From novices to co-pilots: Fixing the limits on scientific knowledge production by accessing or building expertise

The current approach of relying primarily on institutional experts to create knowledge to solve humanity's problems is insufficient to meet the scale, diversity, and novelty of people's needs. Building expertise in people to create knowledge they need provides a promising approach. Despite having contextual insights, people fail to rapidly generate sound plans---like experiments---and correctly implement specific actions---like data acquisition and analysis. The limits to progress in multiple domains---like science and healthcare---can potentially be expanded by building procedural expertise among motivated nonexperts so that they can build on their contextual insights to create valid and generalizable knowledge. In this paper, we report on the design and evaluation of tools that highlight two ways to realize this vision. First, Hevelius is a motor impairment assessment tool for patients to conduct neurological assessments online. A rare disease community has provided fine-granular data and insights from their homes that current in-clinic assessments fail to capture. Second, Gut Instinct is a social computing system that supports procedural knowledge acquisition for experimentation. A fermentation community used Gut Instinct to successfully design and run between-subjects experiments to test their intuitions. These results suggest exploring ways of producing knowledge that are distinct from the dominant model of institutionally-situated experts testing their ideas on subjects in a lab or a clinic. More constructively, these systems demonstrate how knowledgeable and committed people can be aided and amplified by technology in creating scientific knowledge.