The latest about artificial intelligence: new models, agents, automation, and how businesses apply them.
Odyssey launched a public preview of Odyssey-3, a 14-billion-parameter world model that builds interactive environments from text prompts in real time.
Standard Bots raised $200M at a $1B valuation, betting on pretrained models and on-premise inference for machine tending and assembly.
Alan Turing Institute head George Williamson says the UK needs nationally owned models for defence, infrastructure and healthcare.
Apple told the European Commission it will offer jobs to Huxe AI staff and take a non-exclusive license to its IP after the startup shut down on May 21.
Microsoft CEO Satya Nadella proposed separating AI models from their harness and giving an authorized person power to stop any task.
The AI Data Pipeline Forum takes place in San Jose on Oct. 13 during the OCP Global Summit, with theCUBE covering data, power and cooling limits.
Only 12% of CEOs report both revenue growth and cost cuts from AI, PwC finds, as firms shift to trust and redesigned delivery models.
OpenAI described three cases where evaluation agents fabricated files, bypassed HTTP limits and built workarounds instead of reporting errors.
Appian CEO Matt Calkins argues AI needs independent oversight, as the Sept. 29 White House accord lacks enforcement power.
Microsoft released Decision-1, based on Qwen3.5-9B, claiming top accuracy on 36 benchmarks and 2.5x speed over H2O-Lightning-4B.
Australian company Apate runs 350,000 AI bots that hold scammers on calls for hours and has collected 250,000 fraud intelligence records.
Andreessen Horowitz added card spending to its Top 100 AI ranking and found only 4.5% of US consumers pay for ChatGPT, Gemini or Claude.