
Startup Meng Xiaoyi Launches AI Collar That 'Translates' Pet Barks
The 27g PettiChat wearable uses Alibaba's Qwen model to translate pet vocalizations and movements into human language with 95% accuracy.
Latest updates from the world of AI and technology

The 27g PettiChat wearable uses Alibaba's Qwen model to translate pet vocalizations and movements into human language with 95% accuracy.

The US administration approved $9 billion for NVIDIA Blackwell chips and finalized a classified security deal to employ the Mythos model.

DeepSeek announced its V4-Pro model will permanently operate at just a quarter of the original price, set at $0.87 per million tokens.

The second AI supercomputer was assembled in just 64 days, adding 110,000 NVIDIA GB300 accelerators to the cluster.

Facility in Bastrop aims to ensure energy self-sufficiency for the brand's high-density AI data centers.

Preliminary discussions aim to enable the use of custom Maia 200 processors for Claude model inference.

Elon Musk's company's IPO document confirms a $1.25 billion monthly contract through 2029.

Antigravity 2.0 platform orchestrated 93 sub-agents to create an OS from scratch capable of running the classic game Doom.

At Google I/O 2026, Google unveils Gemini Omni for conversational creation and editing of realistic videos from any input.

During the opening keynote of Google I/O 2026, Google unveils Gemini 3.5 Flash with sub-200ms latency, immediately available to developers.

Renowned scientist Andrej Karpathy joins Anthropic's pretraining team, partially pausing Eureka Labs to return to cutting-edge AI research.

Leak reveals Google's new Gemini 3.5 Flash model with sub-200ms latency and prices of $1.50 for input and $9.00 for output per million tokens.