Transcription
According to MIT, 95% of corporate AI projects have failed to replace human labor, with even the most powerful models automating less than 3% of human tasks. And while over half of all employers now regret AI-driven layoffs, nearly 75% of the S&P 500's growth has been driven entirely by AI speculation, raising the question, will we once again be forced to bail out enormous amounts of corporate greed?
While the United States continues to enter periods of severe financial speculation, largely because the consequences are never paid by those responsible, this artificial intelligence boom is once again proving that while history may not repeat, it always finds a way to rhyme. With an emerging AI winter eerily mirroring those of the past, just years prior to the historic .com boom and bust, the United States was already experiencing a full-blown artificial intelligence frenzy driven by geopolitical competition with Japan.
Throughout the late 1970s and early 80s, Japan rapidly consolidated dominance across semiconductors, consumer electronics, and precision manufacturing, culminating in a decade-long government-backed effort to build machines capable of using artificial intelligence to reason, communicate, and respond to spoken language known as a fifth generation computer systems program. The initiative was frequently compared to the American drive to land a man on the moon with the explicit goal of overtaking US computing technology. And as American economists openly predicted that Japan would surpass the United States as the world's leading economic power by the year 2000, US policymakers and corporate leadership increasingly framed artificial intelligence as the decisive pillar of economic survival and global influence.
This culminated in the national expansion of the expert system, an intelligent computer program designed to use encoded knowledge and inference rules to solve problems that would typically require the judgment of leading experts within an industry. As the federal government poured hundreds of millions of dollars into these systems while also encouraging corporate rivals to work together under the banner of national security in traditional corporate fashion, the exact opposite occurred. Rather than grounding artificial intelligence in what existing hardware, data, and logic systems could realistically support, firms raced to signal relevance through prototypes, pilot programs, and inflated expectations, with headlines even claiming that a single expert system could replace the work of 10 specialists.
In reality, these systems could only automate extremely narrow tasks such as screening loan applications or flagging routine decisions and routinely failed when confronted with real-world complexity. This most famously occurred when a medical expert system claimed that a male patient's infection had been caused by a procedure performed exclusively on pregnant women simply because researchers failed to include a rule accounting for gender. And as Japan's fifth generation computer systems program ultimately failed, while expert systems began creating more problems than they solved, jobs were lost, taxpayer funds were wasted, and the American workforce was forced to pick up the pieces of yet another broken hype cycle.
Today, while Japan has since been replaced by China as a nation's boogeyman, hundreds of thousands of lives have once again been uprooted under the language of national security and productivity as artificial intelligence is used to justify workforce reductions long before it is demonstrated any ability to reliably replace human labor. To test whether artificial intelligence can actually automate human labor at scale, researchers from the Center for AI Safety and Scale AI released what is known as the Remote Labor Index. In the test, state-of-the-art models built by companies like OpenAI, Anthropic, and Google were tasked with completing the same work currently performed by paid human workers under identical expectations of quality, accuracy, and consistency. The results showed that even the best-performing model achieves an automation rate of just 3.75%, with Gemini falling below 1%.
And this widening gap between hype and automation is not limited to white-collar work. As fast-food chains have pushed to aggressively replace minimum-wage workers, as if that's the primary driver for rising costs, McDonald's quietly ended its AI drive-through pilot after the system repeatedly failed in basic real-world conditions. Orders were miscounted, including totals as high as 260 chicken McNuggets. Customers were charged for nine iced teas instead of one, and the system routinely placed bacon on ice cream. McDonald's also removed an AI-generated holiday advertisement following widespread backlash, later attributing the issue to its Netherlands operation, stating that the ad was intended to reflect the stressful moments during the holidays in the Netherlands. But we recognize that for many of our guests, the season is the most wonderful time of the year.
In the rush to remove human drivers, Tesla launched its robo-taxi service in Austin 8 months ago. At the time, Tesla promised 500 vehicles in Austin, coverage for half the US population, fully unsupervised rides, an expansion to 8 to 10 cities, all by the end of 2025. Today, the service operates roughly 42 cars in Austin. Availability remains below 20%, and unsupervised operation has either been paused or quietly redefined. In some cases, with a Tesla employee sitting in a vehicle behind the robo-taxi to monitor its behavior. And while it may not be a perfect one-to-one comparison, the robo-taxi is currently crashing at roughly nine times the rate of a human driver, while a competing autonomous driving service admitted during a congressional hearing that after striking a child in a school zone, its vehicles rely on overseas human operators to help make driving decisions when it becomes stumped.
And while responsibility for life-or-death decisions is quickly being outsourced overseas, domestically, artificial intelligence is already producing unbelievably severe consequences. In 2021, a unit of Johnson & Johnson announced what it called a leap forward, adding artificial intelligence to a medical device used to assist ear, nose, and throat surgeons during operations. Prior to the AI update, the FDA had received unconfirmed reports of seven device malfunctions and one patient injury. After artificial intelligence was added, the FDA received reports of at least 100 malfunctions and adverse events. Most involved the system misinforming surgeons about the location of their instruments while operating inside patients' heads. In one case, cerebrospinal fluid leaked from a patient's nose. In another, a surgeon punctured the base of a patient's skull. In two additional cases, patients reportedly suffered strokes after major arteries were accidentally damaged.
And this is not even remotely close to an isolated failure. More than 1,357 AI-enabled medical devices are now authorized by the FDA, double the number approved in 2022. The agency has since received reports involving dozens of other AI-enhanced tools, including a heart monitor that allegedly failed to detect abnormal heartbeats, an ultrasound device that misidentified fetal body parts, and an AI-enhanced stethoscope that incorrectly flagged heart failure in nearly two-thirds of more than 12,000 patients. At the same time, large language model chatbots designed to provide medical guidance have demonstrated widespread hallucinations, while experts warned that clinicians are already losing the ability to independently identify cancers in scans due to growing reliance on AI-powered detection systems.
And while I could spend this entire video demonstrating all the ways that AI has failed to deliver on its promises, let's instead discuss all the ways that AI has actually succeeded. Artificial intelligence has successfully become the most effective blanket justification for the mass replacement of American jobs. Companies like Amazon cut roughly 14,000 employees in 2025 and announced plans to eliminate another 16,000 roles attributed to AI, with senior vice president Beth Kletti calling it "the most transformative technology we've seen since the internet." At the same time, Amazon ranked numbers 1, 11, 24, and 49 among the top 100 companies approved for H-1B workers, totaling 14,365 new hires, more than the number of official layoffs in 2025, and is already on pace to exceed that figure this year. Microsoft made roughly 6,000 H-1B hires during the same period it cut 10,000 American workers. Meta, after publicly restructuring around AI, quietly replaced its assistant M with a human-assisted version monitored by overseas contractors. Amazon even operates an entire division called Mechanical Turk, where thousands of freelancers train and operate AI systems through direct human labor, a name borrowed from the 18th-century chess machine that appeared automated while secretly being controlled by a man hidden inside.
And in this pursuit of supposed AI-driven productivity, more than 1 million workers have been laid off, with October marking the worst month for job cuts since 2003, and 2026 beginning with the highest levels of layoffs since the Great Recession. Which brings us to the second and arguably more pressing victory for artificial intelligence: its successful capture of federal policymaking. According to fiscal data from the US Treasury, the federal deficit has already reached $697 billion and is projected to grow to $1.85 trillion this year. National debt is now expected to rise from roughly $34 trillion to over $55 trillion by 2035, a 62% increase, marking the highest level of sovereign debt ever recorded.
And with productivity growth stagnating, manufacturing hollowed out, the dollar weakening, and household consumption increasingly sustained by credit, the only remaining way for the United States to maintain the appearance of global dominance is through the illusion of growth. And right now, that illusion is being carried almost entirely by artificial intelligence. A September 2025 analysis from Deutsche Bank warned that without AI-related investment, the US economy would already be in a technical recession. For the first time in American history, corporate spending on AI has outpaced consumer spending as a primary driver of GDP growth. Even Elon Musk admitted on the Joe Rogan Experience that the only way to get us out of this debt crisis and prevent America from going bankrupt is AI and robotics, adding that if artificial intelligence fails to at least double national productivity, "we're toast."
With that expectation set, unlike the 1980s when the US government cut its losses of taxpayer funds tied to its early AI ambitions, policymakers today are going all-in, betting that artificial intelligence can correct decades of economic mismanagement. In 2024, nearly 75% of all gains in the S&P 500 were driven by just seven companies, all tied directly to artificial intelligence. Nvidia, Microsoft, Apple, Amazon, Google, Meta, and Tesla now make up more than one-third of the entire index, marking the most concentrated share in US market history, with the hope of eventual profitability already projected to require nearly $2 trillion in additional spending by 2030, a sum larger than the combined annual earnings of the entire S&P 500 tech sector.
Meanwhile, Americans are more exposed to this risk than at any point in history. Data from the Federal Reserve shows that direct and indirect stock holdings, including mutual funds and retirement accounts, now represent 45% of all household financial assets, the highest share ever recorded. Meaning that tens of trillions of dollars in savings, 401(k)s, and retirement plans are now tied not to what AI is today, but to what the nation is hoping it will eventually become. Highlighting that if artificial intelligence companies need the government to appear tough on H-1B visas to mask widespread displacement, well, so be it. If those same companies need to quietly walk back their automation claims and announce elective rehiring, then another historic annual job revision will suffice. If powering multi-billion dollar data centers requires more land, more energy, and more infrastructure, the cost can simply be shifted onto local residents. And if those companies ultimately need to be bailed out, then, as always, it will be framed as a necessity funded by taxpayer dollars, proving that while replacing humans with AI is going terribly wrong by our metrics, the only real ramifications will once again be paid by the American workforce.
Like, at some level, when something gets sufficiently huge, whether or not they are on paper, the federal government is kind of the insurer of last resort, as we've seen in various financial crises. And insurance companies screwing things up. So I guess, given the magnitude of what I expect AI economic impact to look like, sort of, I do think the government ends up as, like, the insurer of last resort. And so what I hope that we're beginning to recognize is that this national pursuit of AI actually has far less to do with automating human workers, at least as of right now, and far more to do with providing this massive blanket statement that removes any form of accountability.
What we are witnessing is these individuals that have billions and billions and billions of dollars, the most power that we could imagine. And they're kind of fitting into these two different camps. One of them is saying, "AI is growing at a rate that we couldn't even imagine. It's it's scary, and there's going to be an apocalypse, and there will be no white-collar workers, and you should be scared, and you should, you know, blah, blah, blah. But the way that you could save yourself is by using our product and spending $20 a month." Now, on the other side is you have this thing that I don't know if you guys have have noticed, but it's almost like this new wave of, "There's going to be a utopia. Don't be scared of AI. Everything's going to work out great. There's going to be so much prosperity that, like Elon Musk said, you don't even need to save for retirement. Why put any money into your 401(k)? Just spend the money now. It's going to be perfect."
And if it's not, then it's going to be horrible. And so what that ultimately does is that you either say, "Okay, I'll give you my money and I'll use your product," or "I'll give you every ounce of power and autonomy and privacy that I have because I guess you have the ability to give me prosperity." At the end of the day, it isn't designed for us. Well, you can recognize within your own life, with every action that we do, there are consequences. You can decide, ultimately, "I'm still going to do it or I'm not going to." If I eat this massive bag of candy and cookies and cake and I drink soda and all these different things, I most likely will have a stomach ache. If I go to bed early, I will most likely feel better in the morning. If I go and work out, I will most likely feel more healthy.
In an entrepreneurial sense, if I build X, it's going to help solve Y. If I build X, it's going to make Y even faster, even better. There's an end result in mind. I have yet to hear of a collective decisive understanding of what AI is actually supposed to do for us. These CEOs and public figures have no idea. You even have Sam Altman saying that whenever ChatGPT comes to, you know, god tier or whatever that it is, they're going to ask ChatGPT, "How do we make this a profitable business model?" That's like the most insane thing that I have ever heard.
What I'm really just kind of trying to do is I I I'm really just trying to put into perspective that if you do not have the end goal in mind, that it is unbelievably obvious. If you are not prioritizing making homes affordable right now, improving the dollar right now, allowing more small businesses to thrive right now, giving college graduates entry-level jobs, making the job market more robust, providing more fair wages, you know, all of these many things. If you aren't doing that to fix it right now, you aren't actually building the bridge with artificial intelligence to get us to that destination. You are building it to provide, you know, an improved bottom line to have investors thrive. It's going to come at the cost of your autonomy and your power and your privacy. And at the end of the day, you might get lucky enough and you might have a crumb thrown your way, but I I I'm at least as of right now, maybe I'm cynical, I'm just not seeing what the real intention of all of this is. And if it doesn't work out, I'm sorry to say it, I see no signs that the bill isn't going to be pushed our way. I think it kind of has to be. The government kind of has to bail it out. And so, yeah, that that makes our lives even worse. And I think they understand that if you paint everything as a state of national emergency, the federal government, as blind as they are and with their pockets being lined as much as they are, you just kind of throw as much money their way as possible. At the end of the day, that's our money.
So I I think that while it is absolutely true AI is not working in the corporate environment right now and there are certainly some use cases that it is thriving in, I really am just kind of trying to provide this this alternative discussion of, we have to begin looking at what is the actual use case for this. What are they actually trying to push for? I think that what really matters is how narratives are being controlled. AI is unfortunately this like blanket statement that not only removes accountability, it also provides distraction and and and it doesn't allow, I think, information to be able to come through.
I personally get it a lot myself. Um, I see a ton of comments that say, "Can't trust this guy. He's just AI." "Another AI-generated video." "More AI slob." "You can tell his video or his background isn't real." Like, this this is real. This is this background. I I am a real person. But when you start having comments like that and tons of people are liking it, or it's just bots, you know, who knows? It still it plants a seed of maybe I can't trust this. And let me adjust this light here real quick. And when people start to think, well, maybe I can't trust what's going on, well, you kind of start to dissociate and you start to remove yourself and you start to to just kind of say, "What is it that I can even do?" I'm just going to give the power to these these corporations and let's just see what happens. We can't let that happen.
I think one of the massive accomplishments that we are seeing is that channels like myself and or channels like mine and and other channels similar, we were discussing how AI AI-driven layoffs actually had nothing to do with AI far before I saw any massive headlines from like CNBC or CBS. I was able to get Dr. Ron Hira, the leading expert on H-1B replacement and American outsourcing. That is a massive accomplishment. Myself and others like me have been able to allow millions of other people to get a a sense of what is actually going on. And I think that's what's allowed us to control the narrative more than I think they thought they'd be able to. I think that's why it's kind of having to be rolled back a little bit quicker than they might have imagined. And so if they keep trying to pivot, or we start having these these products pushed our way with it marketed as "AI not included" or, you know, "Enjoy this coffee mug not made by AI." Well, they market it up. You want it more because it it seems authentic and it seems real. It's the same thing that we have right now. We have to make sure that we say, "You have to stop pushing this product. I don't want to Google something and every time it's some sort of like not at all what I looked up." I I just have yet to actually experience a platform that I regularly use that AI has now been forced onto and it actually works. It just it just doesn't. And so I think that we have to start holding these these platforms accountable.
And so ultimately, I hope that this video was able to provide some value to you. Um, I'm really just trying to do the best that I can. I'm really trying to to just paint the story that I think that we need to hear and maybe we don't want to hear or isn't actively being discussed. And I really hope that it's providing some value. Um, if if this channel is providing value to you, I would really appreciate if you like the video and subscribed if you haven't already done so. It really helps me get the videos out in front of the individuals that really want to make a better tomorrow. I'm actively doing everything that I can to not only allow us to have a more prosperous future, but, you know, I'm really trying to do things in, in I'm really trying to plan for future generations as well. I think that we have to start having that happen and and I'm going to try to lead that in whatever way that I can. So, again, thank you all so much for your time. I hope that this was able to to give you a little something that maybe you didn't already have before. And uh, I look forward to speaking with you next time. Bye-bye.