
Dario Amodei, head of Anthropic, spoke at a meeting of the UN Security Council with a warning about the potential threat of artificial intelligence to humanity, noting its rapid development and risks of loss of control, including possible use to create biological weapons, and stated the company’s readiness to slow down the development of AI in the interests of global security.
AI-generated summary
Anthropic is one of the world's leading artificial intelligence companies, known for developing Claude models. Dario Amodei is its co-founder and CEO.
The head of Anthropic, one of the leading AI companies in the world, Dario Amodei, once again frightened humanity with the possible consequences of the rapid development of artificial intelligence technologies. This time he did it at a meeting of the UN Security Council. “I even believe that AI can pose a threat to humanity as a whole,” TASS quotes a top manager as saying.
“AI is developing very quickly. Four years ago, he could barely write a line of code or solve a high school math problem. He now writes most of Anthropic's code,” Amodei continued, noting that while on this trajectory, the technology will, in a year or two or less, reach what he calls “the land of geniuses in the data center.”
According to the head of Anthropic, advanced AI models can bring many benefits to humanity, but are also associated with such significant risks as the possibility of their inappropriate use to create, for example, biological weapons, as well as the danger of generally losing control over its development.
Because of this, Amodei said his company is committed to slowing down AI development as much as necessary to ensure safety, stressing that such efforts must be "industry-wide and global in scope."
AI outlook — possibilities, not facts
Anthropic will slow down the pace of development of its AI models for security
Likely · Within months

Telegram blocked 65,721 groups and channels on Wednesday, a new low since August 16. During the week, 232,006 groups and channels were blocked, including 587 related to terrorism. Since the beginning of the year, the total number of blockings has reached 26.4 million, of which 178,663 are of terrorist communities.

Users of the 2024 Samsung Bespoke AI 4-Door smart refrigerators reported problems after installing the update: the devices turned off, stopped cooling, or froze during the update process. Samsung acknowledged the deployment error and stopped distribution of the update, without specifying the number of affected devices or compensation.

Google has added to the beta version of Android 17 QPR2 Beta 5 a function to restore access to a smartphone through a Google account without resetting data, reports Android Authority. For the function to work, you must first link your device to your account.

The head of the car service, Andrei Kuzin, said that most popular tips for saving fuel have minimal effect or harm the car. On modern cars with electronic injection, driving in gear with the gas released is often more economical than neutral. A reliable way to reduce consumption is to maintain your car in good condition, check the tire pressure, change consumables in a timely manner and drive smoothly.

The United States and China are collaborating on identifying the top risks of artificial intelligence, including the uncontrolled use of AI and its use by non-state actors in cyber and biological operations, US Treasury Secretary Scott Bessent said in an interview with Fox News.

The head of OpenAI, Sam Altman, said at a meeting of the UN Security Council that artificial intelligence may develop faster than humanity can respond to it, especially due to the automation of the process of creating AI and the risks of self-improving systems.