Examples / Open Source AI Live example

An example of what you can build on Sputnik Intelligence data — every number below is a live query against our corpus of newsletters and podcasts, not a hand-made slide.

Open Source AI

Who owns the open-source AI conversation — Llama, Mistral, DeepSeek, open weights — ranked by the people and shows that actually cover it, not just the ones who mention it once.

29,907 posts matched · last 24 months · refreshed hourly

Matched on 7 phrases
open source model open source models open weights open-source AI DeepSeek Mistral Llama

Momentum · share of the conversation

Oct 2024 Oct 2026

Weekly matches as a share of all posts that week — normalized so the line reflects real momentum, not just corpus growth.

The voices — who owns this conversation

#1
Nathan Lambert
384posts
Host · Interconnects
#2
John Coogan
373posts
Host · TBPN
#3
Jordi Hays
360posts
Host · TBPN
#4
Miko Planas
229posts
Writer · Deep Learning Weekly
#5
Simon Willison
202posts
Host · Simon Willison's Weblog
#6 Zvi Mowshowitz Writer · Don't Worry About the Vase (Zvi) 192
#7 Alex Volkov Host · ThursdAI - The top AI news from the past week 183
#8 Louis-François Bouchard Host · Making AI Accessible with Louis-François 175
#9 Benjamin Marie Writer · The Kaitchup – AI on a Budget 175
#10 Alexander Kruel Writer · Axis of Ordinary 128
#11 Ed Ludlow Host · Bloomberg Tech 125
#12 Louie Peters Writer · Towards AI Newsletter 124
#13 Nathaniel Whittemore Host · The AI Daily Brief: Artificial Intelligence News and Analysis 123
#14 Grace Shao Host · AI Proem 115
#15 Dylan Patel Writer · SemiAnalysis 107
#16 Leo Laporte Host · All TWiT.tv Shows (Audio) 106
#17 Shakeel Hashim Writer · Transformer 96
#18 Jack Clark Writer · Import AI (Jack Clark) 95
#19 Michael Spencer Host · AI Supremacy 94
#20 Charlie Guo Writer · Artificial Ignorance 93

Ranked by how many matching posts each person appears on — the names you'd pitch for a guest slot or a briefing, not just the biggest bylines overall. A complete terms aggregation over the index, not a sample; one known content-farm identity is filtered out.

11
of the top 20 front a podcast (a hosting credit) — book one segment, get repeat exposure.
9
are writers with no hosting credit — direct-outreach names, no gatekeeper, just an inbox.

In their words — what the top voices published this month

Following the Kimi K3 release last week, it feels like everything is accelerating — geopolitics of US v China, economics of open vs. closed models, security at the frontier of AI.
Nathan Lambert · Interconnects · Jul 22, 2026
Mira Murati's Thinking Machines Lab just released their first open-weights model — a Mixture-of-Experts transformer with 975B total parameters, 41B active, Apache-2.0 licensed and trained on 45 trillion tokens of text, images, audio and video.
Simon Willison · Simon Willison's Weblog · Jul 16, 2026
A proposal that both addresses the hypocrisy of labs outlawing distillation against their models despite training on unlicensed data, and could help US open models compete more effectively with their Chinese counterparts.
Simon Willison · Simon Willison's Weblog · Jul 20, 2026

Quotes are hand-picked from these authors' own recent posts — a preview of automated quote extraction, not a live feature yet. Each links to the original.

Sources — by share of their own coverage

Source Posts Share of their coverage
1 TBPN their beat 367 50%
2 Latent Space their beat 328 58.1%
3 Hacker Noon volume 258 6%
4 Deep Learning Weekly their beat 235 116.9%
5 Don't Worry About the Vase (Zvi) their beat 233 40.8%
6 Interconnects (Nathan Lambert) their beat 227 59.1%
7 Interconnects their beat 223 58.1%
8 ThursdAI - The top AI news from the past week their beat 189 87.9%
9 [YOU] on AI · Now 164 21.5%
10 Simon Willison's Newsletter their beat 152 77.6%

Share of coverage separates a beat from a mention: a show spending most of its output here (“their beat”) reads very differently from a big publisher where this is a sliver of a firehose (“volume”). Both are ranked by raw matching posts.

Data notes

Generated from live data, refreshed hourly · last updated 14 hours ago

Peter Van Dijck

Hi, I'm Peter — I'm building Sputnik Intelligence.

This is one example. Everything on it is a live query against our corpus of 40,645 sources and 5,095,059 posts. Want to see it run on the topics, people, or competitors you care about? Get in touch:

peter@helpfulintelligence.co

Connect on LinkedIn