info@innovationlens.org
Most people talk about AI like it’s a faster intern. Jonah Lynch is building something closer to an intellectual compass: a system that can “read” the scientific literature at scale, map what we already know, and point toward the empty spaces where the next discoveries are most likely to happen.
We unpack Innovation Lens, Jonah’s research forecasting platform that uses natural language processing, text embeddings, and geometry in vector space to detect patterns across millions of papers. He explains the core intuition behind prediction in science: some fields are too sparse to pay off, others are so crowded that the easy value is gone, and there’s a Goldilocks zone where the research landscape is ready for a breakthrough. We also talk about validation and benchmarking, why this approach can beat random guessing and even the standard “follow the adviser and find a gap” method, and what it changes for PhD topic selection, literature review, and R&D strategy.
The conversation gets personal too. Jonah shares how leaving the Catholic priesthood pushed him to rebuild his life around quantitative tools and a search for truth that doesn’t rely on authority. From VC decision-making and capital allocation to philanthropy, NSF-style grant impact, and better alternatives to citation metrics, we explore where AI genuinely helps human flourishing instead of just generating content.
If you enjoy episodes about scientific discovery, innovation prediction, and practical AI for research, subscribe, share this with a friend who works in science or investing, and leave us a review. What domain would you want a “map of the future” for?
Help these new solutions spread by ...
Comments? Feedback? Questions? Solutions? Message us! We will do a mailbag episode.
Email: solutionsfromthemultiverse@gmail.com
Adam: @ajbraus - braus@hey.com
Scot: @scotmaupin
adambraus.com (Link to Adam's projects and books)
The Perfect Show (Scot's solo podcast)
Thanks to Jonah Burns for the SFM music.
