Throughout civilization, the greatest periods of economic and societal progress have come when humanity unlocked a new source of productive capacity. Our new initiative, Pioneering Abundance, is pursuing the next great unlock. Read our latest perspective:

“We try and find problems that no one else is going to solve, and then solve them.” Head of @Flagship_PI Matt Davis on how PI applies Flagship's whitespace approach to AI by pursuing what no one else will. Watch Matt’s full 3 Questions:

For thousands of years, using massive amounts of water to flood rice fields looked like the price of feeding the world. It wasn't a law of nature — just a workaround for a problem we couldn't yet solve biologically. In our latest post on The Labs Report, Flagship’s Ignacio Martinez and Alec Asbridge explore what changes once biology becomes readable, and how our Pioneering Abundance initiative takes an ecosystem approach to turn those insights into real solutions. Read:

“Solving the structure of the ribosome” was once a joke — a shorthand for a structural problem ambitious enough to sound absurd. So we chose it as a test. In 36 seconds on a single GPU, LightFold co-folds the E. coli 70S ribosome assembly: 11,000 residues and nucleotides, comprising nearly 150,000 atoms. At ~6,400x speedup and 48x less memory usage than Fold-CP-enabled models, LightFold begins to move long-context prediction from a computational feat to a drug-development tool. Learn more:

We're teaming up with @AnthropicAI and @modal to host a Computational Biology Hackathon. Bring a protein sequence, gene set, or microscopy dataset — leave with an agentic tool built on Modal x Claude Science. Sept 22 | 9AM–7:30PM | Cambridge, MA RSVP:

Today, Flagship launches Pioneering Abundance—a new initiative focused on reimagining how we produce the resources that power our world. For decades, we’ve built around the constraints of scarcity. Now, biology, AI, and deep tech are opening a different path: creating abundance. Learn more: #BiggerLeaps#

The gap between when disease begins and when we can detect it is closing. Flagship Origination Partner Scott Lipnick shares what happens next, in his recent perspective for The Labs Report:

Noubar Afeyan has built a career around an unusual premise: Instead of waiting for the next biotech breakthrough, create the conditions for one. The Flagship Pioneering founder is a 2026 Fierce 50 Leadership honoree. 🔗 @FlagshipPioneer #Fierce50#

AI is beginning to reveal nature’s patterns, helping us recognize connections, ask better questions, and uncover possibilities we couldn't have reached on our own. Polyintelligence — the synthesis of human, machine, and nature's intelligence — is unlocking new forms of problem solving and creativity. So, what breakthroughs become possible when we can finally understand — and build with — nature’s intelligence? @JCoolScience @AnthropicAI @AvakK @afzelius_lovisa @Prologue_Med @bramaccimmyt @CIMMYT

Medical interventions have been blunt because our view of biology has been blunt. That is beginning to change. Flagship Origination Partner Scott Lipnick on what it takes to move from treating disease to getting ahead of it:

.@FiercePharma named Flagship Founder & CEO @NoubarAfeyan to its Fierce 50 list, recognizing his leadership in pioneering company creation, advancing science and technology, and expanding Flagship’s model globally. Read more: #BiggerLeaps#

LNPs are often typecast as having one job — packing therapeutic nucleic acids for delivery. But that’s a vast oversimplification. LNPs have multiple components (including lipids, polymers, and chemical or protein ligands) and functions, including cargo protection, tissue/cell selectivity, liver detargeting, intracellular delivery, and tolerability enhancement. Co-optimizing all of these components while checking all these functional boxes requires intensive bioengineering. @mirai_bio is leveraging machine learning expertise — informed by enormous, proprietary in vivo datasets — to iteratively develop precise LNPs. This robust design and optimization process creates LNPs that home in on targeted tissues while avoiding hepatic capture. Read more:

These figures show peak memory and runtime for a range of co-folding models, all on the same 96 GB GPU. What the grey curves have in common is more interesting than what separates them. They bend sharply upward past about 1,000 tokens and reach the card's ceiling between 2k–3k residues. After this threshold, these models stop working. Why? Because they share an architecture that spends memory as a square of the size of the complex they are attempting to fold. LightFold’s near-linear memory allows that same GPU to keep folding complexes past 15K residues on the same hardware, so its line rises gently. Taken together, where the other models hit a cliff, LightFold finds a slope that allows it to move beyond folding subunits into folding whole assemblies. More here: