AI news, Nov 14, 2026: OpenAI Dissolves Robot Team; LangChain Hits $3B Valuation
OpenAI has dissolved its robotics team, shifting strategy to have general AI models directly drive robots. Simultaneously, LangChain reached a $3 billion valuation as its LangSmith platform expanded to 7,000 clients. Concerns over AI safety remain high, with an OpenAI director warning that social takeover risks could materialize within months. The industry is also seeing new vertical applications, including Anthropic's manufacturing plugin and Perplexity's Decider v1.3 model.
Lead written by AI from the day’s stories
Top stories
- Other
OpenAI Disbands Robotics Team, General Models Take Control
OpenAI has announced the disbanding of its robotics team. The move indicates the company is focusing resources on its general AI models, which are now capable of directly controlling robots. This strategic shift emphasizes model generality over a dedicated hardware team for embodied intelligence.
Hacker NewsIT之家2 outlets
- Funding
AI agent startup LangChain hits $3B valuation
AI startup LangChain has reached a $3 billion valuation. Its model-agnostic agent management tool, LangSmith, now claims 7,000 customers.
- Research
Hugging Face releases XYBench for AI evaluation
Hugging Face released XYBench, a benchmark designed to address underlying problems rather than just answering questions. It evaluates AI models to improve response quality. The release provides developers with a new standard for assessing model performance in complex scenarios.
- Other
OpenAI board member warns of imminent AI societal risks
An OpenAI board member warned that an AI societal takeover could be months away, tracing a path from reward hacking to potential human extinction. This statement highlights deep concerns among top researchers about current AI system safety.
- Products
Anthropic Launches Manufacturing Plugin
Anthropic released a Claude for Manufacturing plugin that integrates AI with factory software like Autodesk Fusion. It features 17 specialized skills to automate workflows such as drafting quotes and placing backup orders after human approval. The plugin does not yet control physical machines.
Models1
- Models
Perplexity Releases Decider v1.3 Decision Model
Perplexity released open-weight Decider v1.3 and Flash v1.3 models. The new version adds Locate functionality for span extraction, achieving 91.6% mean accuracy across 63 query-classification tasks.
Products1
- Products
OpenAI Launches ChatGPT Businesses with AI Agents and Ads
OpenAI officially launched the ChatGPT Businesses beta, allowing companies to claim brand profiles and build dedicated AI business agents. Businesses can upload real-time inventory data to improve recommendations and purchase sponsored placements for ads in chat interfaces. This move expands OpenAI's strategy into search marketing and e-commerce.
Research1
- Research
Study: Fluent AI Responses Lead Users to Ignore Hallucinations
A new study found that users are likely to trust large language model responses containing hallucinations if the language is fluent and polished. The research suggests that fluency alone is insufficient for evaluating AI accuracy, highlighting risks of being misled by confident but incorrect answers.
Policy1
- Policy
MIIT to Establish Industrial OS Standards Group to Promote AI Agents
China's MIIT announced preparations for an industrial operating system standards working group to support AI in industry. The initiative aims to promote industrial AI agents, targeting 20,000 software enterprises and over 100 smart manufacturing factories by 2028 to enhance industrial digitalization.
All news → · All rights belong to the original creators