IndiaAI Mission targets open-source reasoning models as inference breakthrough cuts computing costsThe shift towards test-time compute and reinforcement learning allows sovereign AI programmes to build high-performance systems with far fewer chips.15 October 2026
Reasoning models mark the end of the raw scale era in artificial intelligenceThe focus of frontier AI development has shifted from massive training runs to letting models calculate their way through complex problems before answering.15 September 2026
