Inside OpenAI's Stargate Megafactory with Sam Altman (AI column)

I like to watch documentaries about technology, so my eyes fell on it. documentary Stargate by Sam Altman and Donald Trump. This film was made by the guys at Bloomberg and is an exclusive study of one of the most ambitious technology projects in history. The film shows a massive construction site in Abilene, Texas, where a cost-effective artificial intelligence infrastructure is being built. 500 billions of dollars. The film is hosted and produced by Emily Chang, the lead and executive producer of Bloomberg Originals, known for her interviews with top tech executives including CEOs of Apple, Tesla, Meta and Amazon.

During the film, the following characters were interviewed. 1. Sam Altman - General Director of OpenAI, one of the leading partners of the Stargate project. In the interview, he explained the logic of the investment in size. 500 The amount covers computing power needs for the coming years, based on their growth forecasts. 2. Masaioshi Son Chairman and CEO of SoftBank, who heads the financial side of the project. Son is known for his ambitious predictions about AGI, anticipating it within a decade. 3. Chase Lochmiller – Co-Founder and CEO CrusoeThe company responsible for physical construction. He is a former Wall Street quantum trader who left finances to conquer Everest, where he had ideas for Crusoe. He has degrees in mathematics and physics from MIT and a master's degree in computer science with a specialization in AI from Stanford. Several smaller experts were also interviewed.

Key ideas of the film 1. Scale of the project The campus is planned. 1200 acres in Abilene, Texas, 400 000 Nvidia GPU chips in eight buildings, power consumption 1.2 gigawatt. Initial investment. 100 Billions of dollars with expansion plans 500 billion-dollar 2029 year. 2. Technological innovation Using a closed cycle cooling system without evaporating water, which solves sustainability problems Direct liquid cooling chips instead of traditional methods Transition from CPU to GPU-centered architecture for processing AI tasks (I am not sure if this is an innovation in 2025 year) 3. Energy challenges Modern AI centers are used in 130 kilowatt per rack compared to 2-4 kilowatt 20 years ago

• The request to ChatGPT uses in 10 More Power Than Google Search K 2035 During the year, data centers can consume more 8% of America's electricity 4. Competition and efficiency The film touches on the emergence of the Chinese company DeepSeek, which demonstrated the ability to create effective AI models at much lower costs. This has called into question the need for megaprojects of the scale of Stargate.

If you look at the results of the film,

1. The project is positioned as critical to maintaining U.S. technological leadership in the AI race against China. Trump presented it as a national security project that would ensure American supremacy in artificial intelligence.
2. The project is supposed to help economic development of the United States - parts about creating jobs in the film sound controversial (Which Giga Data Center is Automated to the Maximum)
3. Alman sees 2025 A year is a period when AI agents will do the work we already know how to do. 2026 The year is a time of significant scientific progress thanks to AI.
4. The film shows the risks and uncertainty very well. Financial stability of participants: OpenAI suffered losses in 5 billion dollars 2024 year Environmental Issues: The Contradiction Between Climate Goals and AI Energy Capacity Geopolitical risks: trade wars and dependence on international supply chains

In general, the film shows the scale of the project from the inside, and also considers this project realistically from different angles.

#AI #Engineering #Software #ML #Leadership #Project #Management