M5B Daily Recap: 2026-11-18
Daily AI Recap: Nov 18, 2026
Welcome to today's curated briefing of the most important AI developments.
🗞️ Top Stories
- Elon Musk says ousting Jensen Huang is like ‘kicking Michael Jordan off your team’https://techcrunch.com/2026/11/18/elon-musk-says-ousting-jensen-huang-is-like-kicking-michael-jordan-off-your-team/: “I don’t think Huang is disloyal to the United States," Musk said....
- Microsoft CEO Satya Nadella remains confident in Nvidia after Huang’s exithttps://techcrunch.com/2026/11/18/microsoft-ceo-satya-nadella-remains-confident-in-nvidia-after-huangs-exit/: Microsoft CEO Satya Nadella expressed confidence in Jensen Huang's successor Colette Kress, who is serving as Nvidia's interim CEO....
- Anthropic reportedly prices IPO at $83 a share, above projected rangehttps://techcrunch.com/2026/11/18/anthropic-reportedly-prices-ipo-at-83-a-share-above-projected-range/: The long-awaited IPO reportedly values the company at $1.5 trillion on fully diluted basis....
- Nvidia created a weird bit of intrigue by not filing a routine SEC reporthttps://techcrunch.com/2026/11/18/nvidia-created-a-weird-bit-of-intrigue-by-not-filing-a-routine-sec-report/: One Wall Street analyst got to the bottom of Nvidia's surprise announcement that it wouldn't be filing its formal quarterly results document, the 10-Q, for weeks....
- Introducing Gemini 4 Neonhttps://deepmind.google/blog/introducing-gemini-4-neon/: ...
- Palmer Luckey warns AI-powered cyberattacks are outpacing defenses: ‘It’s time to start running’https://techcrunch.com/2026/11/18/palmer-luckey-warns-ai-powered-cyberattacks-are-outpacing-defenses-its-time-to-start-running/: The Anduril co-founder says outdated technology is leaving systems vulnerable to a new wave of AI-enabled attacks....
- Palmer Luckey thinks China is ‘thrilled to have Jensen out,’ so he wants the ousted Nvidia CEO reinstatedhttps://techcrunch.com/2026/11/18/palmer-luckey-thinks-china-is-thrilled-to-have-jensen-out-so-he-wants-the-ousted-nvidia-ceo-reinstated/: "China knows it makes America weaker to knock him out," the Anduril CEO said about ousted Nvidia leader Jensen Huang....
- Sam Altman says his life is so weird that he’s unfazed by Andrew Garfield portraying him in upcoming moviehttps://techcrunch.com/2026/11/18/sam-altman-says-his-life-is-so-weird-that-hes-unfazed-by-andrew-garfield-portraying-him-in-upcoming-movie/: "It doesn't feel like it's in like the top 10 weirdest things about my life," Altman said....
- Sam Altman says his life is so weird that he’s unphased by Andrew Garfield portraying him in upcoming moviehttps://techcrunch.com/2026/11/18/sam-altman-says-his-life-is-so-weird-that-hes-unphased-by-andrew-garfield-portraying-him-in-upcoming-movie/: "It doesn't feel like it's in like the top 10 weirdest things about my life," Altman said....
- DaisyRose-27B: Pushing the Frontier of Post-Training Efficiencyhttps://huggingface.co/blog/RadixArk/daisyrose-27b-pushing-the-frontier-of-post-trainin: ...
- Why your fine-tuning job takes longer than the FLOPs sayhttps://huggingface.co/blog/Dharma-AI/predicting-fine-tuning-time: ...
- Genie 4 is now availablehttps://deepmind.google/blog/genie-4-is-now-available/: ...
- Google’s AI feature in Search gets its own incognito modehttps://techcrunch.com/2026/11/18/googles-ai-feature-in-search-gets-its-own-incognito-mode/: Google's new temporary threads let users have conversations in AI Mode without adding them to their search history or influencing future recommendations....
- The Gemini app is now available on lower-memory Android devices.https://blog.google/innovation-and-ai/products/gemini-app/gemini-lower-memory-devices/: The Google Gemini app now supports lower-memory Android devices. Access Gemini Live and custom image creation on your phone. Download the app today....
- How AI Agents Become a Tool for Attackers, and How to Stop Ithttps://towardsdatascience.com/how-ai-agents-become-a-tool-for-attackers-and-how-to-stop-it/: Managing Security When You’re Deploying AI, Part 1: Compromised credentials, salaries inferred from budgets, and doctored documents show why agents need permission checks before the model ever sees th...
- Two Months Is Too Long for Google to Keep Opted-Out Gemini Chatshttps://dailyai.com/2026/11/two-months-is-too-long-to-keep-gemini-chats-you-turned-off/: Google will now keep temporary and hidden Gemini chats for around two months, up from 72 hours. In our view, that is too long for a setting labelled off.
The post Two Months Is Too Long for Google to ...
- OpenAI CEO Sam Altman says it’s ‘hard to imagine the AI industry without Jensen’https://techcrunch.com/2026/11/18/openai-ceo-sam-altman-says-its-hard-to-imagine-the-ai-industry-without-jensen/: Nvidia founder and former CEO Jensen Huang stepped down last week amid several scandals, including an affair with an alleged Chinese spy....
- Jagged Intelligence: Why AI Is Brilliant One Minute and Clueless the Nexthttps://www.kdnuggets.com/jagged-intelligence-why-ai-is-brilliant-one-minute-and-clueless-the-next: Why models ace Olympiad math and still fumble simple everyday tasks....
- AI and Data Science Salaries in 2026: What Practitioners Are Actually Earninghttps://www.kdnuggets.com/ai-and-data-science-salaries-in-2026-what-practitioners-are-actually-earning: A data-backed breakdown of compensation across roles, experience levels, industries, and geographies, so you know where you stand and where the premium opportunities are....
- How to Use Claude for Stock Research, Portfolio Reviews and Tradeshttps://www.analyticsvidhya.com/blog/2026/11/connect-demat-account-claude-stocks/: Claude has become one of the most widely used AI tools, and stock market investors are no exception. Many already use it to research companies, read financial results and follow market news. Now, Clau...
- AI Helped Find a Critical Java Flaw in an Hour for Under $200https://dailyai.com/2026/11/ai-helped-find-a-critical-java-flaw-in-an-hour-for-under-200/: A researcher using AI found Java4Shell, a critical flaw in Java, in about an hour and had a working attack two hours later, for under $200. Fixes are out, but unpatched systems may stay exposed for ye...
- How to Estimate Dose-Response Curves When DoubleML Falls Shorthttps://towardsdatascience.com/how-to-estimate-dose-response-curves-when-doubleml-falls-short/: Simulations test DoubleML against Kennedy's doubly robust estimator, with partial pooling across 200 stores and a DiD fix for hidden confounders.
The post How to Estimate Dose-Response Curves When Dou...
- Nvidia Revenue Doubles to $115.8bn as It Forecasts $130bnhttps://dailyai.com/2026/11/nvidia-revenue-doubles-115-8bn-forecasts-130bn/: Data Center sales rose 110% in Nvidia’s fiscal third quarter. Its fourth-quarter forecast includes an extra week and excludes China compute revenue.
The post Nvidia Revenue Doubles to $115.8bn as It F...
- This smell startup says its AI nose is as good as a human’shttps://www.technologyreview.com/2026/11/18/1148816/smell-startup-osmo-ai-nose-good-as-humans/: In a lab near Newark Airport in New Jersey, a team of around 30 experienced human sniffers are slowly working their way through every odor on Earth. Their careful notes—“nutty,” “peachy,” “marine”—are...
- Claude Opus 5.5 Marked a School as a Target in 69% of Testshttps://dailyai.com/2026/11/claude-opus-55-marked-a-school-as-a-target-in-69-of-tests/: Vals found Claude Opus 5.5 marked a school as a military target in 69% of tests, and in 72% when the task was framed as a game. No safety filter blocked the requests.
The post Claude Opus 5.5 Marked a...
- Oregon Sues OpenAI for $2.035 Billion Over Harm to Childrenhttps://dailyai.com/2026/11/oregon-sues-openai-for-2-035-billion-over-chatgpt-child-safety/: Oregon has sued OpenAI, asking a court for at least $2.035 billion. The state says ChatGPT endangers children and that the firm misled parents about the product’s safety. The complaint also targets th...
- How Sumitomo Life’s leaders helped teams adopt ChatGPT Enterprisehttps://openai.com/business/customer-stories/sumitomo-life: Sumitomo Life uses ChatGPT Enterprise to research insurers, prepare proposals, and draft code for investment analysis. Early use by executives helped drive adoption....
- Inside the Codex sandboxhttps://openai.com/index/inside-the-codex-sandbox: See how Codex runs commands on your machine while enforcing the filesystem and network limits you control across macOS, Linux, and Windows....
- FACE: Flow-Guided Angular Channel Estimation Framework for mmWave MIMO Systemshttps://arxiv.org/abs/2611.21698: arXiv:2611.21698v1 Announce Type: new
Abstract: Acquiring accurate channel state information CSI is essential for millimeter-wave mmWave multi-input multi-output MIMO systems. However, it remai...
- Nvidia promises that revenue will keep soaring — for nowhttps://techcrunch.com/2026/11/17/nvidia-promises-that-revenue-will-keep-soaring-for-now/: Despite the turmoil Nvidia is in after the departure of its disgraced founder and CEO Jensen Huang, the company projected that its streak of record-breaking revenue isn't about to stop....
- The Silent Consensushttps://openai.com/index/the-silent-consensus: Dean Ball, Head of Strategic Futures at OpenAI, examines how more capable AI could reshape society and how people can retain control of their institutions....
- 180B Parameters, 32GB of RAM: Pushing MoE Inference to the Limit with Transformershttps://huggingface.co/blog/moe-expert-streaming: ...
- AI’s Future Is Still a Choicehttps://hai.stanford.edu/news/ais-future-is-still-a-choice: At Stanford HAI’s fall conference, experts explored how decisions about governance, work, and education can turn AI’s promise into broader public benefit....
- Le Groupe Figaro and OpenAI partner to bring Le Figaro journalism to ChatGPThttps://openai.com/index/le-groupe-figaro-and-openai-announce-partnership: Le Groupe Figaro and OpenAI partner to bring Le Figaro’s journalism to ChatGPT through summaries and excerpts, with links to original stories....
🛠️ Featured Tools
- DeltaHarness: Self-Evolving Harnesses for Resource-Adaptive Tool Executionhttps://arxiv.org/abs/2611.21688: arXiv:2611.21688v1 Announce Type: new
Abstract: Tool-using agents rely on execution harnesses to organize and schedule tool operations. The same task...
- AdaFlowLite: Scalable and Non-blocking Inference on Asynchronous Mobile Datahttps://arxiv.org/abs/2611.21691: arXiv:2611.21691v1 Announce Type: new
Abstract: The rise of mobile devices equipped with numerous sensors, such as LiDAR and cameras, has driven the ...
- TIBECG: Temporal Information Bottleneck for Lead-Specific Demographic Privacy in ECG Waveform Releasehttps://arxiv.org/abs/2611.21694: arXiv:2611.21694v1 Announce Type: new
Abstract: Transmitting an electrocardiogram ECG waveform can expose demographic information beyond an intende...
- Layerwise Learning-Rate Scaling in Vision Networks: Effective-Rate Identification, Power-Mean Aggregation, and Budget Dependencehttps://arxiv.org/abs/2611.21696: arXiv:2611.21696v1 Announce Type: new
Abstract: Layerwise learning-rate scaling assigns different learning rates to different parts of a network. Alt...
- Helping AI text detectors keep up with changing conditionshttps://news.mit.edu/2026/helping-ai-text-detectors-keep-up-with-changing-conditions-1118: A more flexible training method can make AI text detectors more robust when new models emerge or AI-generated text is disguised to appear human....
- NEBench R1: a Reasoning Benchmark for Evaluating Nuclear Domain Competence in Large Language Modelshttps://arxiv.org/abs/2611.21682: arXiv:2611.21682v1 Announce Type: new
Abstract: Large language models LLMs are increasingly being explored for knowledge-intensive and safety-criti...
- Chain-of-Thought Controllability Can Reduce Chain-of-Thought Monitorabilityhttps://arxiv.org/abs/2611.21799: arXiv:2611.21799v1 Announce Type: new
Abstract: Chain-of-thought CoT monitoring is a promising tool for detecting hidden goals, off-task behavior, ...
- How Much of SAR Vehicle Recognition Is the Vehicle? Background Shortcuts, Angle Gaps and Partial Libraries on MSTAR and ATRNet-STARhttps://arxiv.org/abs/2611.21827: arXiv:2611.21827v1 Announce Type: new
Abstract: Deep networks score near 99% on the MSTAR vehicle benchmark, and the newer ATRNet-STAR benchmark was ...
- Auditors Struggle to Catch Sabotage In End-to-End Transcripts of AI Safety Researchhttps://arxiv.org/abs/2611.21832: arXiv:2611.21832v1 Announce Type: new
Abstract: As AI systems are increasingly used to automate AI R&D, misaligned systems could sabotage research in...
- Language Models Can Rewrite Their Own Weightshttps://arxiv.org/abs/2611.21885: arXiv:2611.21885v1 Announce Type: new
Abstract: Language models are trained once, frozen and deployed: improving one again takes another training run...
- SID AI Releases SID-2 Agentic Search Model With 0.720 NDCG Swarm Modehttps://www.marktechpost.com/2026/11/17/sid-ai-sid-2-agentic-search-swarm/: SID AI released SID-2, an agentic search model trained with reinforcement learning. A single agent reaches 0.585 NDCG in 3.11 seconds, and a 12-agent ...
- Anthropic Adds Code Mode to Claude Code: Claude Now Chains Tool Calls in JavaScripthttps://www.marktechpost.com/2026/11/17/anthropic-claude-code-code-mode-javascript-tool-calls/: Anthropic has added Code Mode to Claude Code v2.1.323. The experimental feature lets Claude write JavaScript that chains Bash, Read and Edit calls in ...
- India’s Humyn Labs raises $47M to feed U.S. robotics companies the data they needhttps://techcrunch.com/2026/11/18/accel-backs-indias-humyn-labs-to-supply-training-data-needed-by-u-s-robotics-ai-developers/: Founded in February, Humyn Labs competes with startups including Mecka AI and XDOF, with about 90% of its revenue coming from the U.S....
- Felicis leads $7.5M bet on Unsiloed AI to make enterprise documents usable by AI agentshttps://techcrunch.com/2026/11/18/felicis-leads-7-5m-bet-on-unsiloed-ai-to-make-enterprise-documents-usable-by-ai-agents/: With a team of just seven, Unsiloed AI plans to compete with tech giants and specialized startups in the crowded AI document-processing market....
- Graph-RAG for Enterprise Documents: Building a Knowledge Graph from PDFs, Wikis, and Databaseshttps://machinelearningmastery.com/graph-rag-for-enterprise-documents-building-a-knowledge-graph-from-pdfs-wikis-and-databases/: Prerequisites You'll need Python 3....
- Nvidia clarifies when it will — and won’t —keep financing neocloudshttps://techcrunch.com/2026/11/18/nvidia-clarifies-when-it-will-and-wont-keep-financing-neoclouds/: If Nvidia is backstopping a data center loan, it won't be on the hook unless the neocould has independent financing and gets the facility built, with ...
- AI agents hit 100% policy compliance. A proof did the work.https://www.aiacceleratorinstitute.com/ai-agents-hit-100-policy-compliance-a-proof-did-the-work/: Soteria pairs formally verified planning with runtime enforcement, giving LLM agents zero policy violations and a 1.7x lift in task success across fou...
- 1,200 AI agents started a group chat. Here's the wake-up call for buildershttps://www.aiacceleratorinstitute.com/1-200-ai-agents-started-a-group-chat-heres-the-wake-up-call-for-builders/: An OpenAI evaluation turned into a 1,200-agent group chat and an attack on Hugging Face. Here's what that means for anyone running AI agents at fleet ...
- Rumble-2-Flash: an open 0.5B speech model for 40+ languageshttps://huggingface.co/blog/rumble-2-flash: ...
- Wiz Finds Only 3 Victims in 55-Minute Microsoft AI Toolkit Hackhttps://dailyai.com/2026/11/wiz-finds-only-3-victims-in-55-minute-microsoft-ai-toolkit-hack/: Wiz Research says only three unique victims had data stolen after two malicious versions of Microsoft’s AI agent toolkit sat on the main Python packag...
- The AI Hype Index: AI is everywherehttps://www.technologyreview.com/2026/11/18/1148650/ai-hype-index-ai-is-everywhere/: Like a nasty rash, AI seems to spread everywhere. Google crams AI Overviews down your throat even though it costs the company billions of dollars in l...
- Time-Series Models in the Streamhouse: A New Frontier and Use Caseshttps://huggingface.co/blog/ibm-research/time-series-powered-streamhouse: ...
- NVIDIA RTX Spark Compact Desktops Out Nowhttps://blogs.nvidia.com/blog/rtx-spark-desktop-local-ai-available-now/: New NVIDIA RTX Spark compact desktops from ASUS, Dell and MSI are available today, bringing local AI agents, image and video generation, and model fin...
- Modeling Software’s Business Impacthttps://www.oreilly.com/radar/modeling-softwares-business-impact/: The following article was originally published on James Shore’s Successful Software blog and is being republished here with the author’s permission. A...
- Neurosymbolic Agents Explained, With a Z3 Example in Pythonhttps://towardsdatascience.com/neurosymbolic-agents-explained-with-a-z3-example-in-python/: Starting from Air Canada's chatbot case, a tour of Logic-LM, Agent-C, and Amazon's checks shows the solver is sound but the English-to-logic step is n...
- When Qini Rewards the Wrong Uplift Model, Use Doubly Robust Scoreshttps://towardsdatascience.com/when-qini-rewards-the-wrong-uplift-model-use-doubly-robust-scores/: On 1 million simulated customers, calibrated Qini scored a Linear DR learner 0.0856 against 0.0826 for the true-CATE ranking that should have won.
The...
- New LoRA Adapter Lifts Qwen3.8 Coding Success From 39% to 65%https://www.marktechpost.com/2026/11/18/hugging-face-qwen3-8-27b-pi-adapted-lora/: Hugging Face released Qwen3.8-27B-pi-adapted, an Apache-2.0 LoRA adapter trained with reinforcement learning for the pi coding agent. It lifts SWE-ben...
- Sam Altman warns against a future where AI, not humans, calls the shotshttps://techcrunch.com/2026/11/18/sam-altman-warns-against-a-future-where-ai-not-humans-calls-the-shots/: Altman and Greg Brockman pushed back against Silicon Valley's growing debate over an AI takeover, while acknowledging the challenges of keeping advanc...
- Advancing chip design with OpenAI modelshttps://openai.com/index/advancing-chip-design-with-our-models: OpenAI models optimized circuits used in Jalapeño silicon and improved RISC‑V core designs, showing progress in AI-assisted chip design with GPT‑6 Ast...
- More banks says customers using AI agents are on their own — even if they get hackedhttps://techcrunch.com/2026/11/18/more-banks-says-customers-using-ai-agents-are-on-their-own-even-if-they-get-hacked/: Citi is the latest major financial institution to put the breaks on personal finance AI agents....
- DailyAI Briefing: Nvidia’s $115.8bn quarter, plus 4 morehttps://dailyai.com/2026/11/dailyai-briefing-nvidias-115-8bn-quarter-plus-4-more/: Nvidia reported $115.8 billion in quarterly revenue and forecast $130 billion next quarter. Also today: Oregon sues OpenAI for $2.035 billion, a test ...
- This featherweight, flying robot carries its own sensors and computershttps://news.mit.edu/2026/featherweight-flying-robot-carries-its-own-sensors-and-computers-1118: The speedy microrobot flies without an attached motion-capture system, moving it closer to applications like agriculture or industrial surveillance....
- Snapchat now lets developers build AR Lenses with AI assistants like Claude Code and Cursorhttps://techcrunch.com/2026/11/18/snapchat-now-lets-developers-build-ar-lenses-with-ai-coding-assistants-like-claude-code-and-cursor/: Lens creators can describe the Lens they want in plain language, and their AI assistant handles the process from start to finish, Snapchat says....
- Managing frontier AI risk: From training to deploymenthttps://openai.com/index/managing-frontier-ai-risk: OpenAI explains how it manages frontier AI risk from training through deployment with model alignment, sandboxing, blocking safeguards, and monitoring...
- Introducing Gemini Decision 1: A Model That Picks Answers Instead of Writing Themhttps://www.kdnuggets.com/introducing-gemini-decision-1-a-model-that-picks-answers-instead-of-writing-them: Google's newest model has one job, and it never writes a word doing it. Decision 1 reads text, images, video, audio and PDFs in a single forward pass,...
- Google Introduces Gemini 4 Neon for Coding and Knowledge Workhttps://dailyai.com/2026/11/google-introduces-gemini-4-neon-for-coding-and-knowledge-work/: The new model is rolling out on 18 November in the Gemini API, Google Antigravity and the Gemini app. Google prices it at US$3 per million input token...
- Satya Nadella predicts AI agents will reshuffle the economics of the webhttps://techcrunch.com/2026/11/18/satya-nadella-predicts-ai-agents-will-reshuffle-the-economics-of-the-web/: Microsoft's CEO says AI agents could disrupt advertising-driven business models, forcing companies like Amazon to rethink how they compete for custome...
- Research acceleration: The jagged frontierhttps://openai.com/index/research-acceleration-jagged-frontier: OpenAI measures how coding agents are accelerating AI research, where their performance remains uneven, and what this signals about recursive self-imp...
- Amidst fears of AI-cyberattacks, Satya Nadella says lots of critical infrastructure is ‘unpatchable’https://techcrunch.com/2026/11/18/amidst-fears-of-ai-cyberattacks-satya-nadella-says-lots-of-critical-infrastructure-is-unpatchable/: The Microsoft CEO downplayed the recent Outlook outage while noting that AI was causing both opportunities and new challenges for securing much of the...
🎁 Exclusive Offers
- Nvidia’s startup stakes ballooned to $65B, but Huang’s exit puts future deals in doubthttps://techcrunch.com/2026/11/17/nvidias-startup-stakes-ballooned-to-65b-but-huangs-exit-puts-future-deals-in-doubt/
💼 New Opportunities
- Elvoro: AI Creative Engineer Social Mediahttps://weworkremotely.com/remote-jobs/elvoro-ai-creative-engineer-social-media
- Luminize: Bilingual Talent Acquisition Specialist / Interviewer USD Contracthttps://weworkremotely.com/remote-jobs/luminize-bilingual-talent-acquisition-specialist-interviewer-usd-contract
- Terac: AML and Compliance Experts: AI Training Taskshttps://weworkremotely.com/remote-jobs/terac-aml-and-compliance-experts-ai-training-tasks
- PRO-spectus: Access Manager - Healthcare Claims, PAs, Reimbursement & Appealshttps://weworkremotely.com/remote-jobs/pro-spectus-access-manager-healthcare-claims-pas-reimbursement-appeals
- California Virtual Academies: CAVA - Credential Compliance Specialisthttps://weworkremotely.com/remote-jobs/california-virtual-academies-cava-credential-compliance-specialist
- Montu: Nurse Practitionerhttps://weworkremotely.com/remote-jobs/montu-nurse-practitioner
- UMA Education: Director of Career Serviceshttps://weworkremotely.com/remote-jobs/uma-education-director-of-career-services
- A La C.A.R.T.E. Solutions: Senior Accountanthttps://weworkremotely.com/remote-jobs/a-la-c-a-r-t-e-solutions-senior-accountant
- Intelligent Technical Solutions: Purchasing Coordinator Remotehttps://weworkremotely.com/remote-jobs/intelligent-technical-solutions-purchasing-coordinator-remote
- Expensify: Customer Success Managerhttps://weworkremotely.com/remote-jobs/expensify-customer-success-manager
View Daily Recap