10 results on this page
SpaceX approached AI coding startup Cognition AI about a potential acquisition, according to people familiar with the matter. The deal would have marked SpaceX’s second major acquisition in recent months as it seeks to strengthen its position in AI.
Uber has begun offering autonomous rides in Zagreb, Croatia, marking its first European AV service. The rides are operated with Pony.ai and Verne, but vehicles still include a safety driver.
Anthropic announced that Claude would add invisible watermarks to AI-generated content to comply with new EU rules. Within hours, coders online began sharing purported methods to override or bypass them.
Calendly is expanding beyond scheduling by entering the crowded meeting note-taking market. It is also releasing Callie, a meeting scheduling assistant.
Anthropic’s effort to label AI-written text has prompted developers to build tools that remove those labels. Business Insider reports that one remover has gone viral on GitHub, raising concerns about the durability of AI-text provenance signals.
Blockbuster debuts by two technology companies show Beijing encouraging its tech champions to raise money from domestic investors. The strategy aims to finance China’s artificial intelligence ambitions while reducing dependence on Wall Street.
Kingsoft Cloud reported second-quarter FY2026 revenue of 3.072 billion yuan, up 30.8% year over year and above market expectations. AI-related customers drove growth, while the company recorded its first GAAP operating profit and continued investing heavily in AI computing capacity.
The community project J-Space Cognition Suite claims to improve DeepSeek V4-Pro-0813 through an inference-time Agent Harness without changing model weights. Its reported gains on several benchmarks have not been independently reproduced, and the project is unrelated to Anthropic's internal J-space interpretability research.
Pony AI CEO James Peng says the company plans to expand its robotaxi fleet into more Chinese cities. He expressed optimism that the broader rollout will help meet growing demand.
A community benchmark reports up to 218.3 tokens per second for Qwen3.8-27B on two RTX 3090 GPUs using vLLM, INT4 quantization, and DFlash2 speculative decoding. The setup achieved a 131K context ceiling, 168–178 ms time to first token, and peak VRAM usage of 22.3 GB per card.