Verified founder journeys, startup histories and public perspectives from AI entrepreneurs, researchers, CEOs and investors worldwide. Every statement is a summary of a public position with its official source.
Positions AI as an instrument for accelerating scientific discovery — protein structure prediction being the first proof that a learned model can solve a 50-year-old grand challenge in biology.
Advocates open foundation models, arguing that a small number of closed providers controlling the AI layer of information is a bigger long-term risk than open weights, and that current autoregressive LLMs alone will not reach human-level intelligence.
Argues most production AI teams gain more from systematically improving data quality and labelling consistency than from further model architecture changes — the data-centric AI movement.
Calls for human-centered AI: research, policy and product decisions evaluated by their effect on human dignity, labour and community — and for public-sector compute so academia can keep doing frontier research.
Frames the current shift as the end of general-purpose CPU scaling: computing must be accelerated end-to-end, and data centres become 'AI factories' that manufacture tokens rather than store data.
Argues India should apply AI on top of its digital public infrastructure — identity, payments and data-sharing rails — to deliver population-scale services rather than compete purely on frontier model training.