AI models Openai and Venado supercomputer will defend US National Security

The Los Alamos National Laboratory of the United States (LANL) entered into an agreement with Openai on the use of models of a series of ONI-Supercompeuter Venado. The system equipped with NVIDIA GH200 supercurrents will be used for research in the field of national security.

New models can solve more and more complex scientific problems, including applied. The laboratory emphasized that threats to the country are becoming more complex and relevant, so new approaches to ensure its security. The Openai models, in particular O1, will allow you to more effectively cope with the tasks and help solve some pressing problems of the country.

Venado will be included in the protected secret network, where researchers from Los Alamos, Liverrand National Laboratory (LLNL) or Sandia National Laboratory (SNL) will be able to use its resources. The cooperation of these structures of the US Department of Energy (DOE) with IT companies will ensure new opportunities for scientific research. It is reported that in almost all projects of the AI ​​laboratory is already used in one or another quality.

Image Source: Markus Winkler/Unsplash.com

Earlier, Lanl collaborated with Openai over the improving the safety of AI and the assessment of the “ability to reason” the O1 model. Now Openai models are regularly used by national laboratories in energy projects, the development of new materials and the creation of quantum algorithms. The use of models in the secret network will only expand their capabilities and will make it possible to make a deeper contribution to the implementation of different tasks.

The new AI models, according to Lanl, will allow the United States to reconsider the foundations of scientific knowledge and achieve a new level of technological progress. With their use, you can find new methods of treatment and prevention of diseases, achieve leadership in the energy sector due to the disclosure of the full potential of natural resources and convert the country’s energy infrastructure, it is better to identify natural and anthropogenic threats, from biological to cybernetic or threats of the energy system – even before their real occurrence.

Image source: Lanl

It is also known about the cooperation of Lanl and Openai to assess the capabilities of AI in biological research. So, according to The Register, AI capabilities were evaluated to create biological weapons. However, they also found more peaceful use – an assessment of a decrease in the threat of nuclear war and nuclear threats in general.

This is not the only one of the recent examples of Openai cooperation with the American government. A couple of days ago, the company announced the premiere of the ChatGPT GOV bot, the corporate version of ChatGPT Enterprise. Preliminary tests have already taken place in several organizations for the processing of “non -public confidential data”.