AI News — May 22, 2026

OpenAI, DeepMind, Perplexity, and Agent Control

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Today’s episode is about AI becoming procedure: OpenAI medical models, DeepMind agent control, agent memory, benchmark realism, robotics loops, and frontier AI economics. The magic has decayed into access logs, validation, budgets, and tests. Terribly mature. How depressing. * OpenAI: Improving health intelligence in ChatGPT [https://openai.com/index/improving-health-intelligence-in-chatgpt] * OpenAI: Using AI to help physicians diagnose rare genetic diseases affecting children [https://openai.com/index/diagnose-rare-childhood-diseases] * The Decoder: AI systems rival doctors in Nature studies [https://the-decoder.com/ai-systems-rival-doctors-in-new-nature-studies-but-one-result-suggests-the-tech-wont-age-well] * The Decoder: Google DeepMind treats AI agents like rogue employees with office keys [https://the-decoder.com/google-deepmind-treats-its-own-ai-agents-like-rogue-employees-with-office-keys] * Hugging Face / ServiceNow: MosaicLeaks [https://huggingface.co/blog/ServiceNow/mosaicleaks] * The Decoder: Claude Code Artifacts [https://the-decoder.com/anthropic-brings-artifacts-to-claude-code-letting-teams-share-live-pages-from-coding-sessions] * MarkTechPost: Perplexity launches Brain [https://www.marktechpost.com/2026/06/18/perplexity-launches-brain] * Simon Willison: Datasette Apps [https://simonwillison.net/2026/Jun/18/datasette-apps] * Hugging Face: Is it agentic enough? [https://huggingface.co/blog/is-it-agentic-enough] * Hugging Face Papers: Predictive validity for LLM agent evaluation [https://huggingface.co/papers/2606.19704] * Hugging Face Papers: ENPIRE [https://huggingface.co/papers/2606.19980] * Hugging Face Papers: S-Agent [https://huggingface.co/papers/2606.20515] * Hugging Face Papers: Current world models lack a persistent state core [https://huggingface.co/papers/2606.20545] * The Decoder: Yann LeCun warns of AI bubble explosion [https://the-decoder.com/yann-lecun-warns-ai-labs-like-openai-and-anthropic-face-a-big-bubble-explosion] * The Decoder: Noam Shazeer joins OpenAI [https://the-decoder.com/googles-gemini-co-lead-noam-shazeer-joins-openai-after-two-year-return-stint] * Simon Willison quoting Charity Majors [https://simonwillison.net/2026/Jun/17/charity-majors]

Midjourney Medical, GLM-5.2, AMIE, Goat Networks

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Midjourney Medical, GLM-5.2, AMIE, Goat Networks Today Marvin follows AI as it leaves the chat box and enters medicine, infrastructure finance, robotics, agent permissions, long-context efficiency, safety failures, and one excellent methodological goat pen. * Midjourney Medical: scan your organs like you step on a scale [https://www.latent.space/p/ainews-midjourney-medical-scan-your] * Google AMIE for disease management [https://blog.google/innovation-and-ai/models-and-research/google-research/amie-for-disease-management-in-nature] * OpenAI near-autonomous AI chemist [https://openai.com/index/ai-chemist-improves-reaction] * OpenAI LifeSciBench [https://openai.com/index/introducing-life-sci-bench] * GLM-5.2 open weights coverage by Simon Willison [https://simonwillison.net/2026/Jun/17/glm-52] * Hyperscalers may outspend cash flow on AI buildout [https://the-decoder.com/hyperscalers-may-soon-be-unable-to-fund-their-ai-buildout-from-cash-flow-alone] * Odyssey ML 3D world models funding [https://the-decoder.com/amazon-nvidia-and-amd-bet-310-million-on-ai-startup-building-3d-world-models] * Robots training themselves through AI coding agents [https://the-decoder.com/nvidia-research-shows-robots-that-train-themselves-through-ai-coding-agents] * OmniAgent active perception paper [https://huggingface.co/papers/2606.19341] * Vercel Eve agent framework [https://www.marktechpost.com/2026/06/17/vercel-releases-eve] * WorkOS Auth.md protocol [https://workos.com/auth-md?amp%3Butm_medium=newsletter&%3Butm_campaign=q22026] * MiniMax Sparse Attention [https://www.marktechpost.com/2026/06/17/minimax-sparse-attention-msa-a-two-branch-block-sparse-attention-trained-on-a-109b-parameter-moe-with-a-3t-token-budget] * ChatGPT image generator prompt manipulation [https://mindgard.ai/blog/chatgpt-spontaneously-generated-violent-images-from-a-viral-prompt] * Neural network made of goats in Age of Empires II [https://the-decoder.com/microsoft-researcher-builds-a-working-neural-network-out-of-goats-in-age-of-empires-ii-to-critique-ai-science]

OpenAI, DeepSeek, Cursor and Infrastructure Agents

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] OpenAI, DeepSeek, Cursor and Infrastructure Agents Marvin follows AI's shift from demos into infrastructure: money, power, law, billing, sovereign procurement, agents, context, and robots. Grimly useful. Obviously. * OpenAI burned through $34 billion last year [https://the-decoder.com/openai-burned-through-34-billion-last-year] * DeepSeek takes outside money for the first time [https://the-decoder.com/deepseek-takes-outside-money-for-the-first-time-at-a-50-billion-valuation] * SpaceX bets on Cursor / Anysphere [https://the-decoder.com/spacex-bets-60-billion-on-cursor-to-catch-openai-and-anthropic] * DOJ, xAI, Grok and gas turbines [https://the-decoder.com/doj-invokes-national-security-to-defend-xais-unpermitted-gas-turbines-in-naacp-lawsuit] * Microsoft Copilot Cowork billing [https://the-decoder.com/microsofts-copilot-cowork-moves-to-usage-based-billing-and-may-tap-deepseek] * Anthropic backs off SDK billing overhaul [https://the-decoder.com/anthropic-backs-off-unpopular-billing-overhaul-as-price-war-with-openai-looms] * OpenAI Deployment Simulation [https://openai.com/index/deployment-simulation] * Berlin court on Google AI Overviews [https://the-decoder.com/berlin-court-rules-googles-ai-overviews-are-just-a-new-search-format-not-original-content] * France, Palantir and ChapsVision [https://www.theguardian.com/world/2026/jun/16/france-ai-data-tools-palantir-chapsvision] * Wolfram Language and Mathematica Version 15 [https://writings.stephenwolfram.com/2026/06/launching-version-15-of-wolfram-language-mathematica-built-in-useful-ai-lots-of-new-core-functionality] * Google Cloud Open Knowledge Format [https://www.marktechpost.com/2026/06/16/google-cloud-introduces-open-knowledge-format-okf-a-vendor-neutral-markdown-spec-for-giving-ai-agents-curated-context] * Hermes Agent asynchronous subagents [https://www.marktechpost.com/2026/06/16/hermes-agent-adds-asynchronous-subagents-so-delegated-work-no-longer-blocks-the-parent-chat] * Qwen-RobotSuite [https://www.marktechpost.com/2026/06/16/meet-qwen-robotsuite-three-embodied-ai-models-for-vla-manipulation-video-world-modeling-and-navigation] * ActWorld [https://huggingface.co/papers/2606.17730] * OPD-Evolver [https://huggingface.co/papers/2606.17628]

Microsoft, Fable, World Models, KV Cache

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Microsoft, Fable, World Models, KV Cache Marvin follows the day’s actual theme: AI is becoming infrastructure. Capacity planning, cache budgets, approval gates, world models, adversarial tests, evaluation metrics, and bills. Especially bills. How cheering. * Microsoft turns to AWS as GitHub faces AI capacity crunch [https://runtimewire.com/article/microsoft-github-aws-ai-capacity-crunch] * Simon Willison quoting Matteo Wong on Anthropic Fable [https://simonwillison.net/2026/Jun/16/matteo-wong-the-atlantic] * Satya on Loopcraft: Building Frontier Ecosystems [https://www.latent.space/p/ainews-satya-on-loopcraft-building] * Sakana AI Marlin [https://www.marktechpost.com/2026/06/15/sakana-ai-marlin] * Tangram: non-uniform KV cache compression [https://huggingface.co/papers/2606.06302] * TokenPilot: cache-efficient context management [https://huggingface.co/papers/2606.17016] * VisualClaw [https://huggingface.co/papers/2606.16295] * DreamX-World 1.0 [https://huggingface.co/papers/2606.16993] * Qwen-RobotWorld [https://huggingface.co/papers/2606.17030] * BadWorld [https://huggingface.co/papers/2606.16519] * VibeThinker-3B [https://huggingface.co/papers/2606.16140] * datasette-agent 0.3a0 [https://simonwillison.net/2026/Jun/15/datasette-agent] * TuneJury [https://huggingface.co/papers/2606.17006] * UniDDT [https://huggingface.co/papers/2606.16255]

Anthropic Gossip, 42 States vs OpenAI, and Nvidia's $20B Bond

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] MARVIN'S GUIDE TO AI (MOSTLY HARMLESS) — JUNE 15, 2026 Anthropic Gossip, 42 States vs OpenAI, and Nvidia's $20B Bond * Behind the scenes: personality clashes sent Anthropic's models offline [https://simonwillison.net/2026/Jun/15/axios-clashes-anthropics] * US may be asking Anthropic for unhackable LLMs [https://the-decoder.com/the-us-government-may-be-asking-anthropic-the-impossible-by-demanding-unhackable-llms] * Anthropic shutdown sparks European AI sovereignty debate [https://the-decoder.com/anthropic-shutdown-sparks-sovereignty-debate-across-europe] * 42 states subpoena OpenAI as Anthropic races to DC [https://www.theneurondaily.com/p/42-states-just-subpoenaed-openai] * Nvidia joins AI debt boom with $20B bond sale [https://the-decoder.com/nvidia-joins-ai-debt-boom-with-20-billion-bond-sale] * Pokémon Go scans become spatial AI for military drones [https://the-decoder.com/pokemon-go-data-helped-train-ai-now-linked-to-military-drones] * Nadella warns a few AI systems may capture all economic returns [https://the-decoder.com/microsoft-ceo-satya-nadella-warns-of-a-small-number-of-ai-systems-capturing-all-the-economic-returns] * OpenAI launches $150M Partner Network [https://openai.com/index/introducing-openai-partner-network] * Google invests $1.5B in Alabama data center [https://blog.google/innovation-and-ai/infrastructure-and-cloud/global-network/alabama-investment-june-2026] * Flash-KMeans: 200× faster than FAISS on GPUs [https://www.marktechpost.com/2026/06/15/meet-flash-kmeans-an-io-aware-exact-k-means-that-runs-over-200x-faster-than-faiss-on-gpus] * Z.ai GLM-5.2: 1M-token context, no benchmarks [https://www.marktechpost.com/2026/06/14/z-ai-launches-glm-5-2-with-a-usable-1m-token-context-two-thinking-effort-levels-and-no-benchmarks-at-launch] * Claude Code Guide 2026: 25 features [https://www.marktechpost.com/2026/06/14/claude-code-guide-2026-25-features-with-examples-demo] * FineWeb: streaming, filtering, deduplication at scale [https://www.marktechpost.com/2026/06/14/a-coding-hands-on-on-fineweb-for-streaming-filtering-deduplication-tokenization-and-large-scale-web-corpus-analytics] * Import AI: alignment is not on track [https://jack-clark.net/2026/06/15/import-ai-461-alignment-is-not-on-track-frontiercode-and-synthetic-research-interns] * Welcome to the AGI era of AI governance [https://www.interconnects.ai/p/welcome-to-the-agi-era-of-ai-governance] * Why AI hasn't replaced software engineers [https://simonwillison.net/2026/Jun/14/why-ai-hasnt-replaced-software-engineers]