Hey guys, it's the last month of the year and our last video of 2025. It's been one hell of a year and for tech some things were exciting, some maybe not so much, but what a time to be alive. So, in our last video of this year, I want to look ahead and see what's around the corner. So, have a seat, grab a cup of tea. We'll be talking about all things global. Global job market, geopolitical influence on the tech industry, the new buzzwords of the next year, and the AI theme of 2026.

Before we dive in, a quick announcement. Our channel is turning one in a few days, and we will be releasing our Christmas special about what's going on behind the scenes. If you're curious to know what it takes to run a YouTube channel while having a full-time job, how much it costs, what the production process looks like, and how my life has changed over the past year, feel free to subscribe and get a reminder so that you don't miss it. And now, for the last time this year, let's dive in.

So what's happening with the job market after 3 years of unbelievable turbulence globally? We are not trending towards prepandemic. We are nowhere near back to 2021. The layoffs are still very much elevated. The demand is concentrated in pockets and if there are any rebounds, they're happening in sector specific and skillspecific industries. I remember how in late 2024, early 25, there were a lot of talks from very reputable media resources that I have been following for many years and they were saying that a new wave of opportunities is just around the corner. That the market is picking up and that we'll be seeing an ocean of opportunities. They're coming any day now. But now going back, I'm realizing that a lot of it was coming from career coaches and people who sell the idea of getting a job in any market and in any economy. I remember being skeptical even back then because on the ground it did not feel like anything was picking up.

It definitely wasn't disastrous, but I cannot say that I noticed any significant changes or improvements. The only thing that was kind of noticeable is that the mass layoffs were not as common. So now it's a good time to kind of look back and think about it in retrospect and see whether those predictions actually came true. So I looked up some data. Tech layoffs in 2025 matched or even slightly exceeded 2024 levels. The reason I'm saying they matched or exceeded is because there is a slight variation in terms of data depending on which source you're looking at. But what I can say with certainty is that they're definitely not lower. On the positive side, there are definitely better than the 2023 peak. So, let's go region by region and study the trends across the world.

Starting with the United States, the theme of the 2026 job market in the US is the hollow middle. The biggest trend of 2025 that will continue into the next year is replacing highcost local headcount with nearshore labor. And for obvious reasons for the United States, the nearshoring happens in Latin America. What US is seeing is the barbell effect. They're cutting a lot of middle management based in the US. But on the other hand, the money being offered for principal architects, staff engineers, group PMs, or any kind of principal role is sometimes beyond reasonable. And I'm saying this as a PM. And then the entry-level market is often just outsourced to Mexico or Brazil.

So, actually, let me take it back. Maybe it's not a barbell after all. It's more of a kettle bell so to say. What we're seeing is that even for product engineering or R&D talent, the senior roles that tend to stay inhouse are the ones that are tied to strategy and business. Something that would be really difficult to outsource because there is a lot of business and market context that you understand when you live in that market as well as the cultural nuance that you just can't teach over Zoom. The point I'm trying to make is that for a senior level engineer, for example, coding is not enough. Unless you're tied to business and strategic decisions, you can very much be replaced by an engineer in Colombia with the same skill level in the same time zone.

Moving on to Canada. No matter what you do for a living, be it coding, product, marketing, or HR, at some point you get to experience the beautiful world of documentation. And as a PM, I have to write feature specs every single month. The problem is people don't like reading feature specs and the work of the whole team sometimes goes unseen because it's a lot easier to see what was built, not read three pages of text. So I thought, why don't I build an app that will turn my feature specs into beautiful landing pages, so neither myself nor my customers have to suffer and people will actually be excited about what my team has built. Let's do it together right now. Watch me do it in a minute. I'm opening UWare and just describing what I want. Build a feature spec visualizer for product managers to convert feature specs into a landing page for a new product feature. Adding some details here around the specs around the wireframe and how I want this to look like. And finishing it off with a CTA. Let's call it, I don't know, for example, visualize. And create the page. And my laptop dies now because it needs a Christmas break. Not a problem. I'm going to move to my phone and continue from there. All right, let's test it. The feature that I'm launching is advanced search filters. The users can now filter search results by date range, category, and custom tags. All of this makes finding specific content 10 times faster.

Okay, that's actually sick. It just turned my feature spec into a landing page for my feature. Let's try another one. I am launching a dark mode toggle. Users can now switch between light and dark themes in settings. It is located under preferences and then appearance. And it reduces the ice strain and saves battery on mobile devices. And now instead of making my customers read pages and pages of text, they can actually see what's new and what was built. And I can publish and share this instantly. I don't need to worry about hosting, deployment, or setup. And the best part is that all of it was built in minutes. Try UWare yourself. Link is in the description. Turn your ideas into apps without writing a single line of code. Huge thanks to UWare for sponsoring this section.

The theme of the 2026 job market in Canada is going to be nearshoring plus selective hiring. In Canada, the number of people currently with jobs is actually flat or slightly up. What this means is that the crisis is entirely on the hiring side. If you have a job, you're most likely safe. Not guaranteed, but most likely. But if you're looking, it's pretty tough out there. postings for senior roles are slightly up and the junior roles are down 25%. And this tracks with the trend around expertton market. So back to those predictions from earlier this year. Yes, there are opportunities and hiring is happening but the hiring is happening for highly critical and very senior positions. Canada just like the US is also experiencing nearshoring and offshoring trend for a huge number of roles. Even among the people that I personally know, I can count quite a few who got laid off because their entire departments were moved to a cheaper location. From what the data is telling and from my personal experience, the roles that are being kept inhouse are tightly coupled to decision making and core architecture rules whereas execution is often pushed offshore.

Moving on to Europe. Europe is going through a very interesting time because they are essentially at risk of losing their manufacturing base. In general, the European Union is deeply concerned about deindustrialization going into the next year. Germany has lost a quarter million industrial jobs in the past few years and this bleeding will accelerate in 2026. The automotive sector which is the spine of the German economy is first of all not doing well and secondly under a threefront assault. The Chinese are undercutting on price. The full shift to electric is much more expensive and difficult to execute than it seemed and is currently unaffordable to be honest. Plus don't forget about the energy crisis and Sweden's bet on being the country of green tech was a pretty bad miscalculation as well.

They're now reversing course after colossal bankruptcies of the green tech giants and mass layoffs in the industrial sector. On the flip side, even though general software hiring is stagnant because of the manufacturing recession, there is a massive demand for roles around AI sovereignty and regulation. I will speak about this in detail in a few minutes, but the demand for cloud architects with EU specific compliance certifications and data guardians has spiked almost 50% as Europe is gearing up for something called the Euro stack. Again, more on Euro stack in a couple minutes.

And now the Gulf and specifically the UAE and Saudi Arabia. The Gulf states are actively promoting the nationalization pivot and they're pushing for Saudis and emiratization in their attack rules. Unlike in mid to late 2010s, you can't just move to the UEE or Saudi Arabia as a mid-level PM or an engineer without a specific domain knowledge because they have enough foreign talent already. The roles are there, but they now require specialized local knowledge or Arabic fluency, which by the way, in my humble opinion, I do think that local languages should be prioritized. No matter how international the market is, no matter how many expats they have, I am of the opinion that the countries should be prioritizing local languages over English. So in my view, it's expected, not surprised. I understand why they do it.

The Gulf is going through the same situation as the US and Canada where cuts disproportionately hit the middle layer, middle management, operations, and execution rules while AI, cloud, cyber security remain fairly protected. At the beginning of the year, I made a video about 2025 tech trends. And one of the biggest ones that I highlighted was the rise of individual contributors. And now that we're approaching the end of the year, I stand by what I said. The biggest leverage in today's job market are hard skills and domain expertise, which makes one an elite level individual contributor. Now, if you're curious why I'm not covering China or Asia as much, it is only because I have never lived or worked in those regions. I simply don't have the same cultural context and contextual knowledge of the Asian market, which is why I prefer to stay in my lane for now at least.

So, a quick verdict. Was 2025 better than 24 across the globe in the context of tech employment? No. For an average worker, it was arguably worse. The whole job hopping every 2 years is a thing of the past. Quit rates have plummeted and the market is quite frozen unless you are in a highly specialized role. As of today and as we're moving into 2026, being a highly skilled individual contributor pays off more than being a middle manager. The trend of 2026 is to accept the reality of distributed teams. I personally think that we're going to stop using the word offshoring or nearshoring in the near future. The satellite offices become the norm. The market will not recover to 2021 levels. Do not expect that. It's not going to happen in the near future. We are entering a multi-year period of job market compression, which means fewer jobs, much higher barrier to entry and a focus on ROI. And if you are an average worker without specific hard skills or industry expertise, my recommendation would be to specialize as soon as you possibly can. Remember, these days narrow is the moat, whether you're building a business or a career.

Moving on to my favorite topic, tech and geopolitics. While the US and China fight over chip supremacy, the rest of the world is fighting to own the infrastructure so they don't become digital colonies of either US or China. Why is infrastructure and data sovereignty so important? Let me give you an example. Imagine there is an AI company in Singapore that is building a very powerful AI model and it works great until they try and sell it to a different country. It gets blocked by Singapore's data residency law because the model was trained on servers in Virginia. It gets shut down in the EU because it doesn't comply with the European AI act and it can't operate in China because well, China. So the startup realizes we can't go global. And this is why the Saudis building those data centers and NEOM aren't doing it just to run GBT faster. They're building it to say, "You want to sell AI to my people? You come to my server, you come to my law with my rules, and you pay me." This framing defines the next era of geopolitics in tech and the rise of the third AI power center. The swing states of tech, bricks, UAE, Saudi Arabia are refusing to be just users of US and Chinese tech. They're building their own third stack to protect themselves from a situation where the US says, "I'm putting a 100,000 sanctions on you. And unless you do as I say, we'll cut you off from the rest of the world." China doesn't have this problem.

China is a superpower and they invest heavily in building domestic AI with their own chips and their own cloud. So, we got the US, we got China, we got Europe. So, what are we left with? Well, just about the entire half of the planet, the global south, give or take. But treating the countries that are in the global south as one unified entity is a big mistake. They're very different. India produces about 20% of the world's data, but only holds 3% of global data center capacity. India is datarich, but infrastructure poor. Latin America is in a very similar situation, but Saudi Arabia and the UAE are actively recovering from the Dutch disease. So instead of buying chips, renting models, and simply relying on US, China or Europe, they're signing contracts with all three to build the relationship. And they're building the entire supply chain on their own ground. The way I see it is that the global south is splitting into two camps. Camp one being the Gulf and specifically the UAE and Saudi Arabia. These are the only non-western powers that have the capital to buy their way out of dependency. The UAE and Saudi are capital rich infrastructure hungry. They can't match India or Latin American population, but they can build gigantic physical data centers. They're the only ones who can write a hundred billion dollar check to OpenAI or Nvidia and force the transfer of technology to the region. And camp number two, the data colonies, India, Brazil, and Indonesia. These nations are very rich in data but poor in infrastructure. Like I said, India generates about 20% of the entire world's data, but they only have 3% of global data center capacity. India exports data, its most valuable asset to the US servers. There it gets refined into intelligence or models and sold back to India as a subscription service. This is a tech equivalent of the cotton trade. Sell cotton cheap and buy expensive shirts. Brazil and Indonesia are facing a very similar story. Massive user bases, massive population, good talent, but their tech infrastructure runs on Amazon and Microsoft clouds. If the US turns off the tap, the tech economy goes dark. So this was the global south.

And finally, Europe. Europe's problem is that on top of losing ground in the manufacturing realm, they got to do something about data, and they better do it sooner rather than later. Europe owns only 4% of global cloud capacity. Every bite of European data, every AI model trained on it, every service delivered to a European, all of it runs on US servers. So the European government was thinking and was like, "Okay, let's fight back." So they're launching a moonshot project called Euro Stack, which is a European cloud built on European servers and under the European law. Euroack is not a single company. It's a policy. And the plan is to glue together European tech assets into one usable stack. It includes European chips, European operating system, European AI models like Mistral and a jurisdictional gate to protect Europe from foreign laws like the US cloud act. Europe's biggest problem right now is fragmentation. If you think about the US, they have three centralized giants, AWS, Azure, and Google. Europe at the same time has 27 different strategies, a dozen startups, fantastic startups especially in Sweden and a European committee. So for Europe, 2026 is about bringing all of this together and making it the year of Euro stack so they can become defensible.

Moving on to business trends of 2026, the rise of NeoClouds and regional stacks. So we spoke about the country level. Now, let's zoom into the market level. Imagine you're moving from building A to building B. Both buildings are on the same street. You get a U-Haul, and U-Haul gives you a massive truck, a massive 26- ft truck with hydraulic lift, climate control, and GPS. But you're moving a one-bedroom condo on the same street, but you're paying for the whole truck. But U-Haul is the only game in town, so you don't have a choice. In 2025, AWS and Google became that U-Haul truck. A startup training an AI model pays for compute plus storage plus database licensing plus data transfer plus billing complexity that they simply do not need at that scale. And then Coreweave and Nebus the Neoclouds step in and they say we're not U-Haul. We will sell you just as much as you need and for 30% less. And what the Neocloud business is doing is that it makes compute costs cheaper. And when compute costs become cheap, the entire structure of tech inverts because suddenly a startup with a $2 million funding can train a world-class model and the barrier to entry falls. Neoclouds are in the newest form of tech business with an extremely high leverage that keeps rising across Europe, the Gulf, and Latin America, which means that we're heading towards a trend of decentralization.

Where does this leave us as we enter 2026? The geopolitical situation of the past 5 years as well as the pandemic deeply affected the tech industry. And if this isn't giving you the cold war vibes, I don't know how else to call it because we see this race unfolding right in front of us. And the race is for the AI infrastructure and data sovereignty. China showed an example of how one can survive and thrive even when you're cut off from the rest of the world. American hyperscalers are actively investing in metal in chips because neocloud model is undermining their business model and their ability to control. The south is looking up and goes you guys have fun over there but we got the population and we got the land. So if you want our data and our users you come to us you open your offices here and you pay us. And last but not least let's move to the AI theme of 2026. 2023 was about knowledge and chats. 25 was about action and agents that could do something with that knowledge and in those chats. 26 is going to be about physics. The theme of 2026 will be physical AI.

I've spoken about this at length and if you're interested, I recommend that you explore reputable research papers because it becomes pretty obvious that the LLM project has reached the plateau of productivity. We have hit the physical limits that prevent us from achieving further massive leaps. So instead of building the next LLM, we're finally entering the era of execution and putting existing AI to use. In 2026, AI will start becoming ubiquitous. We will be seeing it everywhere in kitchen appliances built into new devices. It will be in the cameras at factories and plants. Your power grid is about to get intelligent or smart. Your government IDs will likely be linked to AI agents that file your taxes. All in all, we're about to enter the first year of hardware AI. This means that if you work in manufacturing, logistics, energy, healthcare, appliance production, mining, the most labor intensive industries on Earth, this is your year. You will be seeing a lot of pilot programs, new buzzwords such as spatial computing, and large world models.

In a nutshell, in 2026, artificial intelligence leaves the browser and enters the physical world. Conclusion, there is no going back to 2021. Not for jobs, not for tech, not for geopolitics. We're building new walls. We're building new clouds. Nations are trying to protect themselves from a big power turning off the switch. But despite the setbacks that this creates, I am low-key glad because I think that we will be seeing a much stronger diversification across tech and tech hubs around the planet where each region has something unique to offer. So once again, it's a fascinating time to be alive and let's see how the next year will unfold.

We would like to thank those of you who stayed with us this year, those who believed in us and gave us your time. We cannot imagine getting to where we got in just one year. So from both of us, myself and Maria, thank you and we will see you in our anniversary episode.