Nextomoro is independent intelligence on the AI research field: the labs, the models, the people, and the patterns running across them.
Profiles cover the labs, the models they ship, and the people moving between them. Storyline essays sit alongside, dated and first-person, on the bigger patterns running across the field. The Atlas is the structured-data layer behind the prose: funding rounds, valuations, model benchmarks, who's at which lab.
Why this exists
Most AI media is organized around the wrong units. Newsletters cover stories. Benchmark sites cover models in isolation. Investment research treats labs as financial assets. Almost nobody connects the pieces: which lab a researcher just left, what model a lab just shipped, what that all says about who's about to ship what next. A reader trying to map who's where, working on what, catching up to whom has to assemble that picture from a dozen places.
I also cover the field globally on purpose. The Frontier US labs get the same treatment as the Chinese frontier, the sovereign-AI labs in Korea and India and Germany, and the open-source institutions. Most AI publications cover anything outside the US/UK orbit as a footnote. I'm honest in both directions: hype called on the dominant labs, Chinese, Indian, European, and Korean labs taken seriously on their own terms. No puff pieces, no flag-waving.
How it's organized
Lab profiles, model profiles, and people profiles share a Wikipedia-style descriptive voice. They open with what the entity is, then run through origins, key facts, current standing, sources. Profiles avoid editorializing.
Storyline essays carry the editorial voice. They're dated, first-person, and cover patterns that cross labs: research-talent flows, the coding-agent race, China closing the gap. New essays publish roughly every two weeks.
The Atlas tracks the numbers behind the prose: funding, valuations, benchmarks, model lineups. It refreshes quarterly and lets readers slice the field by category, region, or weight class.
Sources and corrections
Funding figures, valuations, benchmark scores, and "current model" claims are point-in-time. Structured data refreshes quarterly. Profiles get updated when something noteworthy happens: a new model, a new round, a leadership change.
Sources are cited inline. I rely on labs' own announcements, primary press, benchmark leaderboards (Artificial Analysis, LMArena, SWE-bench), founder interviews and conference talks, and Crunchbase or SEC filings where applicable.
If you spot a factual error, a stale benchmark rank, or a lab you think deserves coverage, get in touch.
Who writes this
Nextomoro is editorially independent. Coverage decisions and the editorial line live with a single editor. The Atlas's records are maintained continuously, on the cadence above.
