On Sunday, OpenAI CEO Sam Altman said that some upcoming AI features will be restricted to paying users due to high compute costs, though he pledged the company aims to make its technology more affordable over time.
Altman Outlines Paid Features On X
Altman took to X, formerly Twitter, and said that OpenAI will soon launch "new compute-intensive offerings."
He cautioned that certain features will initially be limited to Pro subscribers, while some entirely new products will carry additional fees.
"Our intention remains to drive the cost of intelligence down as aggressively as we can and make our services widely available," Altman wrote.
"But we also want to learn what's possible when we throw a lot of compute, at today's model costs, at interesting new ideas."
AI Bubble Warnings From OpenAI's Chairman
Altman's comments came days after OpenAI Chairman Bret Taylor compared the current artificial intelligence boom to the dot-com bubble of the late 1990s. He was agreeing with Altman's earlier comments.
Taylor said valuations are overheated but added that, as with Amazon and Google after the internet crash, durable companies will emerge.
OpenAI's Broader Push Into Chips
The discussion around costs also comes as OpenAI reportedly prepares to reduce its dependence on Nvidia Corp. (NASDAQ:NVDA) by moving into custom hardware.
The company is reportedly partnering with Broadcom Inc. (NASDAQ:AVGO) to begin mass production of proprietary AI chips next year, with a $10 billion supply agreement already in place.
Broadcom CEO Hock Tan confirmed on an earnings call that the deal had created "immediate and fairly substantial demand" for its semiconductors.
Read Next:
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.
Photo courtesy: jamesonwu1972/Shutterstock
