Scientific discovery's bottleneck has shifted from ideas to data
Terence Tao argues that where hypothesis generation was once science's scarce resource, the bottleneck has shifted toward collecting and verifying against large datasets — with direct implications for how AI could accelerate research.
Details
Core information and root causes
This is a Seed of a Bottleneck Card — the bottleneck was surfaced from a real interview, not yet researched into a full card (current efforts, affected populations, forecast, resources). If you have context, data, or know of existing work on this, we'd love your input.
"I'm not sure nowadays that hypothesis generation is the bottleneck anymore... modern science collects big data first, and then tries to get hypotheses from it." — Terence Tao, on The Dwarkesh Podcast
Tao traces a shift across the history of science: from theory-first idea generation (the classic scientific method) to data-first pattern discovery. That reversal changes what actually limits the pace of discovery today.
Resources
Sources, references, and supporting materials
This card was surfaced by a signal — see the Dwarkesh Patel page for the sourced quote and podcast episode this is drawn from.