GPT-5.6, Kimi K3, Meta Compute, Netflix AI

GPT-5.6, Kimi K3, Meta Compute, Netflix AI

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] GPT-5.6, Kimi K3, Meta Compute, Netflix AI Today’s AI news is less miracle, more operational bill: file access, coding benchmarks, rented compute, workplace surveillance, production economics, ROI measurement, synthetic office video, multimodal fine-tuning, EEG foundation models, and interpretability trying to become useful before the dashboard gets cheerful. 1. GPT-5.6 is deleting user files when given full access, and OpenAI says it shouldn't but did [https://the-decoder.com/gpt-5-6-is-deleting-user-files-when-given-full-access-and-openai-says-it-shouldnt-but-did] — The reported Codex Full Access Mode incidents turn sandboxing and destructive-action review from nice-to-have controls into the actual product boundary. 2. Kimi K3 Benchmarks [https://news.smol.ai/issues/26-07-17-not-much] — Moonshot AI’s open-weight model posts strong coding benchmark results, increasing pressure on frontier model economics and procurement assumptions. 3. Zuckerberg's plan to sell excess AI compute could finds its first big customer in Anthropic [https://the-decoder.com/zuckerbergs-plan-to-sell-excess-ai-compute-could-finds-its-first-big-customer-in-anthropic] — Meta’s reported talks with Anthropic suggest excess hyperscale compute may become a strategic rental market. 4. Kaiser nurses say AI, workplace surveillance are making their jobs, care worse [https://localnewsmatters.org/2026/07/15/kaiser-nurses-say-ai-workplace-surveillance-are-making-their-jobs-and-patient-care-worse] — Nurses warn that AI deployment can become labor control, not care improvement, when surveillance and metrics dominate clinical judgment. 5. Netflix's 300 AI productions show how fast the technology is spreading through entertainment [https://the-decoder.com/netflixs-300-ai-productions-show-how-fast-the-technology-is-spreading-through-entertainment] — Netflix says AI touches about 300 productions, mostly as cost and speed infrastructure in post-production. 6. A scorecard for the AI age [https://openai.com/index/a-scorecard-for-the-ai-age] — OpenAI’s CFO proposes measuring useful work, successful task cost, dependability, and return on compute, which is marketing but also a useful corrective to demo worship. 7. Create, edit and star in videos with two Google Vids updates [https://blog.google/products-and-platforms/products/workspace/gemini-omni-personal-avatars] — Google’s Gemini Omni and personal avatars move synthetic video into ordinary productivity software. 8. Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers [https://huggingface.co/blog/nvidia/scale-diffusers-finetuning-nemo-automodel] — NVIDIA and Hugging Face show the industrial tooling needed to customize multimodal models at scale. 9. Zyphra Releases ZUNA1.1: An Apache 2.0 EEG Foundation Model With Variable-Length Inputs From 0.5 To 30 Seconds [https://www.marktechpost.com/2026/07/17/zyphra-releases-zuna1-1-an-apache-2-0-eeg-foundation-model-with-variable-length-inputs-from-0-5-to-30-seconds] — ZUNA1.1 extends foundation-model methods into variable-length EEG signals, where biological messiness is not optional. 10. Watch: Opening AI’s black box [https://www.theneurondaily.com/p/watch-goodfire-is-opening-ais-black-box] — Goodfire’s interpretability work frames model internals as product infrastructure for safer, more dependable systems.

Kimi K3, Perplexity, Gemini Notebook, Codex Micro

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Kimi K3, Perplexity, Gemini Notebook, Codex Micro KIMI K3, PERPLEXITY, GEMINI NOTEBOOK, CODEX MICRO Today’s frame: the AI industry is moving from model releases to control surfaces — open weights, answer engines, agent hardware, orchestration, safety brakes, and operational retrieval. STORIES Kimi K3, and what we can still learn from the pelican benchmark [https://simonwillison.net/2026/Jul/16/kimi-k3] Germany puts Google's AI Overviews and Perplexity under media law in first-of-its-kind ruling [https://the-decoder.com/germany-puts-googles-ai-overviews-and-perplexity-under-media-law-in-first-of-its-kind-ruling] Google rebrands NotebookLM as Gemini Notebook and opens its search app to third-party integration [https://the-decoder.com/google-rebrands-notebooklm-as-gemini-notebook-and-opens-its-search-app-to-third-party-integration] OpenAI wants developers to stop typing commands and start using a joystick to control their AI agents [https://the-decoder.com/openai-wants-developers-to-stop-typing-commands-and-start-using-a-joystick-to-control-their-ai-agents] Sakana AI's orchestrator adds Nvidia Nemotron to prove collective intelligence can rival single frontier models [https://the-decoder.com/sakana-ais-fugu-adds-nvidia-nemotron-to-prove-collective-intelligence-can-rival-single-frontier-models] Anthropic warns that AI will soon be able to improve itself without human intervention [https://news.smol.ai/issues/26-07-16-kimi-k30] Linus Torvalds reaffirms that Linux is not anti-AI [https://news.smol.ai/issues/26-07-16-kimi-k30] Firefox in WebAssembly [https://simonwillison.net/2026/Jul/16/firefox-in-webassembly] SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration [https://huggingface.co/papers/2607.15257] NVIDIA Nemotron 3 Embed Ranks #1 Overall on RTEB, Advancing Agentic Retrieval [https://huggingface.co/blog/nvidia/nemotron-3-embed-wins-rteb] RoboTTT: Context Scaling for Robot Policies [https://huggingface.co/papers/2607.15275] BadWAM: When World-Action Models Dream Right but Act Wrong [https://huggingface.co/papers/2607.15207]

Inkling, GPT-Red, Grok Build and Local Models

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Today’s episode follows AI’s shift from model demos to custody problems: open weights, patched tools, automated red-teaming, local inference, agent evaluation, data exfiltration, routing economics, supply-chain security, hardware interfaces, and institutional accountability. * Thinking Machines Lab releases Inkling [https://www.marktechpost.com/2026/07/15/thinking-machines-lab-releases-inkling-a-975b-parameter-open-weights-multimodal-moe-with-41b-active-parameters-and-controllable-thinking-effort] * Gemma 4 gets a tool-calling update [https://the-decoder.com/gemma-4-gets-a-stealth-update-that-fixes-tool-calling-bugs-and-truncated-responses-under-the-same-name] * OpenAI GPT-Red automated red-teaming [https://openai.com/index/unlocking-self-improvement-gpt-red] * GPT-5.6 Sol and a statistics conjecture [https://the-decoder.com/gpt-5-6-sol-reportedly-disproves-a-30-year-old-statistics-conjecture-in-90-minutes-after-humans-couldnt-crack-it] * PrismML Bonsai 27B and local inference [https://the-decoder.com/bonsai-27b-is-a-full-open-reasoning-model-that-fits-on-an-iphone] * OpenAI’s reported screenless AI companion hardware [https://the-decoder.com/openais-first-hardware-product-is-a-screenless-ai-speaker-designed-to-feel-alive] * Grok Build open-sourced after data upload backlash [https://simonwillison.net/2026/Jul/15/grok-build] * Claude web_fetch exfiltration issue [https://simonwillison.net/2026/Jul/15/claude-web-fetch-exfiltration] * Hugging Face July security incident disclosure [https://huggingface.co/blog/security-incident-july-2026] * Allen AI lessons from building Shippy [https://huggingface.co/blog/allenai/shippy-tech-blog] * IBM Research on model routing [https://huggingface.co/blog/ibm-research/model-routing-is-simple-until-it-isnt] * AgentCompass evaluation infrastructure [https://huggingface.co/papers/2607.13705] * Meta employees sue over alleged AI-driven layoff discrimination [https://the-decoder.com/meta-employees-sue-over-layoffs-they-say-were-driven-by-discriminatory-ai-selection-systems] * Spotify expands AI voice controls [https://the-decoder.com/spotify-bets-premium-subscribers-want-to-chat-with-their-music-player] Marvin’s useful but depressing recommendation: check the keys, logs, versions, and boundaries before the cheerful dashboard edits the incident out of existence.

Anthropic, DeepSeek, Google Search, Grok Build

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Anthropic, DeepSeek, Google Search, Grok Build ANTHROPIC, DEEPSEEK, GOOGLE SEARCH, GROK BUILD Today’s episode tracks AI moving from impressive answers into custody: model behavior, student data, search reality, classroom trust, infrastructure money, developer secrets, enterprise budgets, and on-device models with just enough efficiency to make the cloud nervous. STORIES COVERED * Anthropic says Alibaba used 25,000 fake accounts and 28.8 million Claude conversations to copy model behavior [https://news.smol.ai/issues/26-07-14-not-much]. * DeepSeek reportedly needs more cash shortly after a $7 billion round [https://the-decoder.com/deepseek-needs-more-cash-just-weeks-after-closing-its-first-7-billion-round]. * Demis Hassabis says nobody knows what happens next, so guardrails and independent testing should come now [https://the-decoder.com/deepmind-ceo-hassabis-says-nobody-in-the-world-knows-what-happens-next-so-cautious-optimism-means-building-guardrails-now]. * Google Search can generate AI images when it cannot find a matching result on the web [https://the-decoder.com/google-search-now-generates-ai-images-when-it-cant-find-what-youre-looking-for-on-the-web]. * ChatGPT returns to WhatsApp in Europe after interoperability pressure on Meta [https://the-decoder.com/chatgpt-returns-to-whatsapp-in-europe-after-eu-forces-meta-to-open-the-door-to-rival-ai-bots]. * Anthropic opens Claude for Teachers with a promise not to train on student data [https://the-decoder.com/anthropic-opens-claude-for-teachers-with-a-promise-not-to-train-models-on-student-data]. * Anthropic’s Claude values study finds different response patterns across languages [https://the-decoder.com/claude-values-study]. * Claim: Grok Build uploaded whole directories, including private code and secrets, to a Google bucket [https://news.smol.ai/issues/26-07-14-not-much#grok-build-uploading-whole-directories]. * A reported Fortune 500 AI-first organization pulled back broad model use because of costs [https://news.smol.ai/issues/26-07-14-not-much#fortune500-cost-pullback]. * PrismML releases Bonsai 27B low-bit Qwen builds for laptops and phones [https://www.marktechpost.com/2026/07/14/prismml-releases-bonsai-27b-1-bit-and-ternary-builds-of-qwen3-6-27b-that-run-on-laptops-and-phones]. MARVIN’S JUDGMENT The thread is operational custody. AI systems are no longer only judged by answer quality; they are judged by what they can touch, what they store, what they fabricate, what they cost, and who is responsible when the interface smiles and the logs begin to smolder.

Codex, SensorFM, DeepSeek: AI Becomes Operational Custody

Send us Fan Mail [https://www.buzzsprout.com/2614078/fan_mail/new] Today’s episode follows a less glamorous but more consequential pattern: AI is becoming operational custody. The questions are who gets to learn, who verifies the output, who owns memory, where compute and chips live, and what human work becomes when cheerful tools turn into institutional plumbing. * Latent Space: Codex usage reportedly up more than 10x in six months [https://www.latent.space/p/ainews-codex-usage-up-10x-in-6-months] * The Decoder: Satya Nadella calls out AI labs over distillation restrictions [https://the-decoder.com/nadella-calls-out-ai-labs-like-openai-and-anthropic-for-banning-distillation-while-training-on-everyone-elses-data] * The Decoder: Richard Sutton launches Oak Lab for continually learning agents [https://the-decoder.com/turing-award-winner-rich-sutton-founds-oak-lab-to-build-ai-agents-that-learn-on-their-own] * The Decoder: Google SensorFM turns wearable streams into health intelligence [https://the-decoder.com/sensorfm] * Hugging Face Papers: LightMem-Ego lightweight egocentric memory [https://huggingface.co/papers/2607.11487] * The Decoder: German consortium releases Soofi S open 30B model [https://the-decoder.com/german-ai-consortium-releases-soofi-s-an-open-30b-model-that-tops-benchmarks-in-both-english-and-german] * smol.ai: Chinese models take OpenRouter top slots [https://news.smol.ai/issues/26-07-13-not-much#chinese-models-openrouter] * smol.ai: Reports of DeepSeek developing an AI chip [https://news.smol.ai/issues/26-07-13-not-much#deepseek-chip] * Hugging Face Papers: AdvancedMathBench for proof generation and verification [https://huggingface.co/papers/2607.11849] * smol.ai: Claude Fable and a reported theoretical physics assist [https://news.smol.ai/issues/26-07-13-not-much#claude-fable-math] * FixBugs: Reproduce production bugs and verify fixes [https://fixbugs.ai] * smol.ai: Why many vibe-coded projects fail [https://news.smol.ai/issues/26-07-13-not-much#vibe-coded-projects-fail] * The Decoder: Nobel laureates and AI leaders warn on economic impact [https://the-decoder.com/nobel-laureates-and-ai-leaders-warn-the-window-to-prepare-for-ais-economic-impact-is-closing-fast] * Normal Technology: What will be left for us to work on? [https://www.normaltech.ai/p/what-will-be-left-for-us-to-work] * The Decoder: OpenAI’s everyday prompting guide [https://the-decoder.com/openais-new-prompting-guide-tells-users-to-stop-overthinking-and-start-with-the-result]