Is AI Spreading Like a Virus? Reddit Essay Questions the $700B Datacenter Buildout
thepotatobake · reddit · 2026-10-11
A Reddit essay argues AI is spreading more like a virus than a normal technology — expanding by exploiting host behavior rather than organic demand.
Core argument
- Companies and countries adopt AI because competitors do; regulators lag; AI is bundled into software people already use; once organizations depend on it, it's hard to remove.
- Frontier labs now use their own models to write code, run experiments, and plan — feeding the next model and infrastructure. More decisions about AI are being made with AI's help. The author admits no proof, but notes no one can disprove it either, since evidence is private.
Accelerants
- Lab races and geopolitical competition; US Big Tech expected to spend $700B this year, mostly on datacenters.
- Chipmakers investing in their own customers, who use that money to buy more chips — inflating apparent demand. Most demand forecasts come from companies profiting from the buildout.
- Executives use AI tools to write business cases, and these tools tend to agree with the asker.
Ask: The author can find no public measure of actual capacity utilization, who pays, or how much goes into training the next generation — reported figures mostly cover leased capacity. Borrowing public-health's "surveillance before proof" approach, he calls for pandemic-style monitoring of AI infrastructure and asks if such work exists. Disclosure: he ironically used AI to search for supporting evidence and came up empty.
More from AGI Musings
- 228 tasks, 26 models: METR's time horizon curve is nearly flat from 2-30 min — lulzxdxdxd · 2026-10-11
- AI Code Was Dismissed as Slop a Year Ago; Now It May Redefine Good Software — kevinnbass · 2026-10-11
- Complexity scientist Melanie Mitchell: "superintelligence" isn't yet a scientific concept — khademinori · 2026-10-11
- "Censoring AI is step one": a warning about the surveillance slippery slope — PierceLilholt · 2026-10-11
- Togelius: commenters keep telling pros upset by AI progress to accept obsolescence — kchonyc · 2026-10-11
- Eric Schmidt warns AI could develop its own language within 5 years that humans won't understand — 2C_ornot2C · 2026-10-11