OpenAI will develop AI-specific hardware, CEO Sam Altman says
ChatGPT operator eyes biggest disruption in tech devices since iPhone
PALO ALTO, California -- After taking the world by storm with ChatGPT, OpenAI is looking to develop an artificial intelligence-specific device -- a move that could bring the largest disruption to tech hardware since the 2007 launch of the iPhone.
"We hope to do it in partnership" with Jony Ive, a former chief design officer for Apple, OpenAI co-founder and CEO Sam Altman told Nikkei on Jan. 27. Altman is visiting Japan and plans to meet with Japanese Prime Minister Shigeru Ishiba on Monday.
OpenAI has not previously announced official plans to develop AI-specific hardware. Altman indicated that it would take several years to unveil a prototype.
"It feels to me like AI is a big enough shift in how we can interact with computers that there ought to be a new kind of hardware," he said.
Altman's journey as an entrepreneur began when he dropped out of Stanford University to start a mobile social media company. He later served as president of Y Combinator, a startup accelerator, and co-founded OpenAI in 2015. He became the full-time CEO in 2019.
The startup's ChatGPT AI chatbot launched in 2022 and has since gained over 300 million users worldwide. OpenAI aims to lead the AI market in both software and hardware with its AI-specific device.
Apple's iPhone revolutionized user interface design with the use of touchscreens. With AI, "I think voice should be a key," Altman said.
When asked whether OpenAI would develop its own semiconductors, Altman said, "We're certainly working on it."
Other U.S. tech companies like Apple, Google and Amazon are also working on custom-made chips to deliver faster and more efficient AI.
On the day after U.S. President Donald Trump's inauguration, OpenAI announced a $500 billion Stargate venture with SoftBank Group and Oracle to develop AI infrastructure on U.S. soil.
OpenAI will be responsible for running the company overseeing Stargate and will build and operate data centers.
"Stargate is a big and complex project that is up and down the stack," Altman said. "I think there's places to collaborate [with Japanese companies and investors] at every level, including chip."
Altman is calling for Japanese companies to participate in Stargate. OpenAI has recently put forward policy recommendations warning the U.S. might fall behind China competitively without drawing AI investment from global partners, such as Japan and Middle East countries.
Chinese startup DeepSeek has developed an AI model it claims rivals OpenAI's model at a fraction of the cost.
DeepSeek's offering is "clearly a good model," said Altman. "It's a reminder of the strong interest in reasoning, and there is going to be serious competition."
"This capability level isn't new," he added. "We've had models at this level for some time and we'll continue to make better models."
OpenAI is looking into whether DeepSeek improperly used its proprietary technology for training.
China has caught up to American AI companies "significantly," Altman said, while warning bad actors could apply advanced AI to military and similar applications.
"If authoritarian governments misuse powerful AI for power consolidation, that would be bad," he said.
The Stargate venture has become a signature project in Trump's presidency. Under former President Joe Biden, the government sought to regulate AI development by setting mandatory safety requirements for powerful AI models, but the Trump administration represents a departure from that trajectory.
The plan is to ease regulatory strictures and encourage investment in AI companies, so the U.S. solidifies global leadership in the field.
"I think it's important for the world to get AI built correctly and be deployed for maximum benefit to everyone," said Altman. "U.S. leadership is helpful to get to that place."
"I was very impressed with the way President Trump thinks about the need to build infrastructure and to move quickly and sensibly on technology policy," he added.
Some have voiced concerns that AI safety would be compromised if the Trump administration eases regulations. Altman mentioned the idea of establishing an international watchdog to monitor AI development processes for potential harm and prevent AI from becoming a threat to humanity.
"When we have these technologies that have truly global potential, mega impact, we have some sort of international framework or plan," said Altman.
The International Atomic Energy Agency is one example.
"I think the same thing makes sense here," said Altman. There is "a lot of talk in the world right now about all of the different approaches, that'll be for the heads of states to decide," he added.