AI Trends and Related News
Google launches Gemini 3.8 Live dual models, with its voice AI topping benchmarks and support for “thinking while speaking.”
On September 15, Google announced two new speech models, Gemini 3.8 Live and Gemini 3.
ChatGPT technology co-inventor unveils Jev: an AI model that doesn’t generate text and only makes decisions, claimed to be 200 times faster and 400 times cheaper.
Former OpenAI researcher Diogo Almeida, who claims to have co-invented ChatGPT, yesterday (September 15
Perplexity teams up with NVIDIA to bring Portable Computer to Windows, turning RTX PCs into local AI agent workstations.
Perplexity and NVIDIA today announced that they will expand Portable Computer to Window
An overseas user ran Qwen3.8-27B locally on an RTX 3090 and built a first-person zombie shooter in just 5 hours.
Now that local AI models have come this far, many people are surely curious: just how far have they gotten?...
Is AI really moving too fast? OpenAI, Anthropic, Google, and Musk, in a rare show of agreement, support slowing frontier AI development.
Anthropic CEO Dario Amodei published a long essay on September 12 titled "We Must Pace "
Anthropic picks Nasdaq for October listing, IPO valuation soars to 2 trillion USD
Business Insider exclusively reports that Anthropic, the developer of Claude, has selected Nasdaq as
M5 Max vs RTX 4090 vs 5090 local AI video generation benchmark: the results are surprising (LTX2.5 vs MiniMax H3)
Ever since MiniMax H3, the current most powerful open-source AI video generation model, was released, professional AI video generation tasks...
Anthropic’s September Threat Report: AI bot armies mass-produce fake news, seven Chinese AI labs collude to steal models via distillation, and Taiwan is also on the list.
Anthropic released its latest threat intelligence report on September 10, documenting the period from December 2025 to 2026
Even an old laptop with 12GB RAM can run Qwen3.8-27B! A developer used RAMDeck to link 4 devices together and share the memory load.
Many people want to run free open-source AI models on their own computers, but the biggest hurdle right now is...