AI Industry Development & Policy Dynamics

NEAR Protocol launched the Delta League, an open competition designed to enhance AI models by enabling developers to directly edit model weights. The initiative encourages developers to submit models for evaluation on a hidden test set, with verified submissions eligible for further development. NEA

According to NEAR Protocol's official announcement, the protocol recently launched Delta League, an open AI competition enabling developers to directly edit model weights and submit entries for evaluation on a hidden test set. The initiative aims to foster creativity and collaboration within the dev

According to CNBC, on November 5, Chinese AI lab Moonshot AI is in talks with at least one European data center to lease clusters with 8,000 or more Nvidia Blackwell GPUs, primarily for inferencing capacity but also exploring training compute. The move circumvents U.S. export restrictions on advance

According to TechCrunch, AxiomMath announced on Wednesday that it had formally verified the Stacks Project, one of the largest bodies of mathematics ever compiled. The company translated the project's approximately 7,654 pages into roughly 16 million lines of computer-checked code using Lean, a prog

Expedia CEO Ariane Gorin said Thursday the company has not yet seen artificial intelligence agents affecting customer booking behavior, though early signs show potential for industry shifts. In a CNBC interview, Gorin acknowledged AI agents like Meta's Muse and OpenAI's Dots are becoming useful for

Dropbox announced Thursday its acquisition of cloud technology company Tigris Data for $350 million, marking what appears to be its largest acquisition to date. The deal is expected to close in the fourth quarter of 2026 and aims to strengthen Dropbox's AI infrastructure capabilities, supporting the

According to TechCrunch, Box unveiled Box Content Retrieval at its developer conference on Thursday, integrating Nvidia's Nemotron embedding system to power agent-first content searches. The system will deliver queries four times faster, with 20% fewer errors and 90% fewer tokens used, according to