Artificial intelligence promised to make everything cheaper, faster, and more efficient. Instead, it’s becoming a silent financial drain that’s hitting consumers and businesses harder than anyone anticipated. While we celebrate ChatGPT’s latest capabilities and marvel at autonomous agents, the economic reality is stark: AI is making everything more expensive, and most people don’t even realize it’s happening.
This isn’t just about the $20 monthly subscription to OpenAI or Claude. We’re witnessing a fundamental shift in how technology costs flow through the economy, creating what amounts to an invisible tax on digital life.
The Subscription Trap: Death by a Thousand Cuts
The most obvious AI cost explosion is happening in software subscriptions. Every tool you use is suddenly “AI-powered” and commands premium pricing. GitHub Copilot, Notion AI, Grammarly Premium – the list grows monthly. What started as occasional $10-20 add-ons has morphed into a subscription avalanche.
One developer’s experience illustrates this perfectly:
“Five months ago, a developer was spending hundreds every month on AI tools while building a SaaS and running agentic workflows. Then he moved a large part of the workload onto a Mac Mini. The result? $459/month became roughly $3. A $599 machine replaced years of recurring AI costs.” — @0xkerazcity
This mirrors the mainframe-to-PC revolution of the 1980s, when smart companies realized they could buy computing power once instead of renting it forever. The difference? Most consumers and small businesses haven’t figured this out yet.
The Enterprise Reality Check
Large corporations are hitting AI budget walls faster than anyone predicted. The pattern is becoming clear: initial AI enthusiasm leads to massive overspending, followed by harsh financial reality.
“AI costs are exploding as companies rethink the hype. Uber burned its full 2026 AI budget in just 4 months. Firms are now chasing cheaper models, usage caps, and strict ROI while OpenAI gears up for IPO.” — @MrStealth27
This corporate scramble has real consequences for consumers. When Uber burns through its AI budget, those costs get passed down through higher ride prices. When retailers overspend on AI customer service, product prices rise to compensate.
The Infrastructure Arms Race
Behind the scenes, an even more expensive battle is raging. AI requires massive computational infrastructure, and the bottlenecks are creating artificial scarcity that drives prices skyward.
The real constraint isn’t processing power – it’s high-bandwidth memory (HBM). As one industry observer noted:
“memory being the constraint is the real shift here, not the fab count. the bottleneck for AI is HBM, the high-bandwidth memory stacked next to the accelerators, and it is sold out years forward.” — @fromthearena1
This creates a supply chain choke point reminiscent of the semiconductor shortage of 2021-2022. When essential components are scarce, every company using AI pays premium prices, and those costs flow directly to consumers.
Four Ways AI is Emptying Your Wallet
The AI tax manifests in several distinct ways:
- Subscription Inflation: Every software tool now has an “AI tier” that costs 2-3x more than basic versions
- Hidden Processing Fees: Cloud services embed AI compute costs into seemingly unrelated products
- Premium Product Positioning: Companies use “AI-powered” as justification for higher prices across all categories
- Infrastructure Pass-Through: Data center costs and memory shortages increase prices for any digital service
The Historical Parallel: The Internet Bubble’s Lessons
This situation echoes the dot-com boom of 1999-2000, when companies burned through massive capital building internet infrastructure. The Shiller P/E ratio then hit similar levels to today’s 40.88, and the market correction that followed was brutal but necessary.
The key difference: internet infrastructure eventually became commoditized and cheap. Whether AI follows the same path depends largely on how quickly the memory bottleneck resolves and whether open-source alternatives can break the current oligopoly of AI providers.
The Smart Money’s Strategy
While most consumers keep adding AI subscriptions, a small group is taking a different approach. They’re buying local compute power once and using it for years, rather than renting AI capabilities monthly.
This strategy requires technical knowledge most people lack, but it points toward a future where AI costs could plummet – if you know how to navigate the transition.
Looking Forward: The Reckoning Approaches
The current AI spending bubble is unsustainable. Companies are starting to demand real ROI from AI investments, and consumers are beginning to question whether they need 12 different AI subscriptions.
The correction is coming. The question isn’t whether AI costs will normalize, but how painful the adjustment will be and who will bear the financial burden. History suggests that early adopters who bought into the hype will pay the highest price, while those who waited for mature, efficient solutions will benefit from the eventual cost reduction.
The AI revolution is real, but so is the AI tax. Understanding both is essential for navigating the next phase of the digital economy.
Published in Stream · Dispatch #424 · June 6, 2026 · 4 min read.
Reply to paolo@mont3.ch - every email gets a human answer within 24h.