All Stories — Page 30

Liang Wenfeng's latest model release helped lift the round's reported target from 50 billion yuan, with an IPO described as possible in early 2027.

The a16z partner points to an experience problem; OpenClaw shows how setup, model access and permissions shape an agent's first impression.

AWS launched the model on October 5th with managed APIs, prompt caching and cross-Region inference; access is limited to eligible customers.

The October 5th agreement expands a relationship dating to 2020, putting Applied Intuition's Vehicle OS and development tools alongside Nissan's engineering teams.

Hugo Blendl says Higgsfield's tools took about four hours to make a recent 1-minute-20-second episode; Higgsfield's creator program covered generation costs.

The permit lapsed in August; Texas regulators told KXAN Tesla submitted its renewal late and would owe $11,250, while its separate automated-vehicle authorization remained active.

The state-approved pilot starts with two physicians reviewing every prescription and can move toward retrospective checks only after regulatory sign-off.

Olivia Moore's seventh consumer-app report adds U.S. card-spending data, with spending concentrated among people buying coding, creative and productivity software.

The MIT-licensed app coordinates coding agents, local models and visual checks, starting with Three.js browser games on macOS and Linux.

CEO Misha Laskin is pitching the model to companies and governments that want control over where AI runs; Reflection says Beam's weights will arrive later in October.

Shareholders representing 52.02% of Better's voting power backed removing all five incumbent directors. Garg now controls the reconstituted board; an interim CEO candidate has been identified but not appointed.

The framework OpenAI published on September 16th lets staff flag incidents and request public disclosure, while company teams control investigations and publication.

Founded by repeat entrepreneur Yoeven Khemlani and computer-vision researcher Harsha Khurdula, the model is available under Apache 2.0 and is designed to run on one 80GB GPU.

Policy chief Morgan Dwyer told New York City Council that OpenAI will publish its job-impact analysis; the federal bill would let companies submit workforce data to agencies.

Era generates linked company data across systems including Salesforce, Slack and Jira, giving developers a repeatable test environment without using production records.

Reflection AI's first scorecard puts Beam near or ahead of some open models on selected coding tests and behind others; its compute comparison excludes several costs of serving a model.

The Foundation found mostly sandbox edits and unsuccessful proxy attempts, and says it found no evidence its systems or data were compromised.

Tibo Sottiaux put the change at 50 tokens per second, up from 30, for subscribers using OpenAI products and supported partner apps.

After roughly six months in beta, Chakra is generally available. HackerRank says it ran more than 500,000 interviews, with Snowflake, Snorkel and Capgemini among its testers.

Logan Graham described the decision at a New York City Council hearing; Anthropic instead routed the model to vetted defenders through Project Glasswing.

At a New York City Council hearing, an OpenAI speaker cited support for multiple bills but named neither the measures nor the company representative.

Creator Bin Liu is extending HyperFrames from an agent-written video framework into an editor where people can review and revise the same project.

Flywheel AI's Jash Mota and Mahimana Bhatt are using teleoperation to put existing machines to work while gathering data for future autonomy.

BBC sources said Claude remained in Palantir's Maven Smart System as recently as last week, despite an August phaseout target.