OpenAI has launched its largest model to date, GPT-6 Astra, which has reached a critical cybersecurity capability threshold as the company declares that AI has entered the AGI era.
Google has released its third Flash model in six weeks, using rapid iteration to compete in the lightweight AI market while updates to the Pro series remain on hold.
The dark side of the moon has connected two major programming tool protocols, giving developers new workflow options. You can connect directly to programming models like Kimi K3 without any extra conversion.
World Labs, co-founded by Fei-Fei Li, has unveiled Atlas, the world’s first multimodal world model capable of generating consistent 3D worlds from a single image, paving the way for robotics and visual effects.
Anthropic has launched its flagship Fable 5.1 model, delivering major performance gains over its predecessor while significantly cutting the cost of long-running tasks—potentially signaling the start of a new AI model price war ahead of its IPO.
Manus announced on September 1 that it has resumed independent operations, with its original founding team continuing to lead the company. This marks the completion of its separation from Meta.
China’s first fully AI-generated long-form drama, The Later Journey to the West, has premiered with a “produce, review and release as you go” model, exploring a new path for AI-powered television production.
Why Mac, and Not Nvidia?
As AI agents move from task execution toward autonomous transactions, payment is evolving from a settlement tool into infrastructure connecting identity, authorization, usage and value flows.
The AI copyright fight has finally reached the biggest names in AI: Sony and Warner, joined by a coalition of music publishers, are suing Anthropic for up to $150,000 per song — a battle for the ages over "where the training data really came from" has only just begun.
While publicly accusing Anthropic of "attempting to monopolize," Zuckerberg has simultaneously made Meta one of the AI company's largest customers.
Nvidia CEO Jensen Huang says his only regret is not investing more—and sooner—in artificial intelligence, as the company’s equity portfolio of more than $40 billion helps secure computing demand and strengthen its ecosystem position ahead of an anticipated wave of AI IPOs.