The AI Bubble Should Have Never Existed In The First Place

But the reason it does will piss you off.

Photo by SIMON LEE on Unsplash

The AI Bubble Should Have Never Existed In The First Place

But the reason it does will piss you off.

So much money has been poured into AI that we have effectively bet the entire economy on its success. That is a huge problem. Not just because AI is a fundamentally flawed technology that will never be profitable, never live up to the speculation, and is creating a debt bomb on a scale never seen before. No, you see, even if I am wrong and the AI investors are right, we are all still screwed, as AI will take a huge number of jobs and crash the economy from the bottom up. It is a moronic lose-lose situation. But it didn’t have to be this way. Hell, it should never have been this way. If companies had just taken an iota more responsibility, the ticking economic time bomb that is the AI bubble would have never happened, and you would be exponentially better off. Let me explain.

Why does the AI bubble exist? Not the surface speculation, but the actual structural forces behind it. Well, there are four simple yet devastating reasons.

Valuation

The primary driver of the AI bubble is Big Tech, such as Google and Microsoft. These are some of the most valuable companies on the planet. But what makes them so valuable?

It isn’t their profits. For example, Alphabet (Google’s parent company) made just over $100 billion in profit in 2024 and was worth $2.36 trillion at the time. Meanwhile, Saudi Aramco, an oil giant, made slightly more profit in 2024 but was only valued at $1.6 trillion. In other words, investors value each dollar of profit Google makes nearly 50% higher than each dollar Aramco makes. Why?

Ultimately, the big driver of valuation in Big Tech isn’t profit, but the potential for future growth from radical innovation. After all, Aramco is limited by how much oil they can drill up and how quickly we can burn it all and destroy the planet. But Google’s domain is cyberspace, which means it is boundless. Right?

Well, if you look at Big Tech’s profits, they soared in the late noughties but only grew marginally after that. They have spiked since 2020, not from radical, market-changing innovation, but rather from tax reforms that reduce how much revenue they can hide offshore and from tightly squeezing their consumers. Indeed, user sentiment toward Big Tech is at an all-time low because of this. In other words, Big Tech’s real-world growth has only been marginal for nearly two decades, and they are starting to push up against market limits, with no innovative growth in sight.

But, for Big Tech to keep its sky-high valuations, it needs to be seen as being in the process of delivering radical innovation, especially after a decade of comparatively slow growth. Otherwise, investors will realise they have butted up against the market limits, and they will value them similarly to Aramco, wiping hundreds of billions of dollars off their value.

In short, going into the 2020s, Big Tech needed to be seen as delivering on the market-altering innovations investors had bet on. Guess what filled that spot?

The Wealth Build-Up

During that decade of comparatively slow growth, Big Tech hoarded a ton of cash for no other reason than tax avoidance.

By exploiting gaps in international tax laws and tax havens where corporate tax is effectively zero, Big Tech avoided hundreds of billions of dollars in tax. By 2015, Big Tech had amassed over $1.5 trillion in these tax havens, which is roughly the same as the entire Canadian economy. That figure grew significantly from there, so by the start of the 2020s, Big Tech had literal trillions of dollars on hand.

But they couldn’t touch it, because as soon as this money was transferred out, it would be taxed. And what were they going to do? Contribute to the society that made them so damn wealthy? No, they would rather hoard away more money than a first-world country’s economy.

This set the stage. An ever-increasing pressure to be seen as delivering market-altering innovation and enough wealth to make Smaug look destitute. All that was needed was a way to combine the two.

The Release

Trump, in his infinite Big Mac-soaked wisdom, implemented the 2017 Tax Cuts and Jobs Act (TCJA), which enabled Big Tech to access its immense cash reserves.

The TCJA had a one-time “deemed repatriation” tax on all existing foreign earnings for the 2019 tax year. In other words, Big Tech could move its foreign-held wealth into the US and only pay a tiny amount of tax on it. Suddenly, Big Tech could access its trillions of dollars.

This led to a gargantuan return of capital to the US, which initially sounds like a good thing. But it is the market pressures on that capital which made this a truly terrible move. After all, it had effectively just handed trillions of dollars to a sector that was being squeezed and desperately needed to be seen as innovative to justify their valuations. So, guess where they shoved all this cash?

The Bubble

In 2017, Google invented the technology behind Large Language Model AIs and established the generative AI movement, with OpenAI adopting the technology soon after. All of this was just in time for this meeting of market forces.

In 2019, Big Tech used its newfound offshore cash to perform share buybacks and pay down debts, which will be important in a second. But it also enabled them to shove a ton of money into AI to look innovative, which got the bubble going.

For example, in 2019, Microsoft invested $1 billion into OpenAI and then made several more multi-billion-dollar investments in OpenAI in 2020 and 2021. OpenAI used this cash to provide more data and computational power for the technology Google had pioneered and created ChatGPT, which launched in 2022. The first ChatGPT was seen as an exponential leap forward, despite the fact that it was just a very expensive sausage shoved into an old and limited sausage machine.

Because this AI was perceived as groundbreaking, rapidly developing, and innovative to the point of being a new industrial revolution, it fit the narrative Big Tech needed to justify their high valuations. The fact that, in reality, it was none of those things and was wildly unprofitable didn’t matter. So, they shoved more of their money into it, with Microsoft ploughing $13 billion into OpenAI in 2023 and making repeated investments since then.

This created a snowball effect as other investors joined. FOMO dragged all the Big Tech corporations into the AI bubble, as they needed to justify their high valuations, too. But when a sector with high speculative value that makes up the majority of the US economy puts all its development cash in one focused direction, it can make very little sense to bet against that. Sometimes, it is safer to go with the flow. As such, the majority of global venture capital pivoted in the same direction and backed AI.

This sudden influx of cash took AI from a speculative investment into a bubble. But it also kick-started the circular financing we have seen, as it shifted AI valuation not based on results but based on capital expenditure in AI infrastructure, since this was seen as a predictor of who was advancing their AI the fastest. As such, these interdependent AI companies, like Microsoft, Nvidia and OpenAI, used this new cash and higher stock ownership (thanks to all those buybacks) to invest in each other. This would enable cheaper expansion — for example, OpenAI receives cheaper chips from Nvidia — but the circular nature can also dramatically increase their values by passing the same cash around each other, buying smaller and smaller parts of each other every time they do so.

Sure, this creates growth on paper and not in the real world, but that doesn’t matter. This is all about appearing disruptive and innovative, and these circular deals achieve exactly that.

However, this wasn’t enough to keep the AI train running. Remember, it is wildly unprofitable, and as AI scales up, it actually becomes even less profitable. Not only that, but to keep this newfound speculation alive, AI companies had to announce larger and larger planned capital expenditures to obfuscate the fact that the current models are nowhere near good enough to meet the hype. They needed more cash.

But Big Tech had paid down a ton of debt and owned more of its stock due to being able to access its offshore cash relatively tax-free. So, lenders saw them as safer borrowers than before. This enabled Big Tech to raise an insane amount of debt throughout the past year to keep this bubble going. So much so that lenders now view them as risky and incredibly overleveraged. In fact, these lenders are seemingly starting to back out and could effectively cut off this stream of debt that is keeping the AI bubble afloat. If they do, it will pop the bubble and crash the economy, as this debt is tied to every aspect of our financial system (read more here and here).

The Solution

So, market pressure for Big Tech to deliver on innovations it simply couldn’t, combined with two decades of Big Tech’s enormous wealth hoarding through tax avoidance and an ill-thought-out tax amnesty, is what caused AI to go from a small fringe technology to an economy-crushing bubble.

Two simple things could have prevented all of this: a global minimum corporation tax and a 2% wealth tax.

A global minimum corporation tax would render tax havens unviable, making it more tax efficient for corporations like those in Big Tech to keep profits onshore. This not only ensures they are paying an appropriate amount of tax on their profits but also makes their business practices more transparent, enabling better and more efficient tax administration by the government.

Had this been in place, Big Tech would not have hoarded its wealth; instead, it would have been incentivised to invest in genuine growth in its domestic market. They could have done this through stock buybacks, talent acquisition, or even paying their workers more, and in turn growing their domestic economy, which would enable them to grow, too. But such a tax wasn’t in place, which is disappointing when the majority of voters agree with it. A 2021 survey found that 58% of Americans supported a global minimum corporate tax rate of 15%, with support crossing party lines.

But there is still an incentive for Big Tech executives to focus on their company’s value over everything else. You see, rather than taking a salary or income, they instead get paid in more shares, or their shares increase in value. This is because using the “buy, borrow and die” strategy, where they use their stocks as collateral for loans that function as their income, they can get paid without paying a single penny in tax, as loans and share holdings are not taxed. They also don’t plan on paying back the loans, so they just default on them when they die.

Virtually all the Big Tech CEOs do this. Unfortunately, that means they need their shares to keep dramatically increasing in value, not just to get more loans, but to service the interest on loans they have already taken out. This incentivises them to focus on significantly high-risk short-term gains, like the AI bubble, rather than legitimate long-term growth, such as growing the economy they are giants within.

By taxing wealth — for example, taxing assets over $50 million at a 2% annual rate — we could have prevented this mess. Such a tax would mean income from asset growth would be taxed in line with income tax, giving Big Tech and other CEOs no option but to pay their fair share. This would incentivise them to focus on long-term growth, as it would make sense for them to be paid more in salary, and the company needs more tangible growth to be able to do that.

But most wealth taxes were wiped out in the 20th century thanks to neoliberalism. Consequently, nothing was in place, so all the Big Tech CEOs saw how the AI boom could temporarily jack up their share value, enabling them to get even more filthy rich by taking out enormous loans against them. It doesn’t matter if the bubble bursts and millions are left destitute — they don’t plan on paying off these loans anyway!

Thankfully, support for such a wealth tax has spiked in recent years. A 2019 poll by Fox News, of all places, found that 68% of registered voters favoured a 2% annual wealth tax on individuals with a net worth over $50 million.

These two forms of tax should have existed all along. They would ensure the wealthiest don’t hoover up all the gains and that the working and middle classes stay economically relevant. If you are reading this from a Western country, these taxes would have made you happier and wealthier and would have enabled a speedier economic recovery in the aftermath of 2008. In other words, the very reason you have been struggling more and more each year for the past few decades, while the wealthiest get even more wealth, is because taxes like these weren’t in place.

Now, there is another side to tax reform. It is all good taxing the rich and corporations to prevent them from destroying the economy through idiotic speculative investing. But this tax revenue needs to go towards paying for services the people need and investing in the people. If it instead goes towards corporate socialism and continues to make the rich richer and the poor poorer, then we are just solving one problem by creating another.

Really, tax needs not just reform but a rebrand. It needs to be seen and treated as a collective investment and collective ownership for the betterment of the people. By supporting an economy from the base up like this, we can not only create a far more equitable society but also one that actually delivers substantial, tangible, consistent and sustained growth, rather than a revolving door of damaging speculative bubbles.

Personally, I’d argue this is impossible in today’s hyper-neoliberal capitalist society, where democracy and politics are inherently subservient to capital’s growth, not the other way around. The entire system is designed to put profit before people, and that ingrained structural bias needs to be completely flipped before tax reform can be implemented. Otherwise, all the funds will go to the richest 1% and help them grow their capital even more, while they watch the people suffer, as we have seen for the past few decades.

Summary

The AI bubble actually has very little to do with AI technology. Instead, it is a reflection of our economy, tax policy and politics’ structural bias towards capital and its growth, no matter how unreal it is, rather than people. It is also a reflection of the motivations of the wealthiest and most powerful people in the Western world. It is the reason why so many people now hate these billionaires, as they have perfectly demonstrated that they are lucky, greedy morons who can’t see past their own inadequacy.

There is a deep lesson here that needs to be learnt. A subtle but fundamental change in our society must take place to not only recover from this bubble when it inevitably pops but also to prevent it from happening again. More and more people are recognising that fact every day, as we have seen with the majority support for things like a wealth tax. The question is, will the politicians listen to the democratic outcry coming their way, or will they just cosy up to the billionaires again?

---

Thanks for reading! Don’t forget to check out my YouTube channel for more from me, or Subscribe. Oh, and don’t forget to hit the share button below to get the word out!

---

Sources: Reuters, On The Margin, BHRC, Duke Fuqua, Fair Tax, The Guardian, OECD, YouGov, PEW Research Centre, S&P Global