National University of Singapore Releases Open Source Multimodal Language Model NExT-GPT to Advance Multimedia AI Applications

Data to be translated: The National University of Singapore has released the NExT-GPT multimodal language model, which supports processing of text, images, videos, and audio, thereby facilitating the development of multimedia artificial intelligence applications. The model employs a three-tier architecture and undergoes intermediate layer training through MosIT technology, offering open-source contributions that create opportunities for researchers and developers to integrate multimodal inputs. The unique feature of NExT-GPT lies in its ability to generate modal signaling tags, bringing potential applications in content generation and multimedia analysis.

OpenAI Makes Major Announcement: Former U.S. Secretary of the Army Dan Driscoll Appointed Senior Advisor

On November 3, OpenAI announced the establishment of a new position, Senior Advisor for National Security Modernization, and appointed former U.S. Secretary of the Army Dan Driscoll to the role. He will provide strategic advice to OpenAI leadership and collaborate with external partners to explore how advanced AI technologies can strengthen national security and cybersecurity defenses. This move shows that OpenAI is continuing to ramp up its efforts in the national security space.

Saudi Aramco shells out 17.5 billion yuan to pull a national-level AI company into its oil empire

Saudi Aramco will acquire a 49% stake in PIF's wholly owned AI startup HUMAIN, while PIF retains 51%. The deal will be completed by transferring key AI assets and injecting about 9.9 billion Saudi riyals (approximately 17.55 billion yuan), rather than simply throwing money at it. This means Saudi Arabia's national-level AI flagship has been brought into the oil giant's balance sheet, and the world's largest petrochemical company is formally increasing its AI bets.

Trump Strongly Pushes AI to Be Renamed as Super Intelligence, Tech Giants Line Up to Comply

Trump pushed to change the term "artificial intelligence" (AI) that has been used for decades to "super intelligence" (SI). He accused "artificial" of implying falsehood in September, signed an executive order requiring federal agencies to use SI in official communications, and threatened to label those who continue to use AI as enemies. Pressure reached Silicon Valley, with Musk leading to rename SpaceXAI as SpaceXSI, followed closely by Salesforce's Benioff.

U.S. Poll Shows 64% of People Say AI Development Is Too Fast, Nearly 80% Call for Government to Ensure Safety and Control

A recent poll shows that concerns about the rapid development of AI have crossed party lines among U.S. citizens. Most people believe the government should prioritize ensuring AI remains under human control and strengthen protection for workers. 64% of respondents said AI development is too fast, while only a small number felt the pace was appropriate or too slow. Concerns about the risk of AI becoming uncontrollable by humans are driving demands for regulation, making it a central issue in public policy. Several leading tech executives have also called for stronger governance.

OpenAI CEO's Speech at the United Nations: AI May Be Like the Renaissance or Trigger a Revolution-Level Turmoil

OpenAI CEO Altman told the UN Security Council that rapidly advancing AI brings immediate benefits and risks, like the Renaissance or disruptive like the Industrial Revolution. He urged global cooperation on standards for capability and risk assessment, safety, and human oversight of frontier AI systems.....