689 AI communities surface on Luma — 67 with strong evidence, and a long tail of 336 barely-indexed local groups

Asked:

Can you do a deep search on Luma to identify all the AI communities that exist?

A discovery census as of September 1, 2026: 36 web queries across topics, demographics, and regions surfaced thousands of candidate Luma pages; semantic filtering and name normalization retained 689 named ongoing AI/ML communities or recurring calendars, backed by 2,499 search mentions in total. A location was stated for 576 communities and a member or subscriber count for 91. Luma has no public exhaustive index, so this is broad coverage — not a guarantee of "all".

Every community found, one dot each

Dot area = search mentions (1 to 347); grouped by discovery confidence. Hover any dot for the community, its location, and evidence.

high — ≥5 source pages or ≥10 mentions (67) medium — ≥2 pages or explicit calendar title (286) emerging / lightly indexed — single or weak source (336)

Where they gather

Communities per hub, from the stated location field (576 rows carry one). Ambiguous or one-off locations are not force-geocoded.

A further 245 located communities sit outside these hubs — from Sofia to Nairobi to Kuala Lumpur's neighbours — and 113 rows state no location.

The AI Collective is the most-evidenced community by far — 347 search mentions and a stated 200k+ members worldwide. luma.com

High-confidence communities are only 10% of the census but carry 1,580 of the 2,499 total search mentions — 63% of all evidence. luma.com

The Bay Area leads every city with 83 communities, but Singapore alone hosts 34 — more than New York's 25 or London's 24. luma.com

36 explicitly global or online communities span borders, from the 30K-member Build Club to Women Defining AI's 1,000+ virtual members. lu.ma

Method and limits

Confidence measures discovery evidence, not quality or activity. High = at least five source pages or ten search mentions; medium = at least two source pages or an explicit community/calendar title; emerging = a single or weakly indexed source.
Search engines incompletely index Luma's dynamic, private, or unlisted calendars, so some real communities are missing and some appear only through event pages.
Source links are the best available evidence page for each named host — not always the canonical calendar; some are single event pages.
Unresolved fields (location, members, focus) are left blank rather than guessed.

The complete directory

Community	Location	Members (as stated)	Confidence	Mentions	Pages	Source

Data: deep-search census of Luma-hosted AI/ML communities, 689 rows retained from 36 web queries after semantic filtering and name normalization, snapshot 2026-09-01. Search mentions count query-result appearances; source pages count distinct evidence pages. Member counts are quoted verbatim where a page stated one (91 rows). Hub chart groups stated locations by city/region keywords; unmatched locations are counted as "other". Nothing was cut: all 689 rows are in the directory.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.