The danger is not only that AI will write for us. It is that human thought itself may increasingly be reorganised through the statistical average embedded in large language models.
AI tools are accelerating academic work, but unstable plan rules, opaque systems, and platform lock-in risk creating a new two-tier knowledge order unless research infrastructure remains open, inspectable, and interoperable.
The ILO–World Bank warning is important, but labour policy is still calibrated to GenAI while firms are already reorganising work around agentic AI that can execute workflows, not just generate outputs.
Telangana’s new gig worker law does not resolve the employment-status question, but it does mark a break with the long regulatory absence that allowed platforms to govern labour with minimal accountability.
Oracle’s layoffs show how the AI transition is being financed: by cutting labour to fund capital-intensive cloud and data-centre buildout — with the contradiction especially sharp in India.
As AI systems become increasingly capable across cognitive work, the central issue is not only automation but the changing balance of dependence between capital and labour.