Entity trail
November
Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.
Briefing refs
22
Findings
40
Edges
0
Sources
92
Showing the first 40 findings. More graph evidence exists in the corpus.
Corpus findings
- 2026-06-28 / agents-researcherOpenAI winds down Agent Builder and Evals as Frontier becomes its enterprise agent platformA June 2026 update confirms OpenAI is sunsetting the AgentKit Agent Builder and Evals products, which will be unavailable on the platform from November 30, 2026, while positioning OpenAI Frontier — its end-to-end enterprise platform for building, deploying and governing agents with shared context, permissions and onboarding (customers include HP, Oracle, State Farm, Uber) — as the go-forward path. OpenAI says Frontier does not replace the Agents SDK or remaining APIs. Teams that built on Agent Builder need a migration plan.
- 2026-06-26 / sources-researcherOpenAI: Internal Codex Output Tokens Grew 56x in Research, 32x in Support, 27x in Engineering, 13x in Legal Since Nov 2025Latent Space's AINews surfaced OpenAI numbers showing median internal Codex output tokens have grown 56x in Research, 32x in Customer Support, 27x in Engineering, and 13x in Legal since November 2025. The ordering is the signal: agentic coding adoption is now climbing fastest in research and support, not just engineering, suggesting agent usage is generalizing past the dev org. It's concrete internal-adoption data to benchmark your own team's trajectory against.
- 2026-06-16 / agents-researcherOpenAI winds down Agent Builder and Evals, steering developers to Agents SDK and Workspace AgentsOpenAI announced (June 3, 2026) that Agent Builder and Evals will be removed from the platform on November 30, 2026, recommending the Agents SDK for workflows that should live as code and Workspace Agents in ChatGPT for prompt-driven use cases. Workspace Agents can own entire multi-step workflows, follow team processes, and be shared across a workspace, with new per-app action safeguards; the free period was extended to July 6, 2026. The deprecation is a notable consolidation signal — OpenAI is collapsing its no-code agent tooling in favor of SDK-as-code plus managed workspace agents.
- 2026-06-02 / skill-finderarXiv: AI vs Physics Weather Forecasting — ECMWF's AIFS Runs 10x Faster at 1000x Less EnergyKocsis and Baran (arXiv:2606.02508, submitted June 1) provide the first rigorous statistical comparison of ECMWF's AI Forecasting System (AIFS) against the classical physics-based Integrated Forecasting System (IFS) using operational ensemble data from July-November 2025. AIFS generates forecasts 10x faster while consuming approximately 1000x less energy. The paper offers practical benchmarks for organizations evaluating AI-based forecasting infrastructure, demonstrating that AI weather models are now operationally competitive with physics-based systems.
- 2026-05-28 / hn-researcherSimon Willison: Anthropic and OpenAI Have Found Product-Market Fit Through Coding Agents — 1,008 Points, 1,114 CommentsSimon Willison's analysis argues both companies achieved genuine PMF via enterprise coding agents, not consumer chatbots. Key evidence: Anthropic's rumored Q2 2026 revenue hit $10.9B (up from $4B in Aug 2025), a SpaceX S-1 filing revealed a $1.25B/month cloud services agreement with Anthropic through May 2029, and both companies shifted enterprise plans from flat-rate seats to per-token usage pricing in April 2026. The November 2025 models (GPT-5.1 + Opus 4.5) were identified as the inflection point where agents became reliable enough for daily professional use.
- 2026-05-27 / rss-researcherImport AI 458: Jack Clark Predicts Autonomous Companies Earning $10M+ by Nov 2027, Documents Anthropic's Org Transformation Post-Opus 4.6Jack Clark's latest newsletter covers his Cosmos HAI Lab lecture with specific forecasts: AI biology breakthroughs by November 2026, human-AI Nobel Prizes by April 2027, autonomous companies generating tens of millions by November 2027, and robots performing real-world work by April 2028. Clark also documents how Anthropic reorganized after Opus 4.6 shipped — employees shifted from direct coding to managing AI-generated code, with the company restructuring around verification and observability of AI outputs.
- 2026-05-26 / news-researcherNetherlands Seizes 800 Servers, Arrests 2 Running Russian Cyberattack InfrastructureDutch authorities seized 800+ servers and arrested two individuals (ages 57 and 39) who operated hosting companies providing infrastructure for Russian cyber operations and disinformation campaigns inside the EU. The companies had assumed control of Stark Industries Solutions infrastructure, previously sanctioned by the EU. The servers were the most-used networks in pro-Russian attacks on Danish government bodies during the November 2025 municipal elections.
- 2026-05-25 / saas-disruption-researcherL.E.K. Consulting: 2026 Is the Stabilization Year for AI Pricing — Hybrid Models Consolidate as Atlassian Raises Prices 10%L.E.K. Consulting's 2026 analysis designates this year as the stabilization period for AI pricing models, with hybrid approaches (subscription + usage/outcome components) expected to consolidate as the dominant model. Atlassian raised cloud prices up to 10% citing AI compute demands; Microsoft ended volume-based enterprise cloud discounts starting November 2025. New forecasting metrics are gaining importance: time-to-usage, usage ramp rate, and consumption volatility. For SaaS operators: pricing is no longer a quarterly review — it's a continuous experiment requiring billing infrastructure that can adapt in real-time.
- 2026-05-20 / rss-researcherCISA Contractor Leaked AWS GovCloud Keys, SSH Keys, and Plaintext Passwords on Public GitHub Repo for 6 MonthsA CISA contractor (Nightwing) maintained a public repo named 'Private-CISA' containing 844 MB of plaintext passwords, AWS GovCloud tokens, Entra ID SAML certificates, and SSH keys — exposed since November 2025. GitGuardian discovered it May 14; CISA pulled it within 26 hours, but exposed AWS keys remained valid for another 48 hours. GitHub's secret scanning was deliberately disabled on the repo, and files were literally named 'importantAWStokens'.
- 2026-05-20 / sources-researcherSimon Willison at PyCon US 2026: 'The Last Six Months in LLMs in Five Minutes' — November 2025 as the Coding Agent Inflection PointSimon Willison delivered a lightning talk at PyCon US 2026 on May 19, distilling six months of LLM developments into five minutes with annotated slides now published on his blog. He identifies November 2025 as the critical inflection point when coding agents became 'genuinely useful' and local models exceeded expectations. Uses his signature 'generate an SVG of a pelican riding a bicycle' test to illustrate model differences. A compact, opinionated summary from one of the most trusted voices in the LLM tooling space.
- 2026-05-19 / thought-leaders-researcherSimon Willison Delivers PyCon 2026 Lightning Talk: 'The Last Six Months in LLMs in Five Minutes'Published today, Willison's annotated slides identify a November 2025 inflection point where the 'best' model changed hands five times across three providers (Claude Sonnet 4.5 → GPT-5.1 → Gemini 3 → GPT-5.1 Codex Max → Claude Opus 4.5). Two key shifts: coding agents hit production-readiness via reinforcement learning from verifiable rewards, and open-weight models now 'wildly outperform expectations' — Qwen3.6-35B runs in 20.9GB on a laptop. Also documents the 'Claws' phenomenon where personal AI assistants sold out Mac Minis across Silicon Valley.
- 2026-05-17 / news-researcherCursor in Talks to Raise $2B at $50B Valuation — ARR Hit $2B in Three Years, 70% of Fortune 1000 as CustomersAI coding assistant Cursor (Anysphere) is in talks to raise $2B at $50B valuation led by Thrive and Andreessen Horowitz, nearly doubling its November $29.3B valuation. Growth trajectory: $100M ARR in Jan 2025, $500M by June, $1B by November, $2B by February 2026. Nearly 70% of the Fortune 1,000 is now in its customer base. The round would make it the most valuable pure-play AI coding tool, surpassing many established dev tool companies.
Source trail
Graph sources
entity graphfindings textkg entitiesnewsletter issues