Industry skeptics call AI trajectory a 'rug pull'

The "scaling is all you need" era is facing a major vibe shift toward skepticism as promised breakthroughs feel increasingly out of reach.

• –Critics point to Ilya Sutskever’s recent pivot from scaling to fundamental research as evidence that current LLM architectures have reached a plateau.
• –There is a growing perception that AGI declarations from CEOs like Sam Altman and Jensen Huang are more about protecting massive infrastructure investments than genuine technical milestones.
• –The "rug pull" sentiment reflects a divide between hype-driven acceleration and the practical, sample-efficient learning required for true intelligence.
• –Skepticism is mounting over whether LLMs can ever achieve generalization without entirely new, less compute-hungry architectures.