Once again Salesforce CEO Marc Benioff has a prediction about Microsoft and OpenAI’s partnership, however, this one may be more concrete than his more grandiose partisan rants, and it’s not looking good for the future between Copilot and ChatGPT.
According to a recent article from The Information, Microsoft is reportedly testing models from xAI, Meta, and DeepSeek in its Copilot, suggesting that the company might be planning to sever its ties with OpenA, or at the very least, lessening its development and financial exposure with the company. This move isn’t entirely surprising, given Microsoft’s previous complaints about OpenAI’s technology being too expensive and not fast enough to meet its enterprise customers’ needs.
Salesforce CEO Marc Bennioff seized on the reporting to issue the following quote, “Mustafa Suleyman and Sam Altman aren’t best friends.” While salacious in its delivery, Bennioff attributes his claims to a recent appearance of the two on a panel during Davos last year where they seemed to be uncomfortable around each other.
Beyond anecdotal observations, Benioff is correct in pointing out the shift in both rhetorical and functional strategies Microsoft has exhibited in its short partnership with OpenIA. Going from all-in with a $13 billion investment infusion in the relatively small AI company to now testing models from some of the biggest names in the space with its consumer-facing product, Microsoft has made some big moves in the past four years.
While Microsoft may have leveraged OpenAI’s LLMs to get its own version of an AI platform off the ground, it seems like the company is looking to expand and diversify its use of models to accommodate a varying degree of AI-related projects, tasks, and app usage.
Couple this with the news that Microsoft’s in-house models dubbed Microsoft Artificial Intelligence (MAI) are outperforming OpenAI efforts in some instances, which could be cause for its partner to worry about its long-term collaboration outlook up-to and beyond 2030.
Microsoft seems to be taking a page out of Apple’s playbook, aiming for a more flexible business model that allows them to incorporate and switch between the best models for specific tasks. This strategy could potentially give Microsoft the upper hand in the AI race, allowing them to cherry-pick the most efficient and cost-effective models for their needs.
Adding fuel to the fire, Microsoft has been quietly testing its own Small Language Models (SLMs) to be more targeted in their future AI efforts. These SLMs are designed to perform specific tasks using fewer resources than larger models, making them ideal for domain-specific applications4. It’s almost as if Microsoft is saying, “Why rely on someone else’s expensive and slow tech when we can build our own, more efficient models?”
Meanwhile, OpenAI became the face of the $500 billion Stargate project, which is a national effort to boost AI adoption with the inclusion of data center constructions, commercial and consumer integration as well as individual investments such as training and education in the sector.
Yes, Microsoft benefits from the Stargate project tangentially, it’s still something that OpenAI and others will spearhead to the benefit or detriment of their associated brands. OpenAI has struggled with timely and worthwhile updates to their ChatGPT models as of late, and Microsoft has been said to be weary of the company’s ability to not only deliver but sustain itself as revenue strategies as of late have proven insufficient in providing a positive cash flow future for the AI provider.
So, what does this mean for the future of AI? If Microsoft’s strategy pays off, we could see a shift in the AI landscape, with companies moving away from monolithic models like OpenAI’s GPT-4 and towards a more modular approach. This could lead to more competition, innovation, and ultimately, better AI solutions for everyone.
But let’s not get too carried away. After all, this is the tech world we’re talking about, where today’s groundbreaking innovation is tomorrow’s obsolete relic. Only time will tell if Microsoft’s gamble will pay off or if they’ll end up back at square one, scrambling to catch up with the next big thing.
