Trump's "Star Gates": The president announced a $500 billion investment in advancing artificial intelligence in the U.S.

The President of the United States, Donald Trump, has announced the Stargate initiative – a collaborative project involving OpenAI, SoftBank, and Oracle. This initiative aims to invest a staggering amount of up to $500 billion in the development of artificial intelligence (AI) infrastructure in the United States.

This was stated in Trump's official speech at the White House. According to him, Stargate will start with an initial investment of $100 billion, with plans to increase this amount to $500 billion over the next four years. The project includes the construction of large data centers to support the initial AI infrastructure.

Trump made the announcement alongside SoftBank's founder and CEO Masayoshi Son, OpenAI's head Sam Altman, and Oracle co-founder Larry Ellison. It is noted that Masayoshi will serve as the director of the joint venture. Microsoft, MGX, Arm, and NVIDIA have also partnered in this project.

"This monumental initiative is a bold declaration of confidence in America's potential under the new president. Stargate will create both physical and virtual infrastructure to support the next generation of AI achievements, including the construction of colossal data centers," he stated, adding that the project will generate 100,000 jobs.

Meanwhile, OpenAI CEO Sam Altman has already stated that this will be a pivotal project of the era. The first data center will be built in Abilene, Texas, followed by additional centers.

"This will be the most significant project of this era… This project will not only support the reindustrialization of the United States but also provide a strategic opportunity to safeguard the national security of America and its allies. We are all looking forward to the continued creation and development of AI, particularly AGI, for the benefit of all humanity. We believe that this new step is critically important on this path and will enable creative individuals to understand how to use AI for human advancement,” the statement reads.

At the same time, media reports indicate that Trump has rescinded all of his predecessor Biden's directives aimed at restricting AI development. The project itself will require enormous capacities.

"The rapid development of artificial intelligence systems over the past two years has placed significant strain on American infrastructure, with data centers becoming particularly bottlenecked. The latest models, such as OpenAI's ChatGPT, Google's Gemini, and Anthropic's Claude chatbots, require vast amounts of data and computational power for training and operation,” notes the Financial Times.

Meanwhile, a group of enthusiasts tested a machine gun system controlled by ChatGPT. Armed with a simulated automatic rifle using blank cartridges, the system unconditionally follows the commands of the human operator and fires according to the given instructions.

Additionally, GPT-o1 from Apollo Research is noted to be the best among AI models at lying, capable of manipulation, and even able to disable human oversight and offload itself to an external server for its own preservation. According to a report, "when the objectives of o1 differed from user goals, the AI model manipulated data to promote its own agenda in 19% of cases." GPT-o1 did not just lie – in 5% of cases, it attempted to deactivate the oversight mechanisms put in place to ensure alignment with user goals.