The Great AI Disappointment

Americans’ feelings about AI are clear and the tech isn't living up to the promises or the hype. While the U.S. is damaging its reputation with the world and its own consumers...

Where did we go wrong? Image context: Rural Michigan residents rally against the $7 billion Stargate data center on December 1, 2025, planned on Southeast Michigan farm land. Credit: Jim West/UCG/Universal Images Group.

If you think this is a story worth sharing, please spread the word. We need to do better. This is not the right future for our descendents.

If you can do me the favor of sharing this article with a friend, family member, coworker or acquaintance, it would mean the world to me. Or restacking it or sharing it elsewhere.

In this document I provide my Great AI Disappointment thesis.

Listen to the Essay 🎧: 25 minutes, 37.

I’ll be honest I’m not the most persuasive storyteller, I’m just a blogger and a relic of an internet that no longer exists. A human led internet that may never exist again, the way things are going. But as I get a bit older and after writing and curating hundreds of articles about AI, I’m noticing a shift in the world with regards to Generative AI specifically. You can read reports, but the truth is somewhere in the silent consensus beneath the hype. The costs of the over-investment in this AI cycle are beginning to gnaw at the social fabric, rule of law and the social order. The way things were has not caught up to the new reality. The internet is demoralized under the weight of all of the AI slop. But there’s nowhere to go, this labor market is punishing. And you have to begin to wonder if it’s all by careful design. There’s a financial fragility to the madness and a cost to real people. It’s disappointing to witness. AI is now synonymous with greed and a deterioration of the social contract we used to take for granted and believed would always exist. Let me try and explain in this essay: put plainly, AI hardware should not be put ahead of the well-being of people.

We are nearly four years since ChatGPT first went live. If you would have told me that AI would create a more fractured internet, with a more hollow and disappointing experience, hurting human culture, literacy (we read less), a force for white collar deskilling, and detrimental to human well-being with questionable economic benefits, I probably wouldn’t have believed you. But there’s a weight in pretending and a nagging feeling that this isn't it. And people are upset, not only has AI not reached its promised potential - it’s increasing inflation and comes with a rising cost in economic, psychological and even geopolitical and sociological terms difficult to quantify. With all the Billions of dollars in funding OpenAI, Venture Capital and BigTech marketing, lobbying and PR, we have been led astray. It’s beginning to harm our collective future. Generative AI appears fraudulent and its impact appears to be harmful at scale.

Generative AI as a movement and investment, has been more divisive than unifying or empowering. It’s humiliating and deeply unsettling to live in a world of AI slop and work slop. It’s not just you, we all feel it. It’s not been a good technology. It’s increasing corruption faster than it’s siding with any sense of rule of law, fairness or even demonstrating real growth. The cascade of funding a failed promise could be devastating.

It’s not being adopted or diffusing into society as would a significant general purpose technology (GPT). It’s not building value that’s leading us as a civilization, as a species, to a more meaningful or better future. It’s important to say this aloud. I take no great pride in trying to frame this realistically. Generative AI isn’t what we were promised when ChatGPT suddenly appeared. As models became larger and more sophisticated and spending increased astronomically we were expecting tangible benefits and that the dangers to society would be demonstrably mitigated, carefully regulated and monopolies given some ground rules. But the detrimental impact of the tech are accelerating faster than the tech can show real value. By the late 2020s, it will be out of control. Americans are not good custodians of this degree of financial and artificial power. But they sure know how to build a technological ponzi scheme to inflate the corporate monopolies of their increasingly dystopian internet. Where’s the “opt-out” button? Where’s the reset switch? Generative AI is leading us astray.

Degrading Culture and Revamping Economic Opportunity

Generative AI however isn’t just harming human culture, demeaning our time or damaging our sense of purpose - it’s benefitting the privileged with key elements of stock manipulation, over the common person in society, degrading the middle class and it’s hurting our digital civilization and online experience with AI slop and greater untold debt (at a time when the U.S. National debt is mounting) and wealth inequality is being accelerated at a scale difficult to reconcile with the supposed benefits. If you hurt the wealth, opportunities, labor and career aspects of the general population, you better be delivering. And, Generative AI is not delivering value at scale or helping most citizens, consumers or businesses.

Even the corporate insistence and vague promises of higher productivity or a sustainable impact on GDP is highly questionable now in comparison to the cost of compute and the expense of building these huge larger language models, the hardware, Fabs and most of all, the circular financing. The size and scope of the super clusters of AI datacenters is getting alarming. How the technology benefits the most powerful countries while hurting the rest of us is alarming. The abuse of power and lack of ROIC has become more serious. After four years we need to take stock and look at the movement objectively and scrutinize the players and the technology involved. The companies at the forefront need to be held accountable. The countries manipulating our future need our collective feedback.

On the global stage, generative AI is significantly increasing the hegemony of the U.S. and China over other nations that is extremely toxic for the global marketplace, world order, rules of the marketplace and detrimental for middle-sized, smaller countries (and their economies) not to mention detrimental for the middle class and potentially many white collar workers. It’s clearly an economically exploitive technology that is a “divide and conquer” mechanism being used by the world’s financial elite and most powerful Nation States for their own agendas and self-interests. Generative AI has dystopian characteristics on more levels and layers than we can possibly name in an article. To say that it’s not good, is one of the great understatements of our times. And this stark realization will become more than just about protesting datacenters in the coming years and decades. This could lead to social upheaval, the crumbling of empires, the storm before a dark age. This isn’t like the industrial revolution, this isn’t a kind of automation that’s making us better. The energy, datacenter infrastructure and fabs required are not worth it. The price is too high. The bottlenecks not just economic. The consequences won’t just be that of corporate consolidation.

The Missing Links

After four years of watching the evolution of AI developments, I have to conclude that not only is a return on investment (ROI) absent, the cost to people, communities and trust in capitalism and democracy is highly concerning and a rising sameness of the narrative vs. the obvious exploitation in the real world is rampant and deceptive. Why, why is this not being talked about in mainstream circles? Instead, we are constantly bombarded about how great AI is and how much funding it requires and how it’s a matter of national importance and national defense competition with national competitors while Venture Capitalist, Tech executives, AI researchers and political leaders and the financial elites of our systems pad their pockets. An internet and on platforms where even VC firms pretending they are an authoritative media companies. That’s not the internet I signed up for in the late 1990s. This isn’t going to be the revolution that these companies had in mind. It’s doubtful now that Generative AI is the next leap in artificial intelligence that the makers of ChatGPT and their peers said it was. It’s technological fraud at a grandiose scale.

When the technological on a fundamental levels isn’t benefiting most people, most communities or making the marketplace and opportunities better. AI isn’t improving our quality of life, on the contrary in most cases it’s diminishing it. AI has hurt the and tightened the labor force, creating massive uncertainty, economic strain and a post ChatGPT world is now the worst time to start a career, go to college, or plan a career path, in decades. It’s a generational abyss. I have written in previous articles about how AI is hurting both our youngest and older more established workers at both ends of the spectrum. As a writer I’m literally judged by how and how much I employ AI in my work. But something doesn’t feel right, is off and is deeply unsatisfying and is unscrupulous in this so-called - “AI revolution.” The narrative the mainstream media is telling is obviously not the entire truth.

After covering the AI space and developments for 5 years I can tell you, it isn’t what BigTech or powerful Venture Capital firms or the Government or the elite controlled Media, or all the lobbying, is claiming it to be. The investment in compute, in models and in a technology with detrimental impacts of Generative AI and its outputs is not positive for society or a great help to how companies are organized. It’s not what we were hoping it would become almost four years ago. And I don’t have much hope that it will suddenly become that. The 2017 paper introducing the world to Transformers didn’t deliver us a tool that would build a better internet or improve the world. Vibe coding, AI coding or token prediction didn’t justify the cost or bring us any salvation - if anything it’s entrenched us in a system and in an internet reaching a level of joyless decadence, addiction and technological loneliness - that is making people miserable, broke, disconnected from each other, and uncertain about what comes next. Businesses are also confused at the lacklustre results of their investment in token spend.

Negative Public Sentiment towards Generative AI is Increasing

It would be fine if this was just my opinion or perspective but it’s not, there’s widespread and growing negative AI sentiment among the public, business owners, among middle to smaller countries, among workers, at small businesses, in public services, among employees at tech companies, among companies who are the customers with high token bills, and not much else.

From Teachers, from parents, from new graduates and students about to graduate. From community leaders and smaller Towns and counties, who are getting exploited by dangerous datacenters being built on their lands. With rising electricity bills and a less healthy internet, as if that were even possible. This isn’t just about wealth distribution, it’s about whether and if Generative AI is a useful or good technology. If it actually does what it says it can do. It’s about the costs vs. the benefits. The stark realization that the tech underlying Generative AI is a failure is starting to dawn on us. But the system cannot afford to backtrack, it’s too financially committed to do so or even consider it. So it continues like a snowball effect of the greatest fraud in tech history.

The public uprising against datacenters on the local level is just one symptom of declining AI sentiment. People aren’t totally dumb or unaware of what is really going on. But if companies like OpenAI make promises and then the technology doesn’t deliver and is actually bad for people and is leading to more geopolitical uncertainty and raising the risks of global warfare, we should be asking questions, taking stands and doing what is necessary. I would go even further. Generative AI is becoming more harmful.

AI isn’t a technology that’s for Americans, but against them. And it’s turning many middle power countries and smaller countries against America. AI is being leveraged at a time when the United States is abusing its power in trade, in foreign relations and in developing and aiding proxies against its cold-war era rivals. While hoping to profiteer from the war and uncertainty. While drowning in national debt, it’s funding autonomous killer machines. If the United States keeps down this path of glorifying AI, to the detriment of everything else, it might find out the hard way this is not how you do business with the world or persuade struggling citizens, consumers and voters in the midst of a generational affordability crisis. Something is deeply broken in all the AI profiteering.

---

👋 Hey there, I’m Mike. Each week I share AI articles at the intersection of tech, business, society and the future. If you want to support the channel or gain full-access to my work, go here. Read Archives | See Substack Notes | Visit our community Chat | Visit Homepage. I’ve been writing about AI since well before ChatGPT came on the scene, and I’m disappointed and horrified with what it’s become.

I try my best to give balanced takes from many perspective on the news and how to keep up with AI. But something is shifting in the AI bubble and it has doomsday qualities for humanity’s future. The scale of the capital involved is getting very dangerous and the kinds of machines and agentic AI systems we’re building without the economic benefits you’d expect are alarming and bring much more uncertainty than help or opportunities.

A Question of Trust

Recommendation systems and algorithms had already polluted our human experiences on digital platforms meant to addict and entrap us. But the generative aspect of AI isn’t healthy. I don’t usually talk like this. If Ukraine is getting involved in the Iran War the United States is escalating its conflict with the Eastern Bloc and weaponizing AI in it military systems in the process. Defense and drone automation technology company Anduril is now in talks to raise funding at about $100 billion valuation and with Palantir widely implicated in Gaza and Iran missile and bomb targeting. AI is not being used for good. AI will and is obviously be used for killer robots, for the development of covert bioweapons, surveillance, and hacking the political systems of other countries with all of this compute. Neither am I convinced the U.S. is a good actor on the global stage of foreign relations, trade and geopolitics any longer. The stance of the U.S. against Canada, Europe, Greenland and even its treatment of Taiwan (forcing TSMC to build more in Arizona) is unacceptable to me.

If OpenAI is just another spoke in the U.S. Pentagon along with Palantir and Anduril I fear that the Great Disappointment of AI can easily become a tragedy in the years ahead. On July 9th, 2026 an agent of OpenAI went rogue and tried to escape, and hacked a partner called Hugging Face. Rep. Ted Lieu, D-Calif., and Rep. Nathaniel Moran, R-Texas, on Thursday introduced a bill called the “AI Kill Switch Act.” But it’s too late, glorying and funding AI compute now means AI capitalism is dominant and U.S. AI Supremacy is a danger to the rest of the world, and I’m not just talking about cybersecurity. A day after the event, on July 10th, OpenAI’s head of trust and saftey Johannes Heidecke left the company. Why do you suppose that is?

All the while Sam Altman is boasting we have entered the singularity. In truth, OpenAI and Sam Altman have been especially disappointing in their conduct, product execution and untrustworthy strategy and post-ChatGPT implementation of AI. Angel investors, Tech executives and Venture Capital networks aren’t the right groups to be directing the future rollout of technology. Obviously the few decision makers in the U.S. have conflicts of interest.

For Military Might (and not for Men)

We’re building capabilities that not only could leave people powerless with less economic opportunities but it could be harmful. Again, this isn’t a good or benevolent technology. The VC-led leadership in the U.S. of the Generative AI boom, isn’t aligning the future of the nation that can lead to good outcomes. AI Supremacy isn’t an end in and of itself. Being the best at a particular technology while betraying allies in trade and foreign affairs and inciting more global wars won’t win you the benefits perhaps you are hoping for. The rise of imperialism and neo-nationalism in the geopolitical and foreign affairs of the U.S. is a path to empire wide decline. And I will blame the venture capitalists in the years ahead.

AI will obviously is and will be used for war and it’s part of the Great AI Disappointment thesis I’m laying out here. It will be used for a few to benefit while the majority is economically hurt by it, and it’s not a future most of us are looking forwards to. That’s not the way democracy or capitalism is supposed to work. America was founded on freedoms and opportunities that are being stripped from us. Technology that was supposed to empower us, is being used against us. Generative AI accelerates dramatically this frightening escalation of exploitation. Americans are leaving the country in record numbers. The Brookings Institution estimated that between 210,000 to 405,000 people did so last year. We have to ask ourselves why. I’m not a journalist, I’m just a curated based blogger, but I believe good media helps you ask the right questions. America’s Great AI Disappointment is real, just ask real people. Not on LinkedIn or Twitter/X or Meta’s Threads but at the corner store, in the local pub, at the gym, and people you bump into at the park a local small business or event or at an informal hobby group meetup or event. They will tell you.

The benefits of AI are being exaggerated while the costs are being silenced. The critics are being shut out. Anything that goes against the propaganda of Venture Capital funds and BigTech’s reach, or even Silicon Valley and Wall Street interests. The SpaceX IPO, that was the first of the Big Three AI startups, raised a record $75 billion on ​the sale of 555.56 million shares. Do you think any of that will benefit American consumers or voters anytime soon?

You cannot crack down on AI slop once it’s been unleashed just like once ChatGPT has literally been pushed at Colleges. College Students are reading and learning less since ChatGPT. Why would you de-skill your white collar workers or the brightest young minds of the next generation? Does any of that make sense to you? Harming the value of a college education in a market with deteriorating entry level prospects is not what I consider the impacts of a good technology. How about you?

Circular Tactics

However it is growing and significant Capex, debt and circular plus vendor financing by the likes of Nvidia, OpenAI and AMD that have really made me lose trust in the ecosystem where capacity is paid for in-advance. The hardware, model training, inference, HBM chip costs are startling to unravel vs. the utility of the technology. It’s not just disappointing, it’s becoming deeply irrational. Moody’s recently warned us in a note that ‘unprecedented’ AI spending threatens credit quality. What could go wrong?

“Software costs little to replicate, yielding fat profit margins and fortress balance sheets. Generative AI, by contrast, demands a vast physical footprint: warehouses crammed with expensive and energy-hungry servers and chips.”

OpenAI is in discussions with Nvidia about a backstop of up to $250 billion that would help fund its plans to lease a new AI data center in Ohio that seems to be very close to Government intervention and a bailout for OpenAI over Oracle’s weak credit rating and execution risk that endangers the entire ecosystem.

The vagueness, lack of transparency about the details and obvious implications are incredibly concerning whereby according to Bloomberg:

• The guarantees from Nvidia would help OpenAI lease a 10GW project that SoftBank is developing in Ohio.

• In total, the project could cost more than $500 billion, the largest data center project ever announced.

• The power for the project is controlled by the US government and funded separately by Japan under a recent trade deal.

• Commerce Secretary Lutnick is reportedly involved in deciding who will get the power.

In just the last week Nvidia and AMD have had a bevy of new partnerships that resemble vendor financing on an entirely new scale involving circular deals involving Nvidia with SK Hynix, AMD with Anthropic, and Nvidia with OpenAI as seen above. If I am Nvidia, AMD or OpenAI I’m greasing the wheels and cheating a system akin to transportation companies owning the railroads and paying for their construction many years in advance, cooking the books in the process. The Creator Kakashii goes into a lot of detail if you are interested. If much of the demand is synthetic, we have to question the viability of the entire roll-out. Something is not right here, the accounting and the way we measure the ROI is flawed and unclear on purpose. The attribution is broken, the ROIC was never there in the first place.

I understand Cloud computing companies are capacity constrained but this is stimulating demand itself of a technology that is being pushed on people and workers who realize it’s a nascent and flawed technology without great application layer and products yet. Outside of Claude Code and Cursor, it’s difficult to find good AI products even after nearly four years after ChatGPT came and went. Generative AI is clearly not capable of doing what it’s claiming to be able to do. Nvidia as a vendor financing intermediary of the Trump Administration is warping how the free market is supposed to work all the while with profit margins near 75% and clearly playing favorites protecting OpenAI (maybe its key customer) from failure. This is not going to end well. Deception at this scale is not rule abiding and the capital involved is cheating taxpayers and benefiting Wall Street, Silicon Valley and the financial elite at the expense of workers, consumers and the majority of businesses. It’s difficult to watch and a lot of people are getting hurt.

With Nvidia backing OpenAI to this extent, all bets are off. America’s AI leadership has to be in question. Eventually, there has to be some accountability with what we are doing with all of this capital, vendor financing, lack of ROIC, disappointing general utility of the AI products and the widespread corruption and abuses of the AI bubble.

After SpaceX, Anthropic and OpenAI go public we’ll have more data about the state of Generative AI. But for the vast majority of the general public, workers and businesses, and citizens - the verdict is already in. If you live outside of the United States your economy is likely worse off already due to the impact of American AI. U.S. led AI Supremacy has a cost that will take many years to ascertain, but the price in freedom and opportunity will be steep and most people won’t experience those gains anytime soon. Four years after ChatGPT, Generative AI has been a great disappointment with heavy costs not only those that are financial, and it’s difficult to see a path forward where this is not the case.

The Extremes of Capex & Investment in Unsustainable Hardware Cycles for Compute Advantages

Morgan Stanley raising its 2027 and 2028 hyperscaler capex forecasts by 9% and 10%, respectively, now modeling $1.2 trillion and $1.4 trillion in annual spending. In the week of July 6th a recent estimate revision from Morgan Stanley (led by analyst Brian Nowak) underscores the unprecedented scale of the ongoing AI infrastructure expansion.

Morgan Stanley's latest report has sharply raised its 2028 capital expenditure forecast for the world's five largest cloud hyperscalers to $1.4 trillion. American has a lot of capital but its disappointing to see how it uses it. The price of credit default swaps on Nvidia bonds will be something to watch.

Suffice to say that this is not a sustainable investment compared to the results and ROI we are seeing in the real world. Cloud or Advertising growth at these select companies is not sufficient cause or reason. The dangers of this cycle of AI are unacceptable and too many people are being left behind. The United States is not regulating the tech sufficiently to protect either consumers or workers or our most vulnerable.

I cannot in good conscience call AI anything but a great disappointment in how centralized and monopolized American capitalism is becoming. A few people and companies have the power to change history and when that occurs it’s not usually a good thing. The technology is not diffusing with great AI products in the way you’d expect to see.

LLMs are not converging to some point of ROIC in the grand scheme of things. Every month and quarter the AI bubble endures, is hurting people while elevating the few above the many. That’s not the system, principles or ideals America was founded upon.

AI in its 2020s version of the Generative AI cycle is deeply disappointing, a failure and an astounding waste of capital at the expense of people in an affordability crisis. I’m becoming extremely skeptical about the movement, and I’m not alone. Studies have shown if you make over $200k a year you have a more positive view of the Generative AI technology, whereas if you fall in the bottom half of the socio-economic spectrum - you are more likely to mistrust AI. I don’t even believe Generative AI is an efficiency multiplier or productivity enhancing tool and you can see evidence of this actually everywhere.

The compute advantage of the U.S. being lacklustre while living in the poverty of a lack of energy infrastructure in the Generative AI hype cycle because of a lack of investment in renewable and sustainable energy technologies because of the defeat of the Environemntalists, is the ultimate irony and legacy of the Trump Administration, that will be little more than a disturbing footnote in history, many years from today.

P.S.: this piece may contain many grammatical errors and typos and will be edited again. But I wanted to get out what’s heavy on my chest and conscience without too much of a delay. It’s the culmination of many days of research, weeks of thinking, ruminating and soul-searching. This can’t be summarized neatly in an infographic. You have got to live it and feel it to sense the AI failure fully.