I'm a recent high school graduate in a rural area dealing with high gas prices and slow satellite internet. I need a reliable computer to job hunt remotely, learn skills, and participate in today's economy. But exploding RAM prices are making that extremely difficult.

A standard 32GB DDR5 memory kit that cost around $90 not long ago now sells for $350–$600 or more at retail. This isn't just painful for gamers and PC builders — it's pricing out students, rural workers, and everyday people from basic technology. When essential computer parts become luxuries, it creates a bigger divide and locks many out of modern opportunities.

The core issue is clear: Extremely high demand from AI data centers is consuming a massive share of global memory production capacity (especially advanced wafers for high-bandwidth memory). At the same time, supply for consumer-grade DRAM is severely constrained. Three companies — Samsung, SK Hynix, and Micron — dominate the DRAM market with a combined share often exceeding 90%. This concentration, combined with the enormous costs and timelines of building new fabs (tens of billions of dollars and several years per facility), makes the market slow to respond.

There are also real legal and contractual complexities between AI firms and memory producers, along with significant lobbying influence in Washington. These issues add uncertainty, but they don't change the fundamental economics: we have a supply/demand imbalance that hurts regular consumers.Blaming the FTC or punishing existing companies isn't the answer. The smarter, long-term solution is to encourage genuine competition and expand supply.

Governments should:

• Reduce over-regulation and streamline lengthy permitting, environmental reviews, and bureaucratic processes that delay new semiconductor fabs and expansions (while maintaining legitimate safety and environmental standards).
• Lower barriers so new companies and manufacturers can more easily enter the market or scale up production of affordable memory modules.
• Support predictable policies that attract investment in domestic and allied manufacturing capacity.

Positive developments like Google's TurboQuant algorithm — which can compress the memory needed for AI models by up to 6x with no loss in quality — show that software and efficiency improvements can help stretch existing supply. But we still need more hardware production overall.

High prices for RAM, GPUs, and CPUs are making technology less accessible at the worst possible time. Sign this petition if you support breaking down barriers to competition so more companies can produce memory chips and help bring prices back to reasonable levels for everyone — not just big tech and the wealthy.