The biggest tech firms in the world released a deluge of earnings reports this week in an earthquake of a change that can alter the technology industry. The total capital spending of Amazon, Microsoft, Alphabet and Meta amounts to more than 112 billion in the last quarter alone, with the move projected to continue to increase into 2026.
This AI-driven spending spurt took stock markets to unprecedented levels, continuing the hot streak of the Nasdaq, but it also sparked discussions on the issue of sustainability. Co-founder Bill Gates of Microsoft made a sober warning, saying that there would be tons of these investments turning out to be expensive dead ends, rekindling the memories of the dot-com bust.
Breaking Earnings Power AI Ambitions
The performance in the third quarter was a representation of a never-ending need for computing power. The shares of Amazon have shot up to record highs when its cloud computing division, AWS, announced a healthy 20% revenue growth of $33 billion, which is several times higher than the expectations of the analysts.
The company is expanding its data centre, which was called very aggressive by CEO Andy Jassy, and the expansion is planned to be doubled again in 2027, having already doubled the capacity since 2022.
Amazon increased its annual capital spending target to an astronomical 125 billion, indicating even more massive expenses in 2021. E-commerce giant Amazon increased its net income by 38% to an impressive $21.2 billion, with a gain of 9.5 billion in the investment it has made in artificial intelligence startup Anthropic.
The same wave was followed by Microsoft, which revealed 35billion capex, quarterly 5 billion higher than expected-mainly spent on AI data centres. Acknowledging that the company is falling behind demand, finance chief Amy Hood acknowledged that the company had sealed over $400 billion in future sales deals, not including a huge $250 billion obligation by OpenAI.
Today, Alphabet, the parent company of Google, increased its annual capex prognosis to a minimum of 91 billion dollars, with its CEO, Sundar Pichai, stating that AI products are now having 20 times more data than a year ago. The fourth was completed by Meta, which increased its 2025 capital expenditure forecast to $70 billion or greater, almost twice the amount spent last year, in pursuit of its so-called superintelligence, which is capable of outwitting human thought.
These numbers reveal a radical shift: the Magnificent Seven technology giants are placing a bet on AI, and are investing expansions out of strong operating profits of up to just over $109 billion in the last quarter. This boom is self-sufficient, unlike the dot-com boom, whose growth was primarily based on debt, as Federal Reserve Chair Jerome Powell has noted.
Bill Gates Gives a Dumping of Reality
Bill Gates, in an interview, cooled the emotions amid the euphoria. The Microsoft tycoon and philanthropist referred to the latest AI hype as an unquestionable bubble and compared this to the late 90s internet hype.
There are a ton of these investments that will be dead ends, and Gates was right with this declaration, warning that lots of companies are building data centres with prohibitively high electricity costs, just like the me-too companies that squandered capital and failed. He disapproved of hyperbolic analogies to the tulip mania of the 17th century, but agreed, we are in the middle of an AI bubble.
The opinions of Gates are consistent with a recent MIT study that found 95% of organisations make zero returns on their generative AI investments worth $30-40 billion. The majority of pilots fail because they have fragile workflows and a lack of integration; they increase the productivity of the individuals, but do not shift the profit needle. However, Gates is optimistic over the long run: AI has been the most technical thing I have ever seen in my life, with a potential world-changing effect, similar in scope to the long-term influence of the internet.
Nvidia CEO Reps with Optimism of Vicious Cycle
Nvidia CEO Jensen Huang made an aggressive speech at the APEC CEO Summit, allowing nervous investors to conclude that the capex boom is not a hoax; indeed, it is the driver of a self-reinforcing virtuous cycle.
This is why you are witnessing the capex of the world going at such a speed, Huang said, and compared it to the infantile years of the internet. Improved AI models drive the use, and the data of the models are combined to create new models, which will bring more investment in a flywheel that accelerates.
This Nvidia payoff has been out of this world: the chipmaker has reached a market capitalisation of 5 trillion, which was driven by the sales of its H100 and Blackwell processors to power ChatGPT and more. Nvidia is the unquestioned leader in this arms race in AI chip orders and partnerships to the tune of $500 billion and over 100 billion, respectively.
Implications on the AI Era Ahead
By the end of 2025, the AI gold rush will not be declining. Cloud providers have acknowledged that they are capacity-bound, where customers are crying out to get more GPUs to train increasingly larger models. Such innovations as Trainium2 chips by Amazon and the Vertex AI advances by Google, to the events such as the World Series, are examples of practical applications already bearing results.
However, risks loom large. The increasing energy requirements may put strain on the grids, and the regulation is further merciless on the effects of AI on society. This is a high-wire stunt that investors are forced to play; in the short term, there is volatility caused by capex overhang, whereas in the long term, the winners take a decade to establish their dominance.
At least the belief of big tech is unshaken at this point. The cycle is, according to Huang, planned as a radical scheme of the next frontier of computing. It is a trillion-dollar question of the future tech titans that it may end in victory or resonate with the dead ends of Gates.
