The AI Job Apocalypse Could Start in 2027 - Here's How to Invest

I think you should own probably each piece of the stack that I shared. >> What's up everybody? It's LG Ducet here and welcome to Milk Roadi, the daily AI show that can't decide if I'm going to be replaced by an agent or just create hundreds of agents that all sound and look like me and make thousands of podcasts a day. Today is August 17th, 2026. Remember months ago when people said that AI was coming for our jobs? Well, it might finally be coming true and it might be time to pick a few more portfolio companies to express that belief before the market catches on. Today, we're going to sit down with our lead researcher, Martin, to pick through the job numbers, the projections for AI adoption, and look at exactly how different companies are using AI to survive in this new world. Towards the end, and even throughout the show, he'll mention a few companies that he holds in his Milk Road Pro portfolio, which you can check out for just a dollar at the link below. And a reminder that our podcast today is free and it wouldn't be possible without our partners at Saber.money, the stable coin payments platform built for Asia. Keep an ear out for more information about them later in the show. Martin, welcome back to the show. Man, we haven't done a Monday episode together uh ever, but I think it's a good way to kick off the week because I feel like you have some pretty big uh let's call them ideas or maybe projections into the future uh to share with us today. >> Hello, sir. Um yes, I do. Today I want to talk about a labor market because we got some um pretty bad numbers and now people are maybe more afraid about that, you know, labor market apocalypse. Um, and I just want to maybe provide more thoughts about what is going on, why is it happening, what I think that's going to happen next and how how I'm trying to play all this sort of trend that I'm going to share today with you. >> Absolutely. And I think this is something too as we were discussing before. It's like through the winter, this was a huge theme. SAS apocalypse, jobs apocalypse, claude every day is erasing millions of jobs with their new their new releases. But I feel like nobody's updated us on this. Nobody the market doesn't really talk about this anymore. So I'm I'm keen to to know what you think and and kind of what the latest data points are. >> Yeah. Um definitely. So let me just show some charts. Um, this chart went viral last week when A16Z, which is a big investment firm, shared this chart showing that computer agents are now much cheaper and not only cheaper than, you know, um, US labor, but also offshore labor like India. And here on the chart we can see that typical range for computer use agent is between $6 to $8 per hour which is you know go and try to find some workforce around the world at 6 to8 bucks. I mean it's definitely possible but you don't want to always you know um outsource your business operations to you know Philippines or you know Asia. So this chart clearly says okay we have reached a point where even agents are cheaper even than in India. That's the the state where a lot of people outsource their their labor. And now I also want to talk about this six to eight uh bugs bucket because the way it's measured here it's using the most expensive model on the market and it's go it's opening browser and do all the clicking. So it's not even optimized for the cost efficiencies. So, I could argue that you don't need to use the the smartest model out there for everything and not certainly for going opening a browser and clicking and do stuff. You might need to do a critical decisions by the best model out there, but you can do this boring stuff by some, you know, open-source even free models. And so I would argue that even the more realistic actual numbers might be even lower than that. And it just tells me, okay, when all these big investment firms are now really comparing those numbers, we probably reach the point where it might get very confusing and challenging for the labor market because now if I am, you know, in a management of some company and I'm seeing this, I would definitely sort of challenge our position, how we are doing it, how we are approaching it, are we uh sort of adopting this AI fast enough and quick enough. And I'll show some other charts. But now I really want to move this conversation into the labor market because we got some numbers. It they don't look good at all. So let me just show two more um charts I have here. This one it shows US labor market and this one specifically shows six months change in um weekly payrolls. It just means how much is US workforce earning over time. And as you can see here um we are at the lowest numbers. We we were at these numbers in 2012. Sorry. >> Yeah. >> And now we are back there again. And this is not good for the economy because 70% of the GDP is driven by the consumption by these people. And so if these people are not going to making money is you know the whole US workforce in aggregate is not going to making more money then it's not good for the economy like >> I understand you could argue that we don't need it because all this AI boom is now funded by hyperscalers and they have a lot of money. Yes, that's fair. But at the same time, they are not immune to to like this um market this consumer struggling at all. So if they're they're struggling in the market and people don't have money, they don't spend as much money as before like even hyperscalers are going to you know make much less money than before. So it's sort of a it's a virtual loop where it affects everything. So it's not like we don't care like consumers might struggle but hypers scalar still have money that's fine. So it's it's sort of it's getting pretty tricky and it's something that people should pay attention to. Just going to pause there for a second to point out that the market is showing signs of something kind of different happening and our analysts at Milkro Pro are all over it. They spent the last couple weeks making a lot of trades, getting out of some positions, and then getting into a lot of new ones, getting ready for the next wave of robotics, space, or even kind of picking some different AI winners. If you want to see what they have in their portfolios, what positions they're opening, it's just a dollar in Milkroad Pro at the link below. >> It's the Kshake economy, right? Where it's like the stock market keeps going up, the companies that are worth a lot are going to be worth more and yet consumers have less money, right? Right. And this basically this chart sorry this chart just to understand it this is US labor momentum. So this is the growth of the labor force or of their payrolls like of of basically the the amount of money that they make or of the total jobs that they're >> the money the money >> the amount of money that they make. So it's their growth. So this and this is and this is per six-month percent change. Got it. Okay. So this dates back to 2012. So we have basically a 15-year chart almost and shows that in 2012 we hit a low of 1.4%. uh and that I guess a normal range to be in is between one and a half percent and 3% and that there's been a lot of stimulus over time where we've gone up to 3% which I'm assuming is a is a healthy number. there's an outlier of the co years which are it's not there for a year or two because co happened but then it peaked after the pandemic up to almost 7% uh naturally probably as a ton of money came back in as money was printed so um people got a lot of new jobs and and people got paid to come back to work or whatever but now that it's kind of normalized co's done um now we're at a low of that range back down to 1.5% which hasn't it really hasn't been this low other than a few outliers uh in the 2010s right So basically wages growth is at its lowest point in a long time is what you're telling me despite >> the stock market printing alltime highs. >> Exactly. And there is one >> exactly what we thought would happen just we're seeing it in the data what people projected. >> Yes. True. And there is one more chart that was also not not really good for a jobs market which is >> in July there were no new jobs added. In fact, we got less jobs in US in July. So, you know, if you like people are making less money, not not less, but the growth is, you know, stagnant or like lower than before. And now we are not even making new positions, new jobs. So, you know, it's it starts to be concerning. Now if you look at the last 5 years or so you can see that the trajectory is is going down but one month of you know negative um job growth doesn't mean anything. It just highlights that something is going on and as John who is our macro expert he said market was expecting that in July there will be about 80 80,000 jobs created and in fact we got a number that there was um actually minus 23 uh,000 jobs. So it's not as important meaning okay we lost some jobs great but the problem is that the market and all the analysts that were forecasting the growth were not projecting losing jobs and so it's a pretty big disconnect and it means that something is going on and people don't know yet about it. So that's that's the concerning part, >> right? And the concern I'm assuming from this chart, and this is the non-farmms payroll because farming is is very seasonal. So it's unfair to kind of >> have that in in numbers. This is just more like regular positions, let's call them, that aren't seasonal. >> I I'm assuming the assumption is that this trend would continue as well, Martin, right? That it's like usually when you start to see this, it's like you trends continue, so naturally July had a negative job sprint and that going forward there'll be more negative job prints. And this you're going to tell us I guess a bit more about what we all suspect is actually happening. >> Yes, exactly. Actually, I have one more chart that >> is showing something interesting. This there are three charts actually and they all show AI spend per employee per months. LG imagine that there are company some companies that are spending $7.5,000 per employee per months. So in they pay them salary and on top of that they are still spending for the AI usage that you know they they they need they use for they work and that's pretty something and what it what this tells us is that there is only a certain part of the market or of the economy that's really accelerating and that's really taking a proactive approach. approach to this because you can see that the median here is an average 12 bucks per employee per months. So I would argue, you know, that's maybe something. Hey, let's have some AI strategy. Let's just give everyone, you know, access to CH GPT so they can ask, you know, >> where they should go for vacation or stuff like that. But that's that's not the part I'm talking about here. And the difference here is huge. So we are talking about median company uh uh median is at 12 bucks and the top 1% is at 7.5K. So the difference is pretty huge and there is one more important thing that came with this chart and that is those top 1% companies are actually hiring much more people than the median companies and I think that's the key part here because if I am familiar with AI I know how to use it how to leverage it to my business I would like to add as many people as possible just show them how we use it and you know build a great business and that's what I think is happening here because you you can create you can think of it as you can create a digital co-workers or you know assistants that will help you or your customers and you can just scale it but you obviously cannot sell you know another 1,000 clients without bringing more people who who will help you to manage all that but my assump is that the companies who are ahead of the curve, ahead of everyone else, who are okay to invest heavily today in AI are going to have a huge advantage because at some point they will just eat all the slower, less, you know, proactive companies because they just are able to deliver the same amount of work for much less costs. And so that's going to be that economics pressure that's gonna, you know, I guess it's gonna just u make sure that all the companies not adopting will just die, >> right? Yeah. So it's a K-shaped economy also for actual businesses when it comes to AI use and eventually for their balance sheets, right? That's basically what you're saying is that I think this is really fascinating, too, because it's it's it's funny that there's a median the median number is $12 a month of AI spend, which I don't even know what that is. Even a chat subscription is more than that. So I I guess it's just brought down by the ones that are zeros. Um so but but but I like the top 1% part of this chart is that the companies using AI the most are spending like literally I don't know what the average salary is but let's just call it you know they're spending 10 to 15% per employee salary on AI spending for that employee as well. I'm assuming top 1% have higher employee salaries, but still 7 and a halfk per employee um per month is is a lot, but they're clearly I'm assuming they they see some kind of they're hoping to see some kind of benefit from it or or they're going to, which is hopefully what we talked more about. But this is this is a good way to visualize it, man. I hadn't seen this. >> Yeah. And I want to say that I am incredibly bullish on the companies that are in in that top 1% chart because we might not see it today but eventually at some point as they are ahead of everyone else they will get that advantage and they are not going to slow down. And so the the learning curve is quite steep but once you are on it then like everyone else is just not going to be able to compete with you and that's that's something so I want to own the companies when I was listening Q2 calls of my portfolio companies I was always hoping that I will get some answers how they are leveraging AI and I don't want to just you know yes we are using AI but how show me some numbers show me some productivity gains I want to see it. It's okay. I still don't see it on the profit statement. That's fine. But at least show me some, you know, internal gains that you are seeing and how are you using this technology. And I think the the most bullish one the message I got from Q earnings was Uber. they they got a lot of good information how they are using it, how much how much time they save and you know the whole process how they are implementing AI across all their units and so um yeah I'm a huge Uber bull and um I just love what that company is doing but I'm going to talk about a couple more companies at the end that I think are in that you know top one uh percent on on AI spend and I'll get into it. >> Great. This is also very important this chart and that shows that um this the dollar spent on closed models are now going down. It doesn't necessarily mean that um people are using less closed models but because there was a pressure on these frontier labs like open AI or anthropic to you know reduce the prices because they were too expensive and open models were you know open models cost a fraction of cost like tenth of the cost of frontier models. So they were sort of forced to also launch some cheaper models. So this chart get just means that close models because there were market pressure and push to also launch more um cheaper models. So now we can see the chart is going down and so I think it's important because this shows that the intelligence is just getting commoditized because okay maybe you know Fable 5 is the best model out there but only by a by a thin margin. So I don't really care that it's you know 2% better than something else. Yes, for some tasks I might need to use Fable 5, but pretty much for everything else I'm okay to use something that doesn't cost me a fortune. And so here, this chart just shows that even now Frontier Labs were pushed to do it and react to the market needs. And for us end users or for the companies that are using this technology, this is a gift because if there is a competition, it means that you know the prices are going to improve because there's going to be pressure and so it's better for everyone. No, not not for everyone because for open AI on topic they might not be happy right now. But this is what's happening right now, >> right? We've talked about this a lot on the rollup as well and it's something we've been tracking and even even a month ago too uh Martin because I know you don't currently appear on our rollup but I I feel like hopefully we can get you on soon. Uh we talked about this idea that that the open models would eventually or trying to put the closed models out of business by offering almost the exact same thing at at a at a cheaper rate. But now you're also showing that and that was speculative, right? That that that may happen. But now you're showing us that that close closed model usage um spend is actually dipping now. So we are moving to open model. Does this make you bearish? Just pause for a second. Does this make you bearish on the anthropics and the the the open AIS of the world? >> I mean I think that their market share is um going down but their numbers are still going up. So you know there might be more people using uh anthropic and open AI but on the bigger scheme of things I think their market share is going down. Uh so there are rumors that both of them are going to IPO at 1 trillion plus valuations and I think if you are in a business that is going to be commoditized and you are going to compete on price mainly I am not that bullish as everyone else. In fact, like at valuations above 1 trillion, I might be a little bit bearish, but I'm not going to short them obviously. But yeah, I'm I'm less bullish than everyone else, I think. >> Right. Yeah. I mean, this there there is still those rumors of of IPOs coming later this year, but this kind of chart uh not something that they want to see or their investors want to see, but still a lot of time before that happens. So, anyways, let's get back to to to the labor part you were discussing. >> Okay. And I want to show this chart as well because right now we are you know um it's August 2025 so we are somewhere here. Oh sorry it's 2026 >> and so you can still see that the main use of AI is still in that non- aent workloads. This is a prediction for from Goldman Sachs. Sorry. Um, and so you can see that I would say 90% of AI usage is still like, you know, chat interface where people just ask questions and get answers. But like as you can see on this chart, that part of just having chat interface, ask question, get answer is going to be like less important and with a decreasing market share on total AI consumption. And so this only tells me that the things that I have been talking about the 1% of companies are still so negligible to the whole market that it's it's still not relevant at all. But over time, I am pretty sure that the the the market share will change and the agents and the consumer agents and enterprise agents are actually going to take a major major market share at some point. And so, you know, it just adds more scale to what we see today because many people feel like, okay, maybe it's too late or I don't know, like definitely when I talk with my wife, she's always like, oh, I think, you know, I miss the train. I'm like, no way. You know, I think Kevin Baker said that he thinks 500,000 people use agents. 500,000 people and there are 7 billion people on the planet. Okay, so that consumption is just going to explode and so I'm super bullish on compute and I think all the people that are not like must be idiots like this is so obvious to everyone at this point and but at the same time I'm afraid that the trait which is actually showed here those are the five stacks of the AI that everyone keeps talking about. Actually there are six. So there are different layers of the AI stack how you can you know um try to express your views on AI that you are bullish but like everyone knows about you know clouds and hyperscalers and corv and nibus and nvidia and like this is like all known and a lot of people are in that trade don't get me wrong I'm in that trade too because Nvidia is trading at 16 16 times earning s next year. I think the price is just no-brainer. So, everyone should have, you know, some Nvidia. Google looks really good and like there are many companies in these layers that look really good. In fact, my biggest company is in this very bottom land power and shell that's pretty much just building that's connected to the electricity. And then there are like hyperscalers or neoclouds who will just bring their own GPUs and just plug it in and you know it works. But I think what I want to talk about here a bit more is that application layer and the harness layer because that's where the opportunity is. I think and if I >> What do you mean by harness layer? What's the harness layer? Yes, that's a good question because applications you can think of it as you can probably um use chat GPT or something and just in a very simplistic way maybe you create a project or something on cloud and so you are not using chat interface but you are not getting the most of AI let's say so the models is just the intelligence but then you have all the guardrails All the tool all the tools that you know your agent has access to all all the integrators that he can connect to all the skills that he's using what if there is incident who is going to you know um make sure that it works next time there are crunch jobs so you can think of it like now you have got access to the intelligence but someone needs to manage it and like maintain and make sure that it's working properly and that's the harness. I would argue that's probably the most difficult part of it all because now everyone can create whatever like if I want to create new Facebook or you know we just had a team call and one of our one of our teammates said I just built my own exchange so I don't need to use my broker interface because it's super ugly. So I have my own interface and so I just you know send my limit orders to my broker through my interface. So that's the world we are in right now and this harness part is really the most difficult one. So it's not about building an app like you can do it. It's pretty easy but what if something breaks? What if your model you know doesn't work? What if the servers are off? and like a lot of small pieces that make sure that you are getting as much from this intelligence boom as you can. You know, one thing we've talked about a lot on this show is that crypto is quickly becoming a huge part of the global payments infrastructure. And nowhere is that more obvious than in Asia. But if you're actually running a remittance company or a payment business, you know that the hard part isn't moving the stable coins. It's dealing with local banking partners, compliance, liquidity, and all of the operational headaches that come with sending money into places like India and Southeast Asia. That's why today's partner is Saber. They give payment companies stable coinpowered infrastructure to collect and make payouts across Asia without having to build all of that complexity themselves. They've already processed more than $3 billion in transactions across 40 different countries. So this isn't just a concept. These guys are actually doing this for real. If you're building payment infrastructure or expanding into Asia, make sure you check out saber.money. >> Is so Palanteer is is an example of a harness correct? >> Yes. Exactly. Yes. >> Yeah. Exactly. And we learned about that from you from Vincent, I forget. Somebody told us >> it was Vincent, right? Vincent. >> Yeah. >> I just I absorbed the knowledge and I forget who told me. Uh but basically, yeah, we did an episode on Palunteer a few weeks ago with Vincent, which which did very well and it was very informative for me as well and probably the audience, but that was a good example of the harness of basically the rapper for the AI because not everybody wants to go, you know, go into claudin and operate with and and and use it. It's kind of like that layer on top, right? So um that's I totally understand how that's very important especially the the bigger the company or the bigger the the entity the harness seems very important for that. >> Yes exactly and I think actually that in the future what we might see is the applications you can think of it as um I am a company and I just want to make sure that um my customer support inquiries are done by some um agent some AI agent. Okay. So I might go to um service now who has connections to all my data. They know everything about my company and they will do the the harness the the hardest part for me. So okay here is you know user inquiry and like just figure it out and because you know how to do all that harness stuff you will be able to you know answer my user and his ticket is going to be resolved and we are happy you are happy. So that's why I am bullish on like companies like Service Now because they are actually doing this harness part on a huge scale. And the the thing why I'm bullish on Service Now mainly is because they are already integrated in all pretty much all S&P 500 companies. So it's it just makes sense that everyone will want to use AI but not everyone is able to do all that harness stuff. Okay, I'm not talking about you know connecting to claude and asking like okay what I should cook for dinner today. I'm talking about okay I need some job to be done but in a big corporate firms it's not that easy and you cannot just make some changes. So there needs to be someone who takes care of everything that you know all the rights all the governance is is figured out all the safety all the other stuff it's figured out and I think the companies like service now are going to fix it for for those companies >> of course yeah so this is that that's that's a really important part I think for us to to understand especially when you lay out the stack like this right because when we're talking about stock market we we focus so much on the models naturally the cloud right? Like the the land, the power, like we we're still this year so focused on bottlenecks, compute, demand. Uh but those two parts at the top inevitably that's the end goal. That's what's that's that is the whole point of this is the application and then also how people use those applications, right? Like through through what interface, right? So um you know, I think that's very important, right? So uh it's good to shine the light on that. I have a question for you Martin and maybe you have this in a chart somewhere in terms you know we're doing this show kind of focused on the the loss of labor right and the slowdown of labor and you've kind of showed that already a lot of people a lot of projections or rather a lot of optimists or people trying to show some optimism have described this potential labor loss as uh kind of like a a deep valley through which we come out of where yes there will be a lot of job losses a lot of companies will go will cease to exist because they haven't adapted but over time AI I will create new jobs. Right? The same way the internet has created jobs that didn't exist before, the industrial revolution created jobs that didn't exist before. And that that's kind of a a natural cycle when there are new technologies is that a lot of old jobs disappear and over time a lot of new jobs are created. Is that something you foresee especially now that we're talking applications and harness that are that are areas that that will have to have some maybe some level of human operator? >> I am actually on the other side of of this sort of narrative. I think that AI is gonna eat a lot of jobs and yes they are gonna add some but not like it's gonna be much much less than the jobs that are going to be you know replaced by AI and I am pretty confident about that and I think the the market is not seeing it today which is fine and I think the reason why market is still not seeing this is because a lot of companies are still going through that what I called exploration experimental phase where they are trying the technology and see what is going on what's working what's not how to implement it and it's not easy when you are a big company so it takes time but my base case is that in 2027 we are going to start to see the results of that and that's going to you know maybe some companies will still do hire but those will likely hire because they know how to use AI and so they will hire people to to build more products, more services and they are going to eat the lunch of someone else's and then the company who are not using AI are going to struggle a lot because as I showed at the very beginning the costs are going to be reduced dramatically if you use AI versus if you use just a human and at some point the market will punish you if you don't. So I think 2027 is my base case where we start to see like today I showed that the labor market looks weak today but I think you know it might still be volatile. We might add couple more jobs next month and then um some other jobs month later I don't care. But I think that in a year or two we are going to start to see some real weakness in the job market. And that's when I think the the market and the whole economy is gonna start talking about the recession or what is going on. How are we going to figure this out? Um and that's going to be quite interesting and I'm a bit worried that the market will will just fall down pretty quickly once we we will start to see this narrative. So I I prefer to hold the companies that are you know beneficiaries of this and once I see that this narrative is happening and then the people was are starting to realize that this might be a pretty big big problem for not just us for every every government every country because there are going to be a lot of people without jobs then that's the time when I think I'm going to maybe get some cash um some reduce my positions and be more defensive. But I right now I I'm still not thinking that's the case. That's not the trade for today. >> Got it. What What is the trade for today, Martin? I think that that's by this point in the episode that's probably what a lot of people are wondering, right? Is like what how do you express this, right? How do you and and you and and and and you've written a lot in pro and and we discuss this a lot and Kyle talks about this a lot as well as like those winners of the application trade and >> you know Kyle Kyle and Vincent just give people a little sneak like they hold companies like Eli Liy in their pro portfolios right that are actively using AI to develop new drugs and accelerate all the stuff that they're doing and and eat market share there as they do as kind of leading biotech. But obviously that's just one that's still pretty narrow, right? In terms of that's just one business. there's so many ways to express this belief that you're saying especially at this point in time. Um so tell us about kind of your your strategy there. >> Yeah. Um I talk quickly about service now. So that's part of my harness sort of layer. But uh also what's really interesting these days is Epin which is a company that I I already have in my portfolio and their whole business is pretty much they are selling ads on their platform and the better the intelligence the better their model is. And so I think they are hugely benefiting from intelligence getting cheaper. And by the way, I was talking about that some companies are going to be able to leverage AI. Epavin is um 130 billion company or something around that and they have 500 employees. So that's a lean business when you know you have old A-level grade people and they are really um leveraging the intelligence that's getting better because then they can optimize better and their model when their models are better the people that spend money on their exchange is going to bring better returns for advertisers which is which is sort of starting the flywheel and that's why their business keeps growing 20% plus year over year. And it's not just out of nowhere. It's because these guys know how to use AI and they use it so well that they are able to grow so quickly. And I think that's the opportunity. And right now what happened is the Q2 numbers weren't as good as before. And so market dropped pretty drastically. I think it's it's around 320 bucks today. So it's it's down like 20% in the last few weeks. um despite their business still being healthy. The only issue why the numbers weren't not as good as market expected or even guidance was that CEO said that their model keeps improving and they didn't manage to make the last improvement to their model in time. So the numbers are not as good but it's probably you are going to be able to see the the improvements in Q3. So like the business is still good. They were not just able to update their model quickly enough to to meet the you know um expectations of a huge growth. So I think that's really good good time to buy some epin. In fact, I would like to come in here next week and like talk about Applovin a bit more because I think this is really great opportunity and people are still not seeing it and I just love that company. >> It's spel this is you know when you first started writing about this company a few months ago in pro I I it sounded like some kind of McDonald's product which it is not but it sounds like some kind of McDonald's promotional thing. It's literally app loving like a p l o v i n like that's it sounds like a new product but it it's it's from what I understand I learned this from you Martin it's it's it's an enormous player in the kind of like mobile game ad infrastructure of like something it's a it's a layer you never see uh but if you have any kind of games or anything like that on your phone or your kids do or whatever app loving is a huge part of that marketplace and I'd be I'd be very excited to do a deep dive on that because I do feel A it's something that if you search that on YouTube like no one really talks about it and yet like you said it is the market cap is 106 billion right now. So it's a huge company right or not a not a small company let's say um and and they are using AI very actively to kind of uh crank up their profits and also just just do a better job of what they're doing and and like you said they have they have very few employees compared to their market cap. So um that's definitely something definitely we should next week we should do a full episode on them. >> Okay let's do it. Yeah. >> Yeah. Any any other takeaways, Martin, before I let you go about about all this? Like anything else that people should think about as they uh you know, if they buy into your thesis here um and they get ready for for these numbers maybe to get reflected in the market. Any other you know ways for people to think about this? >> I think there are two ways how we can play this. One of them is use and learn how to use AI because only then you will become a superhuman and you will not be irreplaceable anywhere and as you said at the beginning the car shape keyshaped economy is actually going to accelerate I think and so people who are able to use AI are going to be the winners um and then unfortunately There are going to be a lot of people who are not going to use AI or at least not in a way that they could. I mean if you chat with you know chptd that's fine but you should do much more because it's possible and it's definitely gonna you know affect no matter what you do it's going to affect everyone and I'm not just talking about the knowledge workers because you know we are going to have robots soon and now they can also have that intelligence so like I think that the whole economy and the whole world is changing so quickly And like really the winners are going to be the people who take advantage of that and who are rather you know trying and learning as much as they can how they can use AI to stay relevant versus people who don't. So that's one thing. Um and the second one is about your portfolio. Um I think you should own probably each piece of the stack that I shared. So it's not just but you know maybe it's a good to think about the diversification. So like a lot of people have just you know maybe just Nvidia or they they might have just some hyperscalers that's fine that's okay I think those are here to stay for sure but it's okay to think about the diversification part you know maybe because this this last few months like hyperscalers were really struggling but neoclouds were doing really well but maybe at some point in the future applications will really take off and that's the moment you already want to own them as well so it's not like you are not going to play a catch-up trade but you are going to you know front everyone else because you think that what I was talking about today actually makes sense and at some point you will see the applications and harness layer and companies who are taking advantage of that will you will see it on on their income statements and you know earnings report. diversification is key, man. And you are the biggest uh I guess representative of that in our Milkro Pro uh analysts is that you know your your portfolio is pretty steady even through a lot of the these swings because you're you're diversified across a lot of different assets. And again, if people want to check that out, uh it's on Milk Road Pro. It's just a dollar to try it out. So you can check that out at the link below regardless if you're on YouTube or Spotify or whatever you're listening on. Um if you honestly you can just pay a dollar just to look at what else Martin has in his portfolio and if you don't like the rest of the product, you can you can take off. But again, just a dollar to to peak behind the curtain. Otherwise, Martin, uh great to see you again. Great to hear more uh doomsday theories, even though you never said it. You never you didn't say it as a dark world, but uh definitely a lot of disruption coming and especially based on those charts you showed. So, thank you, sir. Uh and we'll see you next week for some for an app loving deep dive. >> Yeah. Thank you, sir. See you next time. >> Want to stay ahead of the biggest technological shift in history? Subscribe now to get insights straight from the sharpest minds in tech and finance. Quickly, you'll note this show is for educational purposes only. Nothing here is financial advice. Investing always carries risk. Never invest more than you can afford to lose. Thanks for tuning in. See you in the next one.