IIT Madras AI head: scaling bigger models without safety systems is worrying
ravi_iitm · x · 2026-09-30
Prof Balaraman Ravindran, head of the Wadhwani School of Data Science and AI at IIT Madras, warns that as AI moves from chatbots to agentic systems that plan tasks, interact with other AI agents and act in the real world, safety safeguards must keep pace.
- He argues even a single life lost to an AI system's unintended action would be too high a cost, calling for strong guardrails and safety engineering, with AI harnesses and narrow custom-built solutions to constrain agent behavior
- The interview also covers agentic AI's impact on cybersecurity and jobs, Indian-language models, Tamil Nadu's opportunities, and IIT-M's AI Studio
- Beyond the bigger-model race, he highlights important work applying AI to understanding the human genome and gene mutations
Related event: IIT Madras Professor: Agentic AI Growth Must Not Outpace Safety Guardrails(2 posts)→
More from AGI Musings
- Analyzing a tweet archive with Sol: identity as compression, emotion as search — mimi10v3 · 2026-09-30
- Wanted AI to book my flights; instead I became the intern — mayanicks0x · 2026-09-30
- Gary Marcus calls excluding scientists from AI planning 'an epic mistake' — GaryMarcus · 2026-09-30
- Scoble on the open-source AI debate: weakening open models weakens defense too — Scobleizer · 2026-09-30
- Hinton Argues Machines Can Have Emotions; Domingos Counters: They Have Reward Functions — pmddomingos · 2026-09-30
- Simons Institute Report Lays Out 12 Actions for TCS Community to Respond to AI Progress — jasondeanlee · 2026-09-30