At a time when tech giants are running to build the most powerful AI models on the planet, Microsoft is choosing another track -one that is a little slower but potentially smarter.
Mustafa Suleyman, Microsoft’s CEO of AI, says the company is not in a hurry to lead the limit of AI development; Instead, it is intentionally choosing to hang a few months behind leaders like Openai. It may sound like a setback, but Suleyman says it’s a strategic feature that can give Microsoft an edge.
“It’s cheaper to give a specific answer when you’ve been waiting for the first three or six months for the limit to go first. We call it off-frontier,” Suleyman said in a recent interview with CNBC. “It’s actually our strategy that really play a very tight second considering the capital intensity of these models.”
A new chapter in Microsoft’s AI Playbook
The comments come as Microsoft celebrates its 50 -year anniversary, announcing several new features to its AI assistant Copilot. These include Copilot Vision, Deep Research and a new “Memory” feature that remembers details of users to customize answers – a capacity previously introduced by Openais Chatgpt.
Microsoft’s AI systems, including Copilot, are heavily powered by Openai’s models. The technology giant has invested nearly $ 14 billion in San Francisco starts, making it a key partner in Microsoft’s AI trip. But Suleyman also made it clear that Microsoft is gradually working on self -sufficiency.
“In the long term, it is absolutely mission -critical that we are able to make AI self -sufficient at Microsoft,” he said. “Until at least 2030 we are deeply cooperation with Openai, who has had a hugely successful relationship with us.”
The long -term view reflects Suleyman’s wider strategy. While the company is still dependent on Openai for large models such as GPT-4, it also develops its own smaller open source models during the Phi project. These models are designed to run on personal computers rather than massive GPU clusters, making them more accessible and cost-effective to implement.
Download this IT leader guide to generative AI from TechPublic Premium
Cracks in Microsoft-Openai Partnership?
The close ties between Microsoft and Openai have begun to show some strain. Last year, Microsoft quietly added Openai to his list of competitors. And in January, Openai surprised many by joining forth with Oracle on his Stargate project of $ 500 billion -a step that ended Microsoft’s exclusive status of Openai’s cloud provider.
Further escalating tensions, Openai’s latest GPT models have allegedly raised concerns within Microsoft about performance and cost. Industrial insiders believe that Microsoft has already begun to develop its own border models, and some observers, including Salesforce CEO Marc Benioff, believes Microsoft is already moving to stand on his own two feet.
The bigger picture
Microsoft may not chase the AI ​​krone, but it still plays a long game. With a powerful AI team, deep investments in infrastructure and a broad customer base, it doesn’t have to lead the race to win it.
By living right behind the bleeding edge, the company hopes to avoid expensive mistakes and offer more stable, reliable solutions to problems in the real world.
“We have an incredibly strong AI team, huge amounts of calculation, and it is very important to us that you know, maybe not develop the absolute limit, the best model in the world first,” he said. “It is very, very expensive to do and unnecessarily to cause this duplication.”