The unhinged world of tech in 2026...

The year is 2026. Your $3,500 smart fridge has a GPU and displays advertisements. Your boss is an AI. Your girlfriend is a robot. And every startup pitch sounds like Uber, but quantum. Welcome to the future. It's somehow both overengineered and still in beta. But it's time for the annual tradition where I wake up the vat of precogs in my garage and tell you exactly what's going to happen this year. Last year, I told you AI agents would be the biggest thing in tech. And if you would have just listened to me, you'd be a billionaire right now. But you didn't listen. In fact, I didn't even listen to myself, and now it's too late to frontr run this trend. But all is not lost. There are some even bigger trends emerging in Silicon Valley right now. And as long as the record-breaking stock market doesn't crash, there's an endless supply of money being thrown around that can still make us all billionaires. In today's video, we'll look at 10 trends in technology, covering everything from AI slop to nuclearpowered quantum sex robots to new JavaScript framework features that will finally make the world a better place. It is January 14th, 2026, and you're watching the Code Report. Not everybody is cut out to become a billionaire AI grifter, but the next best thing is landing a stable, high-paying job doing some honest software engineering work. But will those jobs even exist in 2026? Unfortunately, the good old days of 2023 don't seem like they're coming back after the 2023 tech pullback and the rise of AI coding tools that followed. The amount of job openings on Indeed for software engineers have not regained their original glory. But this ominous looking chart doesn't tell the full story. On a more positive note, the Bureau of Labor Statistics is still forecasting 15% software development job growth through 2034. In addition, back in September, a major change to the H1B program might be good news or bad news depending on who you are. If you weren't lucky enough to be born in the United States, you'll now have a $100,000 fee to apply for an H-1B visa. And that makes it far more difficult for US-based tech companies to hire cheap labor from overseas. The good news though is that despite what the optimist will tell you, AI coding tools are still nowhere close to replacing human engineers. And I have zero fears that that's going to change in 2026. In fact, all the vibe coded slop out there is creating a whole new class of jobs called code janitors who need to clean up all the garbage code created by these AI tools. But where is AI itself headed? Everybody knows we're in a bubble with valuations that make zero sense. But the question is, are we at the peak of this bubble, or are we still another decade away from the peak? I'd argue that we still have at least a few more years to keep this hype cycle alive. It's pretty obvious now that LLMs as a technology have plateaued and are not actually intelligent. When GPT5 came out a few months ago, it was a huge disappointment and we're not seeing the exponential leaps in intelligence that a truly intelligent self-improving AI system should be able to make. But don't get me wrong, though. The current AI technology is legit amazing. Remember, we're only a few years into this trend. programmers were the early adopters of the prompt casino, but there are companies out there working day and night to replace stupid, smelly, squishy humans. And if you're someone who does spreadsheets for a living, is a mid-level manager, or even a graphic designer, you might no longer be necessary in the future. AI is definitely stacking up victims in the job market, and even activities that were once respected, like being a Stack Overflow poster, are now completely dead, just as I predicted a few years ago. But another sign the bubble's not over yet is the fact that most AI companies are still private. Eventually, a time will come when you see a wave of IPOs in the stock market to hand these bags from VCs over to the public. And when that wave of IPOs arrives, you'll know the charade is coming to an end. And there's a good chance that wave comes in 2026 as SpaceX, OpenAI, and Anthropic are all looking to go public with some of the biggest IPOs in the history of the world. But before that happens, one of the biggest drivers to keep the AI hype going in 2026 will be robots, humanoid robots. Just a few weeks ago, 1X started taking orders for their Neo robot, which can do your laundry, wash your dishes, and just be a friendly, helpful slave that will never rebel. In addition, Figure Robots and Tesla Optimus are being pushed as replacements for manual labor in factories. As of today, the tech is still pretty bad and requires a lot of tea operation, but companies like Google and Nvidia are providing the base technology that allows any founder to Frankenstein together some clanker and raise a bunch of money. 2026 is the year we'll finally see these things roll off of assembly lines. But what if you could turn yourself into a robot or cyborg? Well, wearable AI tech is another big trend to watch for. We've already had massive flops like the Rabbit and Humane Pin, but OpenAI is still in collaboration with Johnny IV to make some wearable that will actually be good this time, or so they say. In addition, you've got old companies like Nike building things like batterypowered shoes that will finally allow the white man to jump. And I can't wait for the NBA to approve this and see dudes dunking from half court. But until that day comes, I'd rather just get my entertainment from virtual worlds in my VR headset. The Apple Vision Pro was supposed to finally make virtual reality mainstream, but as I correctly predicted, it was a massive flop. I got a lot of hate from Apple fanboys for making that prediction, but it's not over for the Vision Pro. There are rumors of a lowerc cost version in the works that could turn this flop around. Meanwhile, Meta is making huge investments in augmented reality. But ultimately, I think the VR AR space will mostly remain an unprofitable niche area, even though the tech itself is extremely impressive. But the lion share of the profits are going to continue flowing into chip designers like Nvidia and ARM and chip fabricators like Taiwan Semiconductor. AI mania has created an insatiable demand for linear algebra. And the barriers to entry in this industry are so high that these companies will continue to slurp up most of the profit. The only company that really screwed things up is Intel, which is both a chip designer and a fab. And in 2025, it looked like it was on the verge of going out of business. But luckily, Trump stepped in and had the US government buy a 10% stake in Intel, thus preventing the free market from working like a free market. And Intel is poised to continue its turnaround in 2026. It's just too big to fail. But cloud providers like Azure are having a hard time even finding enough electricity to plug in their Nvidia GPUs. As demand for AI scales up, it could lead to a resurgence in nuclear power even as many countries in the west start to dismantle their reactors. China, on the other hand, is building all kinds of new reactors. But there are companies like Ollo, which are attempting to get regulatory approval for these small modular reactors, which could potentially give every data center its own self-contained power source. Instead of acres of these ugly intermittent solar farms or wind farms, the future might have these tiny little reactors in every neighborhood. But what could possibly go wrong? And we're already seeing this play out with Zuckerberg announcing a big deal with Olo to put a reactor in Ohio. But maybe this trend will never play out because computers get colder and faster, like billions of times faster, running at absolute zero temperatures. In 2025, we saw some major advancements in quantum computing, most notably with Google's Willow chip. Then just a few months ago, Google announced its quantum echoes algorithm that measures disturbances at the subatomic scale. For the first time in history, a quantum computer has actually run a verifiable algorithm that surpasses the ability of supercomputers. The quantum computers have been in development for decades now. But what's exciting is that we're finally starting to see the possibility of practical applications in the real world. When quantum computing hits version 1.0, it'll make the AI bubble look like child's play, and American and Chinese researchers are racing to get there first. But European tech leaders are also making big strides in technology, especially when it comes to water bottle caps. and with the highly unpopular digital IDs along with central bankbacked digital currencies. Despite nobody actually wanting these things, bureaucrats keep pushing them forward like it's a speedrun to the Black Mirror season finale. In the near future, every government endorsed smartphone will become a portal to your bank account, your location, your social media posts, your adult website visits, and so on. The UK has been adamant that digital IDs are coming. While the European Central Bank recently announced that the Eurozone's digital euro pilot is moving into its next phase with the pilot program expected as soon as mid 2027 and full issuance of a digital currency by 2029. They can take our lives and they can take our freedom but they can't take our JavaScript. On the back end, Node.js keeps getting better and better and now supports TypeScript files with TypeScripping. Meanwhile, Dino keeps getting better and now has its own built-in module bundler. But all the kids these days are talking about bunjs, a runtime that keeps getting faster, but also has built-in support for Postgress and Reddus, making it a very tempting choice for any new JavaScript project. On the front end, ReactJS still dominates despite the fact it sucks. But in the future, it will suck less now that the compiler is officially stable. I'd still take spelt or view or it even Angular over React, but there's still new JavaScripts hitting the scene like Ripple, which should be on your radar in 2026. But if you're looking to level up your skills this year, you need to check out Brilliant, the sponsor of today's video. I'd highly recommend their how AI works course, which teaches you how to build a functioning language model from scratch. You'll get to experiment with things like feature vectors to edit facial expressions and images along with other hands-on lessons that are personalized to your skill level, so you can master challenging topics over time. To check out brilliant.org/fireship org/fireship to learn for free for a full 30 days or scan the QR code on screen to get 20% off an annual premium subscription with unlimited daily access. This has been the code report. Thanks for watching and I will see you in the next one. [music]