Apply Now for New Postdoctoral Scholar Positions at the Knowledge Lab at the University of Chicago

We seek outstanding candidates for government and philanthropically funded postdoctoral positions developing large language models (LLMs) and associated AI systems to (1) observe the global system of science and technology at scale; (2) predict future science and technology; and (3) experimentally explore alternative futures and how to configure the world to achieve them. Postdoctoral Scholars will work at the interface of AI research on transformers and novel deep learning designs and realized large-scale systems in collaboration with the Allen Institute for Artificial Intelligence (e.g., its AI-powered Semantic Scholar and Open Language Model, OLMo) and the U.S. Department of Energy’s Trillion Parameter Consortium. This work will be in collaboration with computational social scientists (e.g., economists, sociologists, physicists) furthering the science of science and innovation in order to guide U.S. and global strategic investments in science and technology.

Candidates will work in the Knowledge Lab in the Division of Social Sciences at the University of Chicago, directed by Prof. James Evans. Technological change drives economic change, creating new jobs while making others more efficient, replacing industries in cycles of creative destruction that generate new goods and services to promote human productivity and prosperity. Understanding technological progress and leadership is key to ensuring sustainable prosperity, security, and equity, and buffering against economic shocks. Nevertheless, despite enormous data on science and technology, little is understood about what pathways could ensure advance in critical technology capacity, production, and use. This effort will require improved situational awareness regarding the global techno-scientific environment, strengthened capacity to predict technological advance across that environment, and establishing causal linkages between specific policy levers for technological progress and leadership. We will model and predict the global system of technological development to understand and shape technological capabilities. To enable holistic exploration, we will build a powerful large language model (LLM) atop precise data on global technology funding, research, intellectual property, products, and their mentions, linked by the people, regions, and organizations producing and consuming them, to dynamically account for complex and emergent interdependencies among diverse technological domains. We seek to construct our model to function as a global observatory of the innovation system, and as a virtual laboratory to simulate alternative outcomes (e.g., advances predicted to result from specific funding, educational, or organizational policies). Our LLMs will also support the estimation of structural models to causally identify the impact of funding on interlocking discoveries, overlapping supply chains, global networks of skilled labor, and industrial organizations that drive or impede international technology leadership.