Microsoft Confirms Development of Its Own AI Models, Eliminating Dependence on OpenAI

Despite its partnership with AI developer OpenAI and $13.75 billion in investment, Microsoft is building its own AI models, Microsoft AI CEO Mustafa Suleyman confirmed. He said the company is not aiming to build cutting-edge AI models, but pointed out the positive aspects of this approach.

Image source: Growtika/Unsplash.com

Microsoft’s AI chief said it was “absolutely critical” to the mission that the company be able to pursue AI on its own in the long term, though Microsoft will work closely with OpenAI until at least 2030. He stressed that the company is not making efforts to create breakthrough models.

«”We have an incredibly strong AI team, a huge amount of compute, and it’s really important for us not to spend resources trying to develop the world’s best model first. That’s very, very expensive, and there’s no need to duplicate that effort,” Suleiman said. He noted that building AI models three or six months behind the cutting edge has its advantages. “That’s really our strategy — to be very powerful in the second role, given the capital intensity of those models,” the head of Microsoft AI said.

Cracks have recently appeared in the previously close partnership between Microsoft and OpenAI, shortly after OpenAI announced its $500 billion Stargate project to build a data center in the United States, partnering with rival cloud provider Oracle. Not only did Microsoft lose its status as OpenAI’s exclusive cloud computing provider, it also ceased to be its largest investor after SoftBank led OpenAI’s latest funding round, raising $40 billion.