The number of initial jobless claims in the U.S. for the week ended September 26 was 197,000, slightly below expectations.

Insight Beating AI News Flash: Runway has released the robot model Praxis-1, applying the expertise from training AI video models to real-world robots. It first learns how objects move and how humans perform tasks from a large number of regular videos, then translates that knowledge into robot movements. Runway conducted an object placement test: after pre-training with internet videos, the robot’s average error was 16.1 cm; when using custom-collected teleoperation videos of robots, the error was 16.0 cm. The test covered 93 evaluation sets, with results nearly identical across the two training approaches. If this line of work can be scaled, the data gap plaguing robot training could expand by several orders of magnitude. Collecting data via human-operated robot teleoperation requires specialized equipment, space, and labor, while regular videos are already abundant online. Runway aims to repurpose these videos—originally not intended for robot control—into training materials for robots to master movements. Praxis-1 has been tested on multiple robot models from Noble Machines, Standard Bots, and Ultra, including robotic arms, dual-arm robots, and mobile robots. Runway also showcased that the same strategy can adapt from a studio setting to a kitchen to complete tasks. The model is currently only accepting early test applications; its weights are set to be released in the coming months, with parameter size and more comprehensive independent evaluations still undisclosed.

Per HTX market data, the token of public blockchain network Monad surged more than 17% today, now trading at $0.032671. On the news front, Monad’s official Twitter account released a video yesterday hinting at the launch of its privacy service on October 6.

California Governor Gavin Newsom has signed Executive Order N-10-26, requiring state agencies to continue using the terms "Artificial Intelligence" and "AI" regardless of any potential shift by the federal government to adopting "super intelligence". Earlier, on September 27, Newsom enacted a series of bills, including: banning California public officials from issuing meme coins, and barring trading platforms from listing meme coins featuring public officials’ images; cracking down on cryptocurrency fraud and money laundering, and establishing a victim compensation mechanism; and clarifying procedures for seizing crypto assets from transnational criminal networks. In his signing statement, Newsom specifically criticized former President Trump’s personal meme coin TRUMP, issued in 2025, noting that nearly 1 million buyers have suffered total losses exceeding $3 billion, while Trump himself profited around $636 million.

According to crypto lawyer Ariel Givner, a judge at the U.S. District Court for the Eastern District of New York has dismissed a class-action lawsuit targeting Magic Eden’s operating entity Euclid Labs and its co-founder Zhuoxun Yin. The plaintiffs alleged that the platform, after promoting that the ME token would have features like cross-chain functionality and buybacks, transformed into a gambling platform, leading to the token’s price plummeting by over 98%. However, the judge ruled that the wallet terms signed by users explicitly include valid class-action waiver clauses, which left the federal court without subject-matter jurisdiction under the Class Action Fairness Act (CAFA). The court also held that it lacked personal jurisdiction over the co-founder. This ruling applies only to the pleading stage; other defendants, including CEO Jack Lu, remain on the lawsuit’s defendant list.

Bitget’s PoolX has launched a new lockup campaign, allowing users to lock XAUT, SOL, or XRP to compete for a share of a 100,000 BGB airdrop. The lockup period runs from 14:00 UTC+8 on October 2 to 14:00 UTC+8 on October 9. The XAUT lockup pool offers a total airdrop of 45,000 BGB, with an individual lockup cap of 300 XAUT; the SOL pool has a total airdrop of 35,000 BGB and an individual lockup limit of 10,000 SOL; the XRP pool provides a total airdrop of 20,000 BGB, with an individual lockup cap of 400,000 XRP. Notably, this round’s XAUT lockup pool features a long-term holding bonus mechanism: the system will add an extra bonus to users’ lockup amounts based on their minimum XAUT holdings over the past 15 days. For more details, refer to Bitget’s official platform.

Beating AI News: Perplexity has open-sourced a new context embedding model, pplx-embed-v2-context-9b-preview, built primarily for Retrieval-Augmented Generation (RAG). Traditional embedding models typically split long documents into small chunks and process each individually, but this new model encodes all chunks together, allowing each chunk’s vector to reference the full document context. The model is now available on Hugging Face under the MIT license. For example, a sentence stating “It grew by 20%” is ambiguous on its own, as the subject of “it” is unclear; the new model can interpret this sentence by integrating context from other sections of the document. During training, it is specifically designed to retrieve both answers and supporting paragraphs, enabling downstream AI systems to receive not just answers, but also the evidence needed to verify them. In a blind test conducted on turbopuffer’s context-bench, the new model achieved an Answer@10 score of 45.5%, 14.4 percentage points higher than Voyage Context 4. Its Evidence Recall@10 reached 40.6%, a 5 percentage point lead. The test includes 38,894 long documents and 2,099 queries, with undisclosed questions to reduce the risk of test set contamination by training data. The model remains in preview. Perplexity warns that future weights, APIs, and generated embeddings may change, so vectors produced now may not be directly compatible with future versions.