At the beginning of this year, Salesforce CEO Marc Benioff predicted that Microsoft may emancipate itself from an overreliance on OpenAI for its AI smarts, predicting that the former is already developing its own frontier AI models.
This was shortly after OpenAI unveiled its $500 billion Stargate project designed to facilitate the construction of data centers across the United States for its AI efforts. Aside from gaining more independence from OpenAI, Microsoft reportedly raised concerns about OpenAI’s GPT-4 model being too expensive and not fast enough to meet consumer needs.
A separate report corroborated that Microsoft is building its own AI models. Following the announcement of OpenAI’s $500 billion Stargate project, Microsoft lost its exclusive cloud provider and largest investor status to SoftBank.
SoftBank led OpenAI’s latest funding round, raising $40 billion, which pushed the ChatGPT maker’s market cap to $300 billion.
While details about Microsoft developing its own AI models remain slim, Microsoft AI CEO, Mustafa Suleyman, disclosed more details about its advances in the field. Speaking to CNBC’s Steve Kovach, he indicated that building AI models that are three or six months behind comes with its fair share of advantages.
More specifically, the executive indicated that developing models in this fashion reduces the development cost, allowing the company to focus on specific use cases.
According to Microsoft’s AI CEO:
“It’s cheaper to give a specific answer once you’ve waited for the first three or six months for the frontier to go first. We call that off-frontier. That’s actually our strategy, is to really play a very tight second, given the capital-intensiveness of these models.”
During Microsoft’s 50th Anniversary and Copilot event, the company announced a myriad of new capabilities shipping to Copilot, including Copilot Vision, Deep Research, Pages, Copilot Avatar, and Memory. As such, it’s apparent that Microsoft will need to have more control over the AI capabilities it has integrated across its tech stack, and the best way to do that is by building its own AI models.
Suleyman continues:
“Look, it’s absolutely mission-critical that long-term, we are able to do AI self-sufficiently at Microsoft. At the same time, I think about these things over five and 10 year periods. You know, until 2030 at least, we are deeply partnered with OpenAI, who have [had an] enormously successful relationship with us. Microsoft is focused on building its own AI internally, but the company is not pushing itself to build the most cutting-edge models.”
“We have an incredibly strong AI team, huge amounts of compute, and it’s very important to us that, you know, maybe we don’t develop the absolute frontier, the best model in the world first. That’s very, very expensive to do and unnecessary to cause that duplication.”
To that end, it’ll be interesting to see how Microsoft mitigates the emerging issues with AI as it scales greater heights and demands more computing power, GPUs, and cooling water.
