The tech industry is currently pouring billions into building massive server farms to support cloud-based generative AI. This rush has created a severe shortage of general-purpose memory components, directly impacting the cost of everyday electronics like your Windows PC, smartphone, or gaming console.
Analysts and critics are now questioning whether this massive infrastructure spend is sustainable, labeling it a potential “trillion-dollar hallucination.” As venture capital funds chase returns, the hidden costs are being passed down to consumers through higher device prices and reduced specifications.
The hidden cost of the AI buildout
Memory suppliers have redirected manufacturing capacity toward high-value components for AI servers, such as advanced-layer 3D NAND. This shift has left little room for the standard RAM found in consumer devices. According to TrendForce, suppliers are failing to invest in additional general capacity, exacerbating the shortage.
Gartner predicts this shortage will cause global PC shipments to drop by 10.4% in 2026 and smartphone shipments to decline by 8.4%. Furthermore, prices for these devices are expected to rise significantly compared to 2025 levels, with PCs seeing a 17% increase and smartphones rising by 13%.
The AI consumer tech tax
Electronics manufacturers are now competing for limited supply, forcing them to raise prices or reduce features. Sony has already increased the price of the PS5 by $100, while Microsoft and Nintendo have also raised prices on their respective consoles. Samsung has quietly increased prices across its Galaxy lineup of smartphones, tablets, and laptops.
Apple is not immune to these pressures. The iPhone 18 Pro is tipped for a significant price increase in 2026, potentially adding $100 to $150 before tariffs are considered. CEO Tim Cook has warned that demand for memory components will directly impact pricing. Additionally, analysts warn of “shrinkflation” in tech, where devices may ship with downgraded displays or reduced performance to maintain familiar price points.
A business model under pressure
Critics argue that the current AI economics are broken. Writer Ed Zitron points out that at $200 a month, a user can burn through $8,000 in Anthropic tokens or $14,000 in OpenAI tokens. This scale of subsidy suggests that the actual value of AI may be inflated, creating a bubble ready to burst once market opinion shifts.
Major tech companies are already feeling the strain. Meta has imposed strict limits on internal token usage after finding it was on track to spend billions on internal AI alone in 2026. Two large banks reportedly spent $1 billion on AI experiments without seeing significant returns. Even Cisco executives have stated that AI token costs are far higher than the value they generate at scale.
Apple’s different approach
While rivals build massive data centers, Apple is focusing on edge AI as a privacy-preserving and cost-saving alternative. By working with Google to distill larger models into smaller ones capable of running locally, Apple aims to reduce reliance on cloud services. Siri AI can now hold conversations and access user context directly on the device.
This strategy allows Apple to offer powerful AI features without the ongoing token costs associated with cloud-based LLMs. The company’s Foundation Models framework also enables app developers to run local LLMs using MLX Distributed, further promoting on-device processing. This shift could eventually reduce the pressure on global memory supplies if cloud dependency decreases.
What this means for you
If you are planning to upgrade your PC, phone, or console in the near future, expect higher prices. The memory shortage driven by AI server demand is likely to persist throughout 2026, with TrendForce predicting up to 75% increases on top of recent spikes. Suppliers are hesitant to ramp up general production, fearing unused capacity if the AI bubble bursts.
For Windows users, this means that new devices may cost more or offer less storage and RAM for the same price. The transition toward on-device AI could eventually alleviate some of this pressure, but until then, consumers are footing the bill for the industry’s massive infrastructure bets.
Source: Computerworld
Over to you: Will you delay upgrading your PC or phone to avoid these predicted price hikes, or do you need a new device regardless of the cost?
