Nathan Lambert Says DC Hostility Toward Chinese Open Models Has Reached a Record
Understand the event
Nathan Lambert said anti-Chinese sentiment toward open models in Washington, D.C., had reached record levels. His post did not identify a specific model, action, or evidence for the claim.
Generated by AI from reports · Updated 3 days ago
Read next
More changes in today’s briefingGoogle DeepMind Introduces Gemini 4 Hydrogen for Fast Agentic WorkflowsReport timeline
Follow the reports to explore different sides of the event.
- Microsoft Research Focus #98 Covers AI Translation, a Math Breakthrough, and Coding Agents
Microsoft Research host Anson Ho presents research updates on AI translation, the resolution of the 31-year-old KLS conjecture, and more efficient AI coding agents. A study of 13.5 million coding sessions found fewer than 8% required context compaction, while those sessions accounted for 44% of AI processing costs. Researchers say their APEC method reduced costs by up to 25.4% and execution time by 9.6% in evaluations.
- Jensen Huang’s Departure Raises Questions About Nvidia’s Open-Source AI Support
Jensen Huang’s resignation as Nvidia CEO has raised questions about whether the company will continue his support for open-source AI. Nvidia has named CFO Colette Kress interim CEO, and her approach to the Hugging Face deal and open model ecosystem is unclear. The article says open-source support also serves Nvidia’s business interests as customers such as Google and OpenAI develop their own chips.
- The Jensen Huang Scandal Is Hardly Over
The Verge argues that allegations about Nvidia chip diversion may pose a deeper threat to Jensen Huang and the company than the reported affairs. It points to questions around Megaspeed, Super Micro and New Matrix, while noting that Huang has said Nvidia has seen “no evidence of any AI chip diversion.” The article says further scrutiny could follow, but whether Huang knew of or enabled diversion remains uncertain.
- TechCrunch · AIPicksJensen Huang Resigns as Nvidia CEO Amid Scandal and Investigations
Nvidia confirmed in an SEC filing that its board accepted Jensen Huang’s resignation, and CFO Colette Kress will serve as interim CEO while continuing her CFO duties. The timeline also recounts reports of Huang’s relationships with Megaspeed executive Alice Huang and Nvidia executive Kimberly Powell, as well as investigations into Megaspeed over suspected chip diversion to China. Nvidia, Huang, and Powell have not confirmed the report about Powell.
- Nine South Korean Startups Compete for a Spot in Startup Battlefield 200
Nine early-stage South Korean startups pitched at DETO 2026 in Seoul this week for a chance to enter TechCrunch’s Startup Battlefield 200 next year. Voinosis, which develops voice analysis AI to detect changes linked to brain health, won the competition; the other startups work in areas including healthcare, fintech, energy, semiconductors, and AI.
- Google Launches Gemini Decision 1 for Single-Pass Classification
Google introduced Gemini Decision 1, a multimodal decision model that scores choices in a single pass. It supports up to 1,000 questions per request, returns answers in 150–250 ms end to end, and costs $0.10 per 1 million input tokens, with no charge for output. The project began as an internal Google DeepMind prototype, and a small team shipped the model and API about six weeks later.
- How Engineers Are Building with Jev and System One Models
At the AI Engineer Code Summit, TypeSafe AI co-founder Diogo Almeida argued that fast, structured-choice models such as Jev should be integrated into code to make software more reliable. Glean described using Jev alongside LLMs and classifiers for decisions that can adapt without repeated fine-tuning. UiPath tested ranking document chunks with thousands of Jev calls in under a second, while Workday showed a prototype that predicts user actions as they type.
- LeiphonePicksBaidu Releases ERNIE 5.5 Flash Preview With a Focus on Speed and Cost
Baidu released ERNIE 5.5 Flash Preview, its third major move in a week, with the model available through the Qianfan platform’s web interface and API. The article describes coding, 3D modeling, knowledge work, and content creation capabilities, alongside benchmark results and hands-on game development tests. Baidu says the weights will be released under Apache 2.0 by November 25, with support for SGLang and vLLM.
- Douyin Mall’s Double 11 Festival Showcases Brand Growth Strategies
During Douyin Mall’s 2026 Double 11 shopping festival, brands combined platform campaigns, consumer coupons, creator content and live-streaming to drive sales. The article highlights Charlotte Tilbury’s limited-edition compact, which accounted for more than 60% of its GMV, and Midea’s ranking first in the range hood category during the opening promotion, with brand GMV up more than twofold compared with the previous period.
- Baidu Releases ERNIE 5.5 Flash Preview, Highlighting Coding Capabilities
Baidu released its ERNIE 5.5 Flash Preview model on November 11. The model has 280 billion total parameters, activates 14 billion parameters per token, and supports context up to 1 million tokens. Baidu said it is available through the Baidu AI Cloud Qianfan platform, with API access for developers; model weights are scheduled for release to the global open-source community on November 25.
- JD.com Separates JD Technology Into Cloud Technology and Fintech Divisions
JD Technology has split into two divisions, Cloud Technology and Fintech, both on the same organizational level as JD Retail, according to market reports. The restructuring was announced internally last week. JD Technology’s JoyAI and AI algorithm departments have also moved into Cloud Technology and are listed as a C1 department. JD.com had not responded to the reports by publication time.
- The Jensen Huang Scandal: A Timeline of Reported Developments
TechCrunch organizes reports about Nvidia CEO Jensen Huang and the alleged Chinese spy Alice Huang into a reverse-chronological timeline. It covers Huang’s public admission of an affair, investigations involving Megaspeed and chip exports, and reports about a second relationship with an Nvidia executive, a possible board move to oust Huang, and a potential congressional subpoena. Nvidia, Huang, and the executive have not confirmed the second relationship report.
- AWS Recaps October 2026 Updates for AI Builders
AWS outlines October updates across Amazon Bedrock, AgentCore, and Strands, including new model options, agent reliability features, and cost reporting. Highlights include model fallback in AgentCore Gateway, structured outputs for Claude models on Bedrock, and Strands Box entering developer preview. The recap also describes new pricing for AgentCore memory and an estimated 98 percent drop in cost per ticket after an Amazon team adopted Strands Decider.
- Helsing Unveils Clarity, an AI Platform for Military Target Recommendations
Helsing unveiled Clarity, an agent-based platform that combines military sensor feeds, open-source intelligence, reports, and existing systems to prepare target recommendations for human review. The company says the software runs on customers’ infrastructure, and that a person must authorize each handoff to a weapons system. The article compares these safeguards with reported uses and failures of AI-assisted military analysis in Iran, Gaza, and a near-miss involving a Chinese ship.
- England’s Higher Education Regulator Warns Universities Over AI Use in Assessments
England’s Office for Students is urging universities to use more AI-free assessments, including traditional exams and in-person presentations, and warns it could sanction institutions that fail to develop reliable formats. A joint OfS and Advance HE survey of more than 300 respondents at 127 institutions found students used AI heavily for essays and dissertations, often without detection. The Guardian reported on the survey.
- Latent SpacePicksMercor Raises $750 Million in Series D at a $20 Billion Valuation
Mercor raised $750 million in a Series D at a $20 billion valuation after reaching $3 billion in annualized revenue, according to the AINews roundup. The article also reports that the company has 5 million experts and quotes a statement that it spends three times as much on inference as on employee salaries.
- Zhipu, MiniMax, Qwen Office, Alibaba Cloud and Qoder Offer 50% Off on Tmall
Zhipu, MiniMax, Qwen Office, Alibaba Cloud and Qoder are offering discounts of up to 50% on Tmall on November 11. Consumers can search for “Token” in the Taobao app to reach Tmall’s AI Space Station recharge center. The article says Token became a standalone category on Tmall for the first time during this year’s Singles’ Day shopping event.
- BaoyuPicksA Post Compares Google Docs’ Cloud Shift With Claude Artifacts for Agents
Baoyu argues that Google Docs changed software from a single-computer model to a cloud model, while Claude Artifacts points to software made for agents, with people reviewing and refining the results. He concludes that changing who uses software also changes its form.
- IGF Jurors Reject AI Use as Big Walk Developer Details Its Own
Some Independent Games Festival jurors said they would not vote for games using generative AI in any capacity, while Big Walk developer House House disclosed limited use of LLM-generated code. House House estimates that well under 0.1% of the game’s code was connected to that use, says it disclosed the use in its IGF submission, and plans to develop a more comprehensive AI policy.
- Google DeepMind Reveals Gemini Robotics 2.5 and ASIMOV-Jailbreak Benchmark
Google DeepMind has revealed Gemini Robotics 2.5, a new version of its vision-language-action model for generating robot actions. The report says it was evaluated on more than 200 activities, including serving food with tongs and using a pipette, and describes an Apptronik Apollo 2 robot tying a bow. DeepMind is also preparing an open source safety benchmark for robots operating alongside people.
- Latent SpacePicksAINews: Claude Fable 5.2 Leads Benchmarks as Anthropic’s IPO Roadshow Begins
Latent Space’s AINews issue reviews Anthropic’s release of Claude Fable 5.2 and Mythos 5.2 alongside the start of Anthropic’s IPO roadshow. It reports Fable 5.2’s 16.4% score on SemiAnalysis GDPx, while the cost per task rose to $88.15 from $67.16 despite lower per-token prices. The issue attributes the benchmark result in part to more extensive test generation and subagent checks, and surveys other evaluations, access limits, and reactions.
- What Could China Gain From Jensen Huang’s Nvidia Role?
The Verge examines allegations that Alice Huang, a Megaspeed executive and Nvidia customer, acted as a Chinese intelligence asset and had a personal relationship with Nvidia CEO Jensen Huang. Citing reporting by The New York Times and others, it details concerns about access to information on US export controls and Huang’s lobbying to loosen chip sales restrictions. Nvidia says it has no reason to believe national security or company secrets were compromised, and its board has opened an independent investigation.
- The Vergecast Covers Jensen Huang’s Resignation and a HomeView Review
The Vergecast covers Jensen Huang’s resignation as Nvidia CEO and includes a HomeView review. The source says Nvidia’s board installed CFO Colette Kress as interim CEO after Huang resigned.
- Anti chinese open model sentiment in DC about it hit record levels.
Anti chinese open model sentiment in DC about it hit record levels.
- Hong Kong FinTech Week 2026 Ends, Drawing More Than 48,000 Attendees
Hong Kong FinTech Week x StartmeupHK 2026 concluded on November 6 after a week of discussions, business networking, startup showcases and ecosystem events. It drew more than 48,000 attendees from over 120 economies, with more than 1,100 speakers, over 800 exhibitors and more than 30 delegations. The event featured the Global Fast Track 2026 final, won by Canadian company PragmaClin.
Keep up with AI, one email a day
Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue
Event heat trend
- Current within comparable coverage
- 3
- Peak within comparable coverage
- 1011/10 00:00 (UTC+8)
- 24-hour change
- -50%
Trends compare the same continuously observed participants, which may cover fewer sources than current heat. Point or click to inspect hourly heat, or use left and right arrow keys.