OpenAI made waves this week with several major business deals that could reshape how artificial intelligence companies build their technology. However, the company now faces the challenge of making CEO Sam Altman’s expensive plans work in reality.
The company behind ChatGPT rolled out a series of partnerships this week that involve enormous amounts of money and put OpenAI at the heart of the next generation of machine learning systems.
The announcements started Monday when Nvidia said it would put up to $100 billion toward helping OpenAI create data center space filled with millions of graphics processing units.
The next day, OpenAI shared details about a bigger agreement with Oracle and SoftBank, expanding what they call the “Stargate” project into a $400 billion promise spread across several stages and locations. On Thursday, OpenAI added a business partnership with Databricks, showing the company wants to reach more corporate customers.
The company, best known for its ChatGPT chat program and GPT language models, wants to become much more than that.
OpenAI hopes to join the ranks of major cloud computing companies, even though it spends billions of dollars and depends completely on money from investors to keep growing. The company’s building plans would need enough electricity to power more than 13 million American homes.
Altman has said for a long time that creating the next level of AI technology will need much more computer infrastructure than what exists now.
OpenAI says it’s simply meeting customer demand, which keeps growing. The company believes these investments will eventually make money.
Inside sources say OpenAI expects to bring in $125 billion in revenue by 2029, based on the company’s own predictions.
Major infrastructure challenges ahead
But this strategy comes with big risks when it comes to actually building everything.
Creating 17 gigawatts of power capacity would need about 17 nuclear power plants, and each one takes at least ten years to build. OpenAI representatives say they’re talking with hundreds of infrastructure companies across North America, but nothing is set in stone yet.
The American power grid already has problems, gas turbines are sold out until 2028, nuclear power takes a long time to set up, and renewable energy projects face political obstacles.
“I am extremely bullish about nuclear, advanced fission, fusion,” Altman said. “We should build more … a lot more of the current generation of fission plants, given the needs for dense, dense energy.”
This week made clear just how big Altman’s plans really are, as the OpenAI CEO started putting specific numbers on his ideas, some of them huge.
Industry experts back bold strategy
“Unlike previous technological revolutions or previous versions of the internet, there’s so much infrastructure that’s required, and this is a small sample of it,” Altman said Tuesday at OpenAI’s first Stargate location in Abilene, Texas.
This approach – direct, ambitious, and ignoring traditional thinking – has marked how Altman leads during this new period.
Deedy Das, partner at Menlo Ventures, said OpenAI’s infrastructure partnerships with Oracle might look extreme to some people, but he sees it differently.
“I don’t see this as crazy. I see it as existential for the race to superintelligence,” he said.
Das explained that data and computing power are the two most important things for making AI bigger, and he praised Altman for understanding early how much infrastructure would be needed.
“One of his gifts is reading the exponential and planning for it,” he added.
Past breakthroughs in AI haven’t come from better computer programs, he said, but from having access to massive computing power. That’s why companies like OpenAI, Google and Anthropic are all trying to build bigger systems.
Your crypto news deserves attention - KEY Difference Wire puts you on 250+ top sites
