Bending the Curve of Discovery: AI in Science Today and Tomorrow

Scientific and technological progress has long driven economic growth and led to dramatic improvements in living standards around the world. In his 2025 Nobel lecture, economic historian Joel Mokyr attributed this to a “positive feedback loop”: new technology drives waves of scientific advances that, in turn, fuel further innovation and progress. But he also warns that sustaining this positive loop is never guaranteed.

Recent history illustrates Mokyr's point. The U.S. and E.U. entered the 1990s with relatively similar economic growth rates, but the two have since diverged substantially. The U.S. maintained roughly 2% annual increases in output per capita, largely powered by successive science and technology booms and the broad diffusion of innovation across the economy, while the E.U. has mostly stagnated.

Today, artificial intelligence is beginning to touch nearly every discipline of modern inquiry — from folding proteins and mapping genomic topologies to discovering novel crystalline materials and predicting atmospheric fluid dynamics. Yet the fundamental challenge remains: moving beyond passive pattern matching toward active hypotheses generation and verifiable experimental design.

Economists classify breakthrough technologies like the steam engine, electricity, and the semiconductor as general-purpose technologies (GPTs). Their primary economic value comes not just from direct productivity boosts, but from transforming how downstream sectors innovate.

AI represents a uniquely reflexive category: an invention for inventing itself. When computational agents can automate literature syntheses, simulate molecular dynamics, and run parallelized robotic lab trials around the clock, the cycle time of scientific discovery shrinks by orders of magnitude.

Despite exponential gains in model parameter scale, significant institutional and epistemological friction points persist. High among them are data provenance bottlenecks, the reproducibility crisis in automated benchmarking, and the steep compute requirements that threaten to concentrate fundamental discovery within a handful of well-funded industrial labs.

To maintain open and resilient research ecosystems, engineering infrastructure must prioritize explainability, cryptographic data integrity, and decentralized compute access.

As we look ahead to the next decade, the symbiosis between human scientists and autonomous cognitive architectures will define the trajectory of human knowledge. By pairing human intuition, taste, and ethical judgment with the exhaustive exploration capacity of AI, humanity stands poised to bend the curve of discovery on the most pressing challenges of our time.