Hayden Field and Tom Warren, reporting for The Verge:

For years, Microsoft’s AI business leaned hard on its early and exclusive partnership with OpenAI. But the drama-filled marriage slowly devolved into a situationship, and the pair effectively separated in late April (though Microsoft is still OpenAI’s primary cloud partner — for now). This year’s Build had the vibe of a freshly single divorcée posting a thirst trap on Instagram. “It’s always fun to be at developer conferences in times of great change,” Microsoft CEO Satya Nadella said onstage Tuesday, adding that events like this are about “coming to grips with the new opportunity.”

AI chief Mustafa Suleyman, in an interview with The Verge, put it even more bluntly.

“The goal is to prove that we can become one of the top four labs in the world,” Suleyman said. “There’s three labs that matter, Google DeepMind, OpenAI, and Anthropic. We are not one of them at the moment, and that’s always been my intention. It’s why I came here. I want to build the very best frontier models in the world, fully multimodal, and in order to do that, we have to prove that we can do everything that we need to from the ground up, and we’re not just going to take from others.”

One of Microsoft’s first steps at Build was indeed to play catch-up on AI models. Suleyman unveiled MAI-Thinking-1, the company’s first reasoning model, along with six other new models focused on image, voice, transcription, and coding. Microsoft said the medium-size MAI-Thinking-1 model, which will likely be marketed to primarily enterprise clients, is “built from scratch for serious math, coding, and real-world enterprise deployment.” Microsoft is years behind both OpenAI and Anthropic here; OpenAI began releasing reasoning models in the fall of 2024. But Suleyman emphasized its performance on benchmarks like coding and its price point, saying it was cheaper than OpenAI equivalents on some tasks — a big deal in the age of the AI money squeeze, which has inspired a lot of complaints with customers.

I think the main takeaway from this story is that Microsoft desperately wants to be known for anything other than its consumer products, Windows and Microsoft 365 (née Microsoft Office). That’s why it invested in OpenAI through Azure usage credits. Satya Nadella, Microsoft’s chief executive, has concentrated his company’s efforts on cloud computing, and when ChatGPT launched, Microsoft had no real choice but to back OpenAI and become the inference provider. As for OpenAI’s end of the deal, Microsoft integrated GPT-4 into Bing and Windows under the Copilot suite of products. (Remember Sydney?) Microsoft had to get something out of giving OpenAI millions of dollars in compute.

But Microsoft never wanted to be here. It had wanted to abandon Windows and Microsoft 365 for years, just to be pulled back in during the artificial intelligence boom due to the OpenAI investment. It felt pressure to get something out of the investment and capitalize on the boom itself, so it haphazardly reinvested in product, much to the chagrin of its users. Windows itself is underdeveloped, even by Microsoft’s own admission. (The company is finally remedying this after years of complaints.) Its core product — the piece of software it is known for — has languished in the face of frankly useless AI additions, like AI images in Notepad. (What?)

This was never how Microsoft drew up its plans, and now that OpenAI is building its own Stargate data centers with Oracle — if it can afford them — and Microsoft is released from its initial commitments, the company is easing back into being a model and inference provider. The whole Microsoft Build presentation on Tuesday signaled a shift away from consumer products and a concerted effort to appeal to developers. Microsoft has missed out on the OpenClaw phenomenon. It has kneecapped its GitHub Copilot plans due to high demand. Every major AI lab builds for macOS first and Windows when they remember it exists. And it’s not like normal consumers have given Microsoft a vote of confidence for its AI features. Clearly there’s room for a different strategy here.

I think new models are a great place to start. Microsoft is an incompetent product organization. Nobody there is driven by great products, unlike Apple or maybe even Google. Leave the Windows team to clean up the mess and start anew — build inexpensive, competent models and sell them to developers. The results will trickle down to users eventually in the form of better software, so long as agentic development is the future of software engineering. At a time when OpenAI seems close to collapse after its inability to market to enterprise users, Microsoft seems uniquely positioned to leverage its existing relationships — every technology company uses GitHub in some form — and massive cloud infrastructure. It doesn’t need to build data centers; it can subsidize inference with ease if it controls the models.

Microsoft’s divorce from OpenAI brings the potential to declutter Windows and bring plentiful inference to enterprise customers. I think that’s a good place for Microsoft to be.