yet coding long sequences are doing more than fairly basic Math on their own, a more relevant one. But these systems will continue to improve at all cognitive tasks, okay? And that's an important distinction because a lot of the cognitive labor, you know, it's kind of going away from the trends that we can see, okay? The shared goal of the field of artificial intelligence is to create a system that can do anything. And she says here that I expect us to soon reach it. And if I'm right, how should we think about the coming obolence of work? So now here's where we talk about the future of society. So she says that it's worth noting that upfront that even today, work is far from the only way to participate in society. But nevertheless, it has proven to be the best way to transfer wealth and resources. It provides personal goods like social connection, status and meaning. And of course, it offers social goods like political stability. And here's where we get into the crazy things given this. Should we meet the possibility of its loss with sadness, fear and joy, or hope? The overall economic effects of AGI are difficult to forecast. And. And here I will focus on the question of how people will feel without work, whether they will or can be happy. And there are obviously other vital questions, like how will people be able to meet their material needs? Many have examined this question, and this is the important part here, okay? And this is why I launched a community and why we're working on all these things. Many have examined this question, okay, with no final answer yet, as adopted by official policy for this contingency by any government. So there is not a single government that is currently looking at this problem and thinking, okay, this is going to be an issue in a few years. How are we going to tackle this? And I think that that is important because many people think that, okay, the government is going to be there to take care of me. But I would say the government aren't really babysitters. They just kind of look over you every now and again to see if you're doing all right. And if you're not about to die, the government doesn't really care about you that much. Of course, depending on where you live, this is something that can be argued, of course, but I do think that this is for the majority of societies and the majority of things. It is very, very true. And of course, you can see I will go ahead and assume that people can meet their needs through universal basic income or other transfers and will solely concentrate on the question of whether people can and will be happy, or at least as happy as they are now without work. So here's where we talk about the obsolescence of knowledge work. So you can see I expect AI to get much better than it is today. And research on AI systems has shown that they predictably improve, given better algorithms, more and better quality data, and of course, more computational power. Labs are in the process of further scaling up their clusters, the groupings of computers that the algorithms run on. And machine learning is a young field with an enormous amount of low hanging fruit in terms of discoveries. I meant to say fruit there. And, and this means that researchers can continuously find improvements on the algorithms of these AI systems. So the point here, and I think this is one that it kind of goes back to what Leopold Aschenbrenner said, is that when we take a look at the fact that there is a lot of low hanging fruit, this kind of shifts our opinion. It's not a field that is mature to the point where we are at the edge of everything that we know, or at least what we think we know. And it's very unlikely we're going to get a breakthrough for another 20 to 50 years, like in some fields. This is a field that when we look at the fact that if we know that we scale up another 10 times, we're about to get a whole lot more. We know that if we improve the data quality and the amount of data, we're about to get a whole lot more improvements. We also know that there's a lot of inefficient algorithms that we do use. And we know that there are literally a million different ways to improve these models in terms of the algorithmic efficiencies, in terms of how they respond. So the point here is that there is a lot of low hanging fruit in this industry and we're only at the frontier of what these models are going to be capable of, since we haven't exhausted all of those things. If we had exhausted the compute capabilities, all the algorithmic discoveries, and we were at the wit's end, and these are the kind of systems that we had, then maybe potentially it could be argued that we were near the end in terms of the growth, especially in terms of the curve, maybe the growth would start to slow off like that. So that is why I think this is really important, because there is a lot of low hanging through that could truly give us a lot more exponential gains. And you can see while there is an enormous amount of data that has already been fed through them, there is still more to be found and it can also be generated by these systems themselves. So given the Scaling laws, we can reasonably foresee that these systems will, will keep getting better, at least until these inputs run out. So at what rate will they get better? Language models are not, for the most part continuously improving. They get better in discontinuous jumps. A rough analogy to the current LLM process is that making a new model is like baking a cake. You figure out your data and algorithms, like mixing the batter. Then you pre train the model, that is to run it on a large number of computers for several months that put it in the oven. Then you do some post training, like frosting and decorating the cake. And of course this can adjust the model in certain different ways to make it more harmless, honest, or to make it good at certain specific skills or use cases. But what matters to most of us for the model's capabilities, at least right now, of course, is the underlying cake. And this can't be easily adjusted without starting over and baking something entirely new. So when it comes to the rate of progress, when these models seem to plateau, you should actually assume that that just, just means that the next model is in the oven, but just hasn't come out yet. And basically what they're stating is that whilst, yes, you know, some people do talk about how there is this iterative deployment at OpenAI, from what we can see, growth in LLMs is kind of like this. We hit a staircase, we hit a staircase, and we hit a staircase because essentially what you have is you have, let's say, for example, we have GPT5, which brings us to this level, then we've got GPT4, then we get GPT5. Okay. And the reason it's kind of like that is because like they said, you have this middle bit here where the model is being, you know, trained, whatever, or you know, mixed or whatever you call it, whatever. And of course at the end you've got the post training, and of course here you've got the pre training. So this is essentially where, you know, you're training the model for this entire period of time. So what's interesting is that this period of time where you train the model this entire period of time, do apologize for the colors there, but training the model, and then of course the model gets released after the post training and then that's when we get those new jumps. Which is why it's very hard to say that right now things are plateauing because these models do take several months in most cases to train, especially because the clusters can't support that much work going on at the moment. So here we can See, many expect AI to eventually be able to do every economically useful task, which is a truly insane statement. So many expect AI to do every economically useful task, which is absolutely crazy. And remember, she agrees. Now this is Anthropic's chief of staff here and it says given the current trajectory of technology, I expect AI to first excel at any kind of online work. Essentially anything that a remote worker can do. AI will do better copywriting, tax preparation, customer services and many other tasks are or will soon be heavily automated. I can see the beginnings in areas like software development and contract law. Generally tasks that involved reading, analyzing and synthesizing information and then generating content based on it. It seems ripe for replacement by language models. So basically right here, this gives you a first insight into some of the careers that are under not scrutiny, but under the microscope in terms of where things are going to be truly automated. You might want to note this down. I will have of course, all the notes on in my school community so that you guys can easily read this and digest the information and an easy to use matter. But guys, this is some key, key statements that people are not paying attention to. Because the main thing here is that anything on a computer, anything remote, anything online, those things you have to think now about, okay, if I'm a remote worker, if I'm someone that's enjoying my life, traveling around the world, or that's what I want to do, I need to think about these industries here and how I can position myself, okay, to not be automated away or to not at least be less economically valuable than an AI system. Sure, you might be 10% better than an AI system, but you have to remember that it's much easier for a company to replace your high salary with an AI system that costs much less. And I do think in the future what we will see is AI systems that are truly, truly effective. And of course that can run for 24 hours. Remember, you're just a human. You can't run for 24 hours like these AI systems can. And with all the AI agents being built on the back end, trust me, guys, things are truly bright for disruption. So you need to think, okay, how can I do this? Okay, and how can I manage myself better in order to at least not be in that place where I'm less economically valuable than this thing. And here's where we have some more stuff. So here's where we get to the obsolescence part. And this is where we actually start to look at certain industries and how things are going to be changing. So essentially what she talks about here is that, you know, this is not going to come for every industry at the same pace. It's not like one day everyone's working and next day everyone's not. We're going to have a situation on our hands where this area is now affected. This area is now affected. This area is now affected. And it's going to definitely be proportional to AI system releases. For example, when GPT5 does release, I do expect a wider range of jobs to be not completely automated, but very nearly on that chopping block level to where, you know, there's going to be another wave of potentially layoffs or potentially, you know, companies that used to hire people, not hiring people. And the problem is, is that these effects are downstream, meaning that if someone was going to create a company in the future and they choose not to create a company, you don't experience that level until like maybe 12 to 48 months after, because they. The companies that would have replaced the old ones that wouldn't hire you, those ones no longer exist. So overall, that entire market share has died down. So essentially the pace of improvement in robotics lags significantly behind cognitive automation. It's improving as well, but more slowly. So this is basically what they're saying is that if you do anything with your hands that is pretty intricate in a guided, specific situation, for example, you know, surgeons, gardeners, plumbers, jewelry makers, hairstylists, as well as those who repair ironwork or who make stained glass, might find their handiwork contributing to our society for many more years to come. And this is, of course, you know, very, very true. I do think that, you know, people in these industries, they have a unique set of skills that, you know, for a robot to do as well as a plumber, it's very, very hard considering, you know, the fact that it needs to navigate its way into a house, deal with the customer, be able to identify what's going wrong, stop the water flowing, whatever issue it may be, then coordinate with all of these kind of things. Like if you try to get a team of, you know, engineers to get a robot to do that, that would cost a lot and it would be so expensive to do that. But, you know, an AI system that's on a computer running, that can take control of a computer, that is something that is vastly easier, which is why a lot of the people who work at desks are more at risk than someone who does a manual labor job. And this is something that I think this trend is important to pay attention to. So, of course, regulated industries like medicine or the civil service will have human involvement for longer, but even there, I expect an increasingly number of human workers who are supplemented with AI systems working alongside them. So basically with this one is that if you work in a regulated industry or you're thinking about getting into one, I think those ones are still pretty good for the future because regulatory boards and stuff like that, they do take a lot of time to allow people to get into their industries. You know, for example, like to become a doctor, there's all these tests you have to take to become a lawyer. You got to pass the bar, you got to do this. It's this whole kind of thing, whether you want to call it status. Even if the AI system is better, there's just this human kind of aspect where even if AI systems are better because of the regulatory bodies and how AI might be viewed at that time, AI is probably not going to be penetrating those industries very, very soon. Like it's not going to be, okay, AI's here, boom, let's use it. There's still regulatory bodies that people, you know, follow. And it's going to be really hard to change those because small changes and small issues with AI systems, unless they're like 100% and unless they have no bias, no this, no that, no that, it's going to be really hard to change those legacy industries because they are very, very resistant to change. And they were built upon those, you know, strict foundations. So, you know, careers that have like, you know, all those regulations around them, I think those are going to be secure for a very, very longer time. Now, of course we talk about industries that are going to be preferred for humans. So for example, you know, those jobs that, you know, where you want a human to do it, even if an AI system might be better. So for, for example, jobs that might fall into this category include counselors, doulas, caretakers for the elderly, babysitters, preschool teachers, priests and religious leaders, even sex workers. Much has been made of AI girlfriends. But I still expect that a large percentage of buyers of in person sexual services will have a strong preference for humans. And I think this is an interesting term here, nostalgic jobs. And I think that term actually does make sense. So you can see that these jobs right here, even if you have like an AI system that is, you know, just amazing, you know, sitting a baby, let's say it's great as a babysitter, I would not want an AI system babysitting my child. I would want it to have the human to human connection. I would want it to understand what other humans want. This is just going to be something that maybe as I grow older and as I, you know, get older, into my older years, I'm just like that old person who's, who's scared of robotic technology. And I'm like, no, I want a human looking after my kid or whatever. Or maybe, you know, even people growing up now, they still find it weird. But over time those values can shift as robots become more and more useful and as they become more and more human, like, so it's going to be interesting to see how things are going to be, you know, done in that aspect. But these jobs right here, they're essentially valued because there is a human right there. So if I was going through a problem, I wouldn't really want to talk to an AI system because the whole thing about like counseling and stuff like that and being, you know, talking to a therapist is that there's a human who actually empathizes with you. And no matter how smart an AI system might be, it hasn't had its own, you know, subjective or objective, whatever you want to say, experience of the real world. So it's not going to be able to empathize with you on that way that an actual person that can be like, oh, you know, I, I understand that. I went through that. You know, I've had this experience, I had that happen to me when I was 21. And that like things like that are just completely priceless. So those things there, like those areas and categories are going to really do well, which is why it's important to kind of look at this because a lot of people don't understand truly the dynamic on how things are going to change given the nature of AI going. Now for the rest of this article, there are some interesting things. There is the psychology of employment, which is, as automation rolls out across these industries, how are people going to feel? And basically what they're stating here is that evidence shows, unsurprisingly, in fact, no, this is actually quite surprising is that basically they're trying to look how unemployment actually affects people. Okay. And it says that one study that tried to tackle this looked at the effects of unemployment caused by the collapse of the Spanish construction industry on mental and physical health. And this particular study was attempting to disentangle the causality because people who lose their job during a nationwide collapse of an industry will avoid the selection effect. These individuals are no more likely to have mental or physical issues than other members of the population. And it says by looking at large scale survey responses before and after the crisis, they found that unemployment did appear to increase the likelihood of reporting poorer health. And it says here it does seem that overall unemployment makes people sadder, sicker, and more anxious. But it isn't clear if this is an inherent fact of unemployment or a contingent one, okay? Because basically the problem is this, right? If you, right now you're unemployed, okay? Why are you sadder? Think about that for a second, okay? Why are you sadder? You're sadder because, number one, you might have less status because you don't have a job. And number two, you know, you don't have any money, okay? You don't have any money coming in. So you might think, okay, I'm not bringing any value into society and I'm not making a living for myself. And of course, I might be homeless and yada, yada, yada. And you can see right here, you've got all these things that they talk about, okay? You know, the financial effects are probably the main one. Like if I gave you right now $10 million and I said, you can never work a day in your life, you know, you're never allowed to have a job even if you wanted one. You just are not allowed to work. Most people would laugh, run off with a 10 million, and be like, this guy doesn't know what he's talking about. And that's the main point here, is that it's hard to understand whether or not humans derive meaning from employment or it's just the contingent effects, okay? And you can see right here is that of course this might not occur in the context of universal basic income. And of course it's confounded with the shame aspect of being fired or laid off when you really feel like you should be working, as opposed to the context where essentially all workers have been displaced. Intuitively, it seems that there should be more negative psychological effects from losing a job in a way that feels like a personal failing or that sets one apart from one's peers versus losing a job in a blameless way or at the same time in the same manner as one's peers. Basically, what they're stating here is as well, is that whilst you might think that a lot of people who lose their jobs during this revolution might be sad if everyone loses it at the same time like people did in the pandemic, people actually don't increase their psychological distress because everyone's in the same boat. You can see here they found that individuals who were temporarily laid off in April of 2020 reported lower levels of distress compared to their peers who remained employed. And you can see here, the widespread nature of layoff normalized the experience, reducing the personal blame and fostering a sense of shared experience. And of course, financial strain was mitigated by government support, personal savings and reduced spendings, which buffed against potential distress. So basically they're stating that there might not be as much angst and as anxiety in the future if everyone is in the same boat. But I do think that this is going to be nuanced because there are going to be a lot of industries that are affected before other ones. As we previously discussed, knowledge work, cognitive labor, jobs, jobs on a computer. Those ones are going to be automated heavily away first and it's not going to be all at once. Some companies are going to choose to implement them. Some companies are going to say, nope, we're 100% human run, never going to use AI. That's going to be their marketing gimmick. Those are the companies that you kind of want to work for and be aligning your options with. But it's going to be kind of interesting as well. Like if you're looking to work for a company, try and see what their CEO's view are on AI. Of course every CEO's incentive is to make more money, so they're probably going to do that. But you'd be surprised at how things work in the future. And this is where we get into two of the most interesting concepts here. Okay? And this is where she talks about in banks, the culture books. And this is where, you know, they are completely post scarcity society. Money is viewed as crude and irrelevant for allocating resources. Living space, raw materials and energy are produced in abundance for its citizens and the capacity of its means of production are ubiquitously and comprehensively exceeded. Every reasonable demands. It's not unimaginative citizens could make. Yet culture at least has one need that this abundance cannot satisfy. That feeling was the urge to not feel useless. Basically, to cut the nonsense, this passage highlights the fundamental and AI need to feel like they are contributing to something larger than themselves. Even in a paradise where all material needs are met, the psychological need still persists. And she states here that, you know, you can think about it right now, do you do anything that you are notably worse at than other people just for the sheer value of doing it. She talks about how, you know, being a ballerina, even though she's in her mid-20s and you know, being a ballerina is long behind her, but moving her body like that still brings her joy. And this is where she talks about post AGI. So she says. A renowned AI researcher once told me that he is practicing for post AGI by taking up activities that he's not particularly good at. Jiu jitsu, surfing and so on, and savoring the doing, even without excellence. This is how we can prepare for our future. Where we will have to do things from joy rather than need, where we will no longer be the best at them, but. But we'll still have to choose how we fill our days. We will also not need to choose how to fill our time alone. In the context where we are all out of work and where this is one of our main worries, it means we built relatively aligned artificial general intelligence for the same reasons I expect us to reach AGI are expected to progress beyond this point to where we have superhuman systems. For the same reason these systems will be helpful. If anything, we should expect that these systems will will be able to help with the problems that they create. And basically what they're stating here is that if we get superhuman systems that replace us, there's no reason to think that they cannot help us if we don't have meaning in this future world. And I think this article is very important. Like I said, I'm going to have like a, you know, a short private blog where I summarize everything into a nice area on my school. And this video, you know, is likely going to be released on the school community. Going to be released on the school community a few days earlier because this is the community where we discuss how we can actually navigate this stuff. And, and I don't want you guys to feel like, look, there's nothing I can do. There is so much that you can do in order to be proactive, especially in the last few years of work. There are investments that you can make, there are ways that you can transition your career from one area to another. There are mindsets that you can use, there are frameworks that you can apply to your life to ensure you're not one of the people that get first automated. Because whilst automation is something that is coming, it might not get you first if you're able to move out of the way and you're still able to be relatively valuable in the economy. That's what the entire thing is about. So let me know what you think about this entire document. It is definitely a long one. Maybe I did go over my time limit here, but I think this is probably one of the most important articles because I think by at least 2027 we're truly going to see the trajectory of where things are going because at that stage we will likely have had at least four to five years by 2030 of AI development. We will likely have the biggest clusters being built, will likely have maxed out the physical hardware. So we're likely going to see where those developments are going to be in the sense that any more improvements are going to be hard to come by. And therefore we know that the jobs that currently exist for humans are probably going to be there for quite some time. So this was something that most people just did miss, but I'm pretty sure that this article was something that I think is definitely valuable to those of you who are concerned about the future of post AGI economics. Let me know some of your thoughts on how all of this is going to go down and what your plans are, and if you enjoy this video, I'll see you in the next.