When Machines Design, Prove, and Discover
This issue traces a shared arc: AI systems are moving from assistants to autonomous agents that design molecules, and formal verification is emerging as the trust layer that keeps their output honest.
Newsletter
The weekly issue archive
One issue a week: a curated, sourced digest of what's new across neurosymbolic AI, AI-driven science, formal methods and verification, and AI in the life sciences. Every item is summarized and linked to its original, and every issue is published here in full.
-
-
Reasoning, Rigor, and the Real Limits of AI for Science
This issue examines where AI is genuinely reshaping science—from reasoning agents and math-grounded genomics to protein dynamics and clinical integration—and where the hype outruns the evidence.
-
Editing Life, One Protein at a Time
This issue tracks how AI is moving from designing molecules to editing them—resizing proteins, steering generative models toward real properties, and confronting the data and infrastructure gaps that stand between prediction and validated science.
-
Proteins on Demand, Agents Under Test
This issue tracks how machine learning is learning to resize, steer, and condition proteins with preserved function, alongside the harder engineering problem of making AI agents and AI-driven labs actually reliable.
-
Verified Autonomy: When Proof Meets the Lab
This issue traces a single through-line — the push to make AI-driven science and agentic systems verifiably correct rather than merely tested — from industrial investment in proof languages to formally grounded research agents and the frontier of self-driving labs.
-
Verified Minds: Provers, Agents, and the Neurosymbolic Turn
This issue tracks the convergence of formal verification and LLMs, the rise of general-purpose AI science agents, and fresh evidence sharpening the case for neurosymbolic hybrids.
Prefer a feed? Subscribe via the newsletter RSS feed, or follow everything on this site through the combined feed.