Particle.news
Download on the App Store

AI Boom Draws Bubble Warnings as Study Finds Major Polls Can Be Skewed

New evidence exposes a split over whether the AI surge is risky speculation or soundly based.

Overview

  • A Dartmouth study published in PNAS shows small numbers of AI-generated responses—sometimes just 10 to 52—could have flipped outcomes in seven major 2024 U.S. national polls, with current detectors failing to spot the bots and per-response costs around five cents.
  • Alphabet CEO Sundar Pichai called parts of the AI investment rush irrational and said no company is immune to a potential correction, even as he argued Alphabet is positioned to withstand one.
  • JP Morgan’s 2026 outlook contends the AI upswing rests on solid fundamentals, noting Big Tech capex could reach about $500 billion by 2026 and warning investors that underexposure may be the bigger risk.
  • Reporting highlights circular capital flows across the sector—spanning cloud providers, chipmakers and model developers—which analysts say raise concentration and correction risks alongside mounting infrastructure constraints.
  • Enterprise adoption continues to outpace worker training, with PwC reporting low frequent use among employees and Rockwell finding 95% of manufacturers plan AI investments to tackle skills gaps and quality control, as executives and officials also flag possible hiring slowdowns tied to productivity gains.