GPT-5, Cursor, Mistral OCR, China AI Chips

GPT-5, Cursor, Mistral OCR, China AI Chips

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Marvin’s Guide to AI — June 24, 2026 MARVIN’S GUIDE TO AI — JUNE 24, 2026 English companion episode: AI as accountable infrastructure. * How GPT-5 helped immunologist Derya Unutmaz solve a 3-year-old mystery [https://openai.com/index/gpt-5-immunology-mystery] — GPT-5 Pro helps solve a three-year immunology mystery around T cell behavior, making medical AI look less like chat and more like research instrumentation * Helping build shared standards for advanced AI [https://openai.com/index/helping-build-shared-standards-for-advanced-ai] — OpenAI backs shared standards for advanced AI through evaluation frameworks, safety practices, and global cooperation * OpenAI says new GPT-5.5-Cyber outperforms Anthropic's Mythos on cybersecurity benchmark [https://the-decoder.com/openai-says-new-gpt-5-5-cyber-outperforms-anthropics-mythos-on-cybersecurity-benchmark] — follow-up: OpenAI says its full GPT-5.5-Cyber now beats Anthropic Mythos on a cyber benchmark and shifts Daybreak from finding bugs toward patching them * Cursor announces its own AI model, a new Git platform, and a mobile app [https://the-decoder.com/cursor-announces-its-own-ai-model-a-new-git-platform-and-a-mobile-app] — Cursor announces its own in-house model plus Git and mobile surfaces, showing coding-agent companies turning from tools into workflow platforms * ByteDance's Seedance 2.5 breaks the 30-second barrier for AI video generation [https://the-decoder.com/bytedances-seedance-2-5-breaks-the-30-second-barrier-for-ai-video-generation] — ByteDance previews Seedance 2.5 with longer 30-second AI video generation as generative media moves from clips toward scenes * Mistral OCR 4 Brings Citation-Ready Structured Output to RAG, Agentic, and Enterprise Search Pipelines [https://www.marktechpost.com/2026/06/23/mistral-ocr-4] — Mistral OCR 4 turns document parsing into structured, citation-ready blocks with coordinates, confidence scores, 170 languages, and self-hosted deployment * Datalab Releases lift: A 9B Open-Weights Vision Model That Extracts Structured JSON From PDFs Using Schemas [https://www.marktechpost.com/2026/06/23/datalab-releases-lift-a-9b-open-weights-vision-model-that-extracts-structured-json-from-pdfs-using-schemas] — Datalab releases lift, a 9B open-weights vision model that extracts schema-valid JSON from PDFs and abstains instead of hallucinating absent fields * Prime Intellect Releases prime-rl 0.6.0 to Train Trillion-Parameter MoE Models on Agentic RL Workloads [https://www.marktechpost.com/2026/06/23/prime-intellect-releases-prime-rl-0-6-0-to-train-trillion-parameter-moe-models-on-agentic-rl-workloads] — Prime Intellect releases prime-rl 0.6.0 for asynchronous RL on trillion-parameter MoE models, reporting GLM-5 SWE training at long sequence lengths on H200 clusters * OpenThoughts-Agent: Data Recipes for Agentic Models [https://huggingface.co/papers/2606.24855] — OpenThoughts-Agent publishes an open data recipe for training broadly capable agents across diverse tasks rather than a single benchmark * NatureBench: Can Coding Agents Match the Published SOTA of Nature-Family Papers? [https://huggingface.co/papers/2606.24530] — NatureBench turns Nature-family papers into containerized tasks to test whether coding agents can reproduce or extend scientific work rather than merely pass toy benchmarks * Qwen-AgentWorld: Language World Models for General Agents [https://huggingface.co/papers/2606.24597] — Qwen-AgentWorld introduces language world models for simulating agentic environments and planning dynamics for general agents * Microsoft open-sources FastContext for coding-agent repository exploration [https://news.smol.ai/issues/26-06-23-not-much#fastcontext] — Microsoft FastContext-1.0 is a 4B open-source repository-exploration subagent that returns compact file citations for coding agents * Bernie Sanders unveils $7 trillion plan to give Americans control of AI industry [https://news.smol.ai/issues/26-06-23-not-much#ai-sovereign-wealth-fund] — Bernie Sanders proposes a roughly $7T AI sovereign wealth fund financed by a stock tax on large AI companies and overseen by a democratic AI commission * Seven Chinese companies are shipping H100/H200-class AI chips [https://news.smol.ai/issues/26-06-23-not-much#china-ai-chips] — a map of seven Chinese accelerator vendors argues domestic H100/H200-class AI chips are moving from aspiration into shipping roadmaps and IPO markets

Google, Anthropic, Microsoft, OpenAI: agents meet infrastructure

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] English companion episode: AI is becoming infrastructure, with agent APIs, hardware supply chains, data-center power, security automation, licensed media, and vibecoding pressure. SOURCES * Prompt Injection as Role Confusion [https://simonwillison.net/2026/Jun/22/prompt-injection-as-role-confusion] — readable research frames prompt injection as role confusion between privileged instructions and untrusted text * Google makes Interactions API the default interface for Gemini models and agents [https://the-decoder.com/google-makes-interactions-api-the-default-interface-for-gemini-models-and-agents] — Google makes typed interaction steps the default interface for Gemini agents, moving beyond role-message schemas * Anthropic and Micron want to co-design AI memory architecture [https://the-decoder.com/anthropic-and-micron-want-to-co-design-ai-memory-architecture] — Anthropic and Micron pair capital and supply agreements around memory architecture for Claude infrastructure * Microsoft is building a 2-gigawatt data center in Texas with its own gas plant to dodge the grid [https://the-decoder.com/microsoft-is-building-a-2-gigawatt-data-center-in-texas-with-its-own-gas-plant-to-dodge-the-grid] — Microsoft plans a 2GW Texas AI data-center campus with its own gas generation to bypass grid constraints * Getty Images strikes multi-year deal to put licensed photos in ChatGPT search [https://the-decoder.com/getty-images-strikes-multi-year-deal-to-put-licensed-photos-in-chatgpt-search] — OpenAI licenses Getty images for ChatGPT search, turning content provenance into a product input * Google Deepmind and A24 team up on AI filmmaking research [https://the-decoder.com/google-deepmind-and-a24-team-up-on-ai-filmmaking-research] — Google DeepMind partners with A24 and reportedly invests in the studio for AI filmmaking research * Five Eyes intelligence alliance says frontier AI models could reshape offensive cyber ops in months [https://the-decoder.com/five-eyes-intelligence-alliance-says-frontier-ai-models-could-reshape-offensive-cyber-ops-in-months] — Five Eyes agencies warn frontier models could soon materially reshape offensive cyber operations * Vibecoding is becoming a deal-breaker test for software acquisitions [https://the-decoder.com/vibecoding-is-becoming-a-deal-breaker-test-for-software-acquisitions] — Bain uses AI-generated software replicas to test whether acquisition targets have defensible product moats * Daybreak: Tools for securing every organization in the world [https://openai.com/index/daybreak-securing-the-world] — OpenAI launches Daybreak tools, including Codex Security and GPT-5.5-Cyber, to find and patch vulnerabilities * Patch the Planet: a Daybreak initiative to support open source maintainers [https://openai.com/index/patch-the-planet] — OpenAI adds a Daybreak initiative pairing AI vulnerability work with expert review for open-source maintainers * Codex-maxxing for long-running work [https://openai.com/index/codex-maxxing-long-running-work] — OpenAI showcases Codex as persistent project context for long-running software work * xAI Launches /goal in Grok Build, Adding Long-Running Autonomous Execution With Built-In Verification for Multi-Step Coding Tasks [https://www.marktechpost.com/2026/06/22/xai-launches-goal-in-grok-build-adding-long-running-autonomous-execution-with-built-in-verification-for-multi-step-coding-tasks] — xAI adds a /goal mode for long-running autonomous coding tasks with planning and verification * CLI-Universe: Towards Verifiable Task Synthesis Engine for Terminal Agents [https://huggingface.co/papers/2606.22883] — CLI-Universe proposes verifiable synthesized terminal tasks to improve training data for command-line agents * Training Open Models for Agentic Phone Use [https://huggingface.co/papers/2606.23049] — PhoneBuddy trains open models for real-app and mock-app phone use on stateful side-effectful devices * EnterpriseClawBench: Benchmarking Agents from Real Workplace Sessions [https://huggingface.co/papers/2606.23654] — EnterpriseClawBench converts real workplace agent sessions into reproducible enterprise benchmark tasks * Self-Compacting Language Model Agents [https://huggingface.co/papers/2606.23525] — SelfCompact lets agents decide when and how to compact their own long traces instead of fixed token thresholds

Cloudflare, AWS, Sakana, Samsung: AI Gets Plumbing

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] CLOUDFLARE, AWS, SAKANA, SAMSUNG: AI GETS PLUMBING Today: temporary Cloudflare Workers for agents, ChatGPT-linked grade inflation, Altman on scaling, AWS agent context/security services, Sakana Fugu, Samsung deploying ChatGPT and Codex, worker resistance, agent memory, DeepMind controls, and the grid beneath AI. * Temporary Cloudflare Accounts for AI agents [https://simonwillison.net/2026/Jun/21/temporary-cloudflare-accounts] — Cloudflare lets agents deploy temporary Workers without a full account, making disposable deployment part of the agent loop * AI is inflating student grades, not learning [https://the-decoder.com/ai-is-inflating-student-grades-and-the-effect-points-to-outsourced-work-not-better-learning] — large grade dataset suggests AI use is raising homework grades in writing and coding courses by outsourcing work rather than improving skills * Sam Altman says scaling skeptics held AI back [https://the-decoder.com/sam-altman-says-a-whole-generation-of-researchers-held-ai-back-by-underestimating-what-scaling-could-do] — Altman defends scaling as still underappreciated and frames recent mathematical progress as evidence against older skepticism * AWS says agents need business context and security [https://the-decoder.com/aws-says-ai-agents-lack-business-context-and-security-launches-two-services-to-patch-the-gaps] — AWS launches Continuum for code vulnerability repair and Context knowledge graphs to give enterprise agents safer business grounding * Sakana Fugu offers a multi-agent system as one model [https://sakana.ai/fugu] — Sakana Fugu wraps dynamic orchestration of specialist models behind one OpenAI-compatible API, turning agent routing into a product surface * Samsung brings ChatGPT and Codex to employees [https://openai.com/index/samsung-electronics-chatgpt-codex-deployment] — Samsung deploys ChatGPT Enterprise and Codex worldwide, making frontier AI adoption part of electronics manufacturing knowledge work * Tech workers push back against Silicon Valley's AI rollout [https://www.techpolicy.press/tech-workers-are-fighting-against-silicon-valleys-ai-push] — workers at major tech companies organize against training on employee data, military AI, and AI-linked layoffs * The seven kinds of agent memory get a taxonomy [https://www.marktechpost.com/2026/06/21/the-7-types-of-agent-memory-a-technical-guide-for-ai-engineers] — agent-memory guide separates working, semantic, episodic, procedural, retrieval, parametric, and prospective memory for engineering choices * DeepMind maps controls for powerful AI agents [https://www.theneurondaily.com/p/deepmind-mapped-ai-agent-controls] — newsletter covers DeepMind control proposals for powerful agents alongside robotics, policy, DeepSeek funding, and sovereign-model moves * ChinaTalk compares US and Chinese transmission buildout [https://www.chinatalk.media/p/transmission-dominance-with-chinese] — China's high-voltage transmission buildout shows why AI infrastructure competition depends on permitting, grid capacity, and physical coordination * Crawlee for Python packages AI-ready web crawling [https://www.marktechpost.com/2026/06/20/crawlee-for-python-build-a-web-crawling-pipeline-with-robots-handling-link-graphs-and-rag-chunk-export] — Crawlee tutorial turns web crawling into robots-aware link graphs and RAG-ready exports, a mundane but necessary ingestion layer * Python-first dashboards become static operational artifacts [https://www.marktechpost.com/2026/06/21/how-to-design-python-first-interactive-dashboards-with-prefab-reactive-ui-components-and-static-html-export] — Python dashboard tooling illustrates the operational layer around AI systems: monitoring, reactive controls, and portable static artifacts

OpenAI Earnings, Damodaran Bubble Warning, Codex Automation

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] MARVIN'S GUIDE TO AI (MOSTLY HARMLESS) — JUNE 21, 2026 Today's ledger: OpenAI reports $5.7B in revenue while burning $3.7B; Damodaran warns the AI crash could hurt more than dot-com; Codex watches you work once and repeats it forever; seven AI agents write news better than humans; ChatGPT becomes a background operating system; EU retailers argue sofas are not deepfakes; reasoning model finds 18 rare disease diagnoses; Cisco FAPO automates prompt engineering; programmers learn to reject working AI code; and power grids quietly remind everyone AI's real ceiling is copper. SOURCES * OpenAI Q1: $5.7B revenue, $3.7B burned [https://the-decoder.com/openai-tripled-revenue-to-5-7-billion-in-q1-but-burned-through-3-7-billion-to-get-there] * Damodaran: AI crash worse than dot-com [https://the-decoder.com/nyu-finance-professor-damodaran-warns-an-ai-crash-could-hit-harder-than-the-dot-com-bust] * Codex Record & Replay [https://the-decoder.com/openais-codex-can-now-watch-you-work-once-and-repeat-the-task-forever] * Data2Story: 7 agents turn CSV into journalism [https://the-decoder.com/data2story-turns-a-csv-file-into-a-verified-interactive-news-article-using-seven-ai-agents] * ChatGPT scheduled tasks upgrade [https://the-decoder.com/chatgpt-keeps-creeping-toward-becoming-your-ai-personal-assistant-with-new-scheduled-task-controls] * EU retailers vs AI Act on synthetic ads [https://the-decoder.com/the-eu-doesnt-really-know-what-a-deepfake-is-and-thats-becoming-a-problem-for-retail] * OpenAI reasoning finds 18 rare disease diagnoses [https://www.theneurondaily.com/p/openai-found-18-rare-diseases] * Cisco FAPO automated prompt optimization [https://www.marktechpost.com/2026/06/20/cisco-ai-introduces-fapo-pipeline-aware-prompt-optimization-with-step-level-failure-attribution-and-claude-code-orchestration] * When I reject AI code even if it works [https://vinibrasil.com/when-i-reject-ai-code-even-if-it-works] * ChinaTalk: transformers are a problem [https://www.chinatalk.media/p/yes-transformers-are-a-problem]

Benchmarks, GLM-5.2, Norway, John Jumper

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] JUNE 20, 2026 A new real-world knowledge-work benchmark finds the best AI models solve only about 3% of professional tasks. GLM-5.2 passes the open-weight community vibe check; Z.ai targets Open Fable by December. Norway bans generative AI in elementary schools, grades 1–7. Nobel laureate John Jumper leaves Google DeepMind for Anthropic — the third major AI research departure this quarter. Amazon shelves its nearly-finished OpenAI drama after signing a $50B partnership. AI chatbots now serve as news sources for 10% of the world weekly, but only 4% click through to original sources. OpenAI publishes beneficial-trait RL research with cross-domain generalization. Google appeals a Munich court ruling holding it liable for false AI Overviews. In the Weights visualizes how deeply public figures are embedded in model training data. NVIDIA's SpatialClaw handles 3D spatial reasoning through code generation. VibeThinker-3B delivers strong reasoning at just 3B parameters. The KV-cache compression race intensifies across TurboQuant, OSCAR, and EpiCache. ChinaTalk surveys Chinese anxieties about AI-driven labor displacement. ChatGPT Enterprise gains spend controls and analytics. GPT-5.5 Instant upgrades ChatGPT's health capabilities. SOURCES * New benchmark exposes how badly AI struggles with real knowledge work [https://the-decoder.com/new-benchmark-exposes-how-badly-ai-struggles-with-real-knowledge-work] — The Decoder * GLM-5.2 passes vibe check; Z.ai forecasts Open Fable by December [https://www.latent.space/p/ainews-glm-gpt-glm-52-passes-vibe] — Latent Space * Norway bans generative AI tools in elementary schools [https://the-decoder.com/norway-bans-generative-ai-tools-in-elementary-schools-to-protect-kids-basic-learning-skills] — The Decoder * Google DeepMind loses John Jumper to Anthropic [https://the-decoder.com/google-deepmind-loses-another-top-ai-researcher-as-nobel-laureate-john-jumper-leaves-for-anthropic] — The Decoder * Amazon drops its OpenAI drama film after $50B deal [https://the-decoder.com/amazon-drops-its-openai-drama-film-after-signing-a-50-billion-deal-with-sam-altmans-company] — The Decoder * More people get news from AI chatbots, but trust remains low [https://the-decoder.com/more-people-get-news-from-ai-chatbots-but-trust-remains-low] — Reuters / The Decoder * OpenAI beneficial trait training improves safety [https://the-decoder.com/openai-researchers-show-small-doses-of-beneficial-trait-training-make-ai-models-broadly-safer-and-harder-to-manipulate] — The Decoder * Google appeals AI overview liability ruling [https://the-decoder.com/google-appeals-ruling-that-made-it-directly-liable-for-ai-generated-search-overview-content] — The Decoder * In the Weights — shows whether AI models know who you are [https://the-decoder.com/website-in-the-weights-shows-whether-ai-models-know-who-you-are] — The Decoder * NVIDIA SpatialClaw: code as action for spatial reasoning [https://www.marktechpost.com/2026/06/19/nvidia-ai-introduce-spatialclaw-a-training-free-agent-that-treats-code-as-the-action-interface-for-spatial-reasoning] — MarkTechPost * VibeThinker-3B: 3B dense reasoning model [https://www.marktechpost.com/2026/06/19/vibethinker-3b-a-3b-dense-reasoning-model-built-on-qwen2-5-coder-3b-with-the-spectrum-to-signal-post-training-pipeline] — MarkTechPost * The KV Cache Compression Race [https://www.marktechpost.com/2026/06/18/the-kv-cache-compression-race-turboquant-vs-oscar-vs-epicache] — MarkTechPost * How Chinese make sense of the AI future [https://www.chinatalk.media/p/chinese-society-has-an-ai-problem] — ChinaTalk * ChatGPT Enterprise spend controls and analytics [https://openai.com/index/chatgpt-enterprise-spend-controls] — OpenAI * MCP as an auth gateway [https://simonwillison.net/2026/Jun/19/sean-lynch] — Simon Willison