The Literature Abyss
A new field can mean 40,000 papers. Months of chasing citations later, you still can't tell the foundational work from whatever everyone else happened to cite.
We onboard research groups one at a time. Cohort 1 is forming now — join our Discord to follow along.
10M+
Researchers each rebuilding the same map from scratch
23%
Of work time spent re-reading what others already mapped
$150B
Spent annually on "catching up"
2 yrs
For a new lab member to inherit what came before
That process hasn’t changed in 50 years. It is slow, solitary, and painful. And it looks like this.
A new field can mean 40,000 papers. Months of chasing citations later, you still can't tell the foundational work from whatever everyone else happened to cite.
Three people in your department reviewed the same literature last year. Nobody knew. The models they built stay locked in their heads — and leave when they do.
A new postdoc spends two years just catching up on the field — rebuilding a mental model that already lives in the departing postdoc's head.
Dive into articles, updates, and perspectives on building a more connected research ecosystem.
View all articles
HRA KG: What the Human Reference Atlas Knowledge Graph Maps, Enables, and Leaves Out
A researcher's guide to the Human Reference Atlas Knowledge Graph (HRA KG) — what it maps across 71 organs, what it enables, and where its limits lie.
12 Aug, 2026 • 11 min read
Why "Just Use ChatGPT" Doesn't Work for Real Scientific Discovery
ChatGPT is built to summarize, not to discover. Why real research needs provenance, relationships, and freshness that general chat LLMs can't provide.
7 Aug, 2026 • 9 min read
Papers with Code: What It Built, What It Meant, and What Comes Next
Papers with Code shut down in July 2025. What it built, what the shutdown cost ML research, and which tools now cover the gap it left behind.
3 Aug, 2026 • 9 min read
PubMed Knowledge Graph 2.0: What It Connects, Enables, and Where It Falls Short
A review of PubMed Knowledge Graph 2.0 — what its papers, patents and trials integration enables, where the pipeline falls short, and how to use it well.
29 Jul, 2026 • 10 min read
This is not a newsletter signup. It is an application. We read every one. We invite in cohorts of 50, prioritising labs that want to map their own work first and contribute to their public knowledge graph — not just use a product.
We'll review your application within 7 days. Share your link to move up — every accepted referral bumps your position.
Looks like you've already applied. Here's your spot and your referral link to share.
50.4% filled · 248 seats remain
After applying, you’ll receive a personal referral link. Each colleague you refer moves you higher in the queue.