May 19, 2026
Cash burn meets comment inferno
AI is too expensive
AI’s money bonfire has commenters split between bubble panic and ‘just wait’ cope
TLDR: The article says big tech is spending mind-bending amounts on AI without proving the money will ever come back, with Microsoft’s reported OpenAI bill becoming the jaw-dropper. Commenters are split between calling it a coming bait-and-switch bubble and insisting this is just the messy, expensive beginning of something huge.
The article’s big accusation is brutally simple: today’s AI boom may be wildly expensive and nowhere near paying for itself. The writer argues that giant companies have poured staggering sums into data centers, chips, and partnerships while staying suspiciously vague about how much money AI is actually bringing in. The flashpoint is Microsoft, which reportedly spent around $100 billion on its OpenAI relationship and related infrastructure — a number that made commenters instantly reach for the popcorn.
And oh, the crowd had opinions. One camp went full doom mode, warning that users are living through the “bait” phase before the inevitable “switch” — higher prices, worse deals, and maybe ads shoved into everything. Another commenter predicted the real rug pull would be advertising, which is about the least surprising and most depressing internet prophecy imaginable. The vibe? People are deeply suspicious that today’s shiny AI freebies will become tomorrow’s expensive dependency.
But the thread was far from a one-sided roast. Critics of the article pushed back hard, basically saying: so AI is doomed if it’s expensive, doomed if it’s cheap, and doomed if it succeeds? Others compared it to Uber’s early money-losing years, arguing that if AI really does automate a huge amount of office work, the eventual payoff could still be massive. The wildest comment compared the whole race to owning nukes: maybe profits don’t matter if everyone thinks the tech is too powerful to miss. In other words, the numbers sparked a classic internet brawl: bubble, empire, scam, destiny — pick your fighter.
Key Points
- •The article argues that AI is currently not economically viable for most companies other than firms benefiting from hardware and data center buildouts.
- •It says hyperscalers have invested over $800 billion in the last three years and describes projected additional spending of roughly $700 billion in 2026 and $1 trillion in 2027.
- •The article compares those investment levels with the combined recent annual revenues of Microsoft, Meta, Amazon, and Google, which it lists as $1.599 trillion.
- •Citing Bloomberg and testimony from Microsoft executive Michael Wetter, the article says Microsoft has spent about $100 billion on its OpenAI partnership.
- •The article states Microsoft’s capex since fiscal year 2023 reached $293.8 billion and uses OpenAI capacity and Copilot subscriber figures to question whether AI revenue can justify current investment.