According to CoinWorld, Wu has learned that a Grayscale article points out that the restricted access to Anthropic's updated model highlights the risks of centralized AI control. Grayscale believes that as AI becomes an increasingly important economic resource, a few companies will control the most advanced AI systems, and the influence of governments and AI labs on model access will continue to rise. Against this backdrop, decentralized AI networks represented by Bittensor and its TAO token are likely to gain more attention. Grayscale predicts that demand for decentralized AI like Bittensor will continue to rise as investors seek alternatives.

According to Bijie.com, Alibaba (09988.hk) has released the Qwen-Robot series of embodied intelligent models, comprising three models: Qwen-Robotmanip (VLA operation model), Qwen-Robotnav (VLN mobility model), and Qwen-Robotworld (world model). This is the first complete embodied intelligent model series in the Qwen family. The three models respectively equip robots with dexterous hands, navigational legs, and a thinking brain. They can be deployed individually or operate collaboratively, providing a reliable "universal foundation" for the real-world deployment of robots of different forms.

According to CryptoWorld, U.S. Federal Judge Rita Lin in San Francisco dismissed XAI's trade secret lawsuit against OpenAI, ruling that XAI cannot retaliate. The judge determined that XAI could not prove OpenAI instigated former XAI engineer Xuechen Li to leak the Grok chatbot's secrets, and that asking job seekers about their past work experience during interviews is standard industry practice. OpenAI stated that Xuechen Li never joined the company and that OpenAI did not obtain XAI's secrets. Furthermore, on May 18th, a federal jury ruled against Elon Musk in his lawsuit accusing OpenAI and Sam Altman of deviating from their non-profit mission and seeking $150 billion in damages.

According to Odaily Planet Daily, the Bank of Japan (BOJ) stated at its monetary policy meeting today that the BOJ Policy Board reviewed the development and operation of the Japanese government bond market and discussed its future policy on purchasing government bonds. In principle, long-term interest rates should be determined by the financial markets; therefore, it is appropriate for the central bank to purchase government bonds in a predictable manner, while maintaining sufficient flexibility to support the stability of the government bond market. Based on this approach, to improve the efficiency and stability of the government bond market, the BOJ decided by a 7-1 majority vote to take the following measures: From now until January-March 2027, the BOJ will, in principle, reduce its monthly government bond purchase program by approximately 200 billion yen per calendar quarter. From April 2027, its monthly government bond purchase program will remain at approximately 2 trillion yen. If long-term interest rates rise rapidly, the BOJ will take flexible measures, such as increasing the scale of government bond purchases, conducting fixed-rate government bond purchase operations (both of which are not subject to the monthly government bond purchase program), and conducting funding supply operations based on pooled guarantees. Furthermore, the Bank of Japan will not conduct a medium-term review of its direct purchase program of Japanese government bonds. However, the central bank is prepared to adjust the pace of purchases at its monetary policy meeting, based on the fundamental principles of the purchases and other factors such as the performance of the Japanese government bond market, if deemed necessary. (Jinshi)

In the past hour, total liquidations across the network reached $13.74 million, with $10.07 million from long positions, with $3.66 million from short positions. BTC liquidation $4.22 million, ETH liquidation $5.80 million.