画像クレジット: Image via source article. Used under fair use for news commentary. · source
You might be wondering why Microsoft moved Build to a smaller venue this year. The official line is something about intimacy and connection. But honestly, I think it's more telling than that.
Microsoft's annual developer conference kicks off June 2nd in San Francisco, and The Verge is reporting we'll see new AI models, a reasoning model from Microsoft AI, and something called a Copilot "super app." The headlines are predictable. AI this, AI that. But I've been thinking about what's underneath all the announcements, and I'm not sure the coverage is asking the right questions.
Here's what caught my attention: Tom Warren at The Verge has been attending Build since it was called the Professional Developers Conference. His read? He can't remember a more pivotal moment. That's not hype. That's someone who's watched Microsoft cycle through countless reinventions saying this one feels different.
The smaller venue isn't a flex. It's a retreat. Trust in Windows and GitHub is, by Warren's account, at an all-time low. That's a pretty damning assessment for a company that built its empire on developer loyalty.
I initially thought the venue change was just logistics, maybe cost cutting dressed up as strategy. But after reading Warren's framing, I'm seeing it differently. Microsoft needs to look developers in the eye and convince them the company still cares about their problems, not just about shipping AI features to hit quarterly targets.
関連記事
More in AI Models
The company promises new reasoning capabilities and a Copilot 'super app,' but the technical details remain frustratingly sparse.
Aisha Patel · 1 hour ago · 5 min
The Stargate project breaks ground in Michigan, and I've got questions about what all that compute actually gets used for.
Robert "Bob" Macintosh · 2 hours ago · 3 min
The AI company's rapid expansion of access to its vulnerability-finding model raises questions about what changed, and what we still don't know.
Aisha Patel · 3 hours ago · 5 min
The company said Mythos was too risky for public release. Now it's handing out access like conference swag.
Sources are telling The Verge we'll hear about new AI models baked into Windows, plus a new reasoning model from Microsoft AI. The details are thin, which is frustrating. We don't know if these are incremental improvements to existing Copilot capabilities or something genuinely new.
The reasoning model is interesting, though. Microsoft has been playing catch-up to OpenAI's o-series models, and a homegrown reasoning system would signal they're serious about not being entirely dependent on their partner (and competitor, sort of). But I should be honest here: I don't have enough information to know whether this is a real technical leap or just marketing language for "we made the model think a bit longer before answering."
There's also mention of "agentic OpenClaw-like tools," which, tbh, is the kind of phrase that makes me want to ask a lot of follow-up questions. OpenClaw has been getting attention in developer circles for its approach to autonomous coding agents. If Microsoft is building something similar, that's worth watching. But "like" is doing a lot of work in that sentence.
Microsoft has been struggling to articulate a coherent Copilot story for over a year now. There's Copilot in Windows, Copilot in Microsoft 365, Copilot in GitHub, Copilot in Dynamics, Copilot in whatever else they can bolt it onto. The user experience has been, let's say, fragmented.
A "super app" sounds like an admission that the current approach isn't working. Consolidation could be good! But it could also be a mess. Super apps work in markets where users want everything in one place (think WeChat in China). American users have historically been more resistant to that model. We like our apps separate and specialized.
I'm curious whether Microsoft has actual user research suggesting people want a unified Copilot experience, or whether this is an internal decision driven by engineering convenience. Those are very different motivations, and they tend to produce very different products.
The Verge notes that "major changes to Windows 11 have already started appearing," which is true. We've seen hints of a more AI-integrated shell, new developer tools, and some under-the-hood improvements to how Windows handles local AI inference.
Microsoft also just announced the Surface Laptop Ultra with Nvidia's RTX Spark, so there could be more Windows on ARM news coming. This matters more than it might seem. The ARM transition has been bumpy (honestly, that's generous), and developers need to know Microsoft is committed to making it work. Half-hearted ARM support is worse than no ARM support, because it splits the ecosystem without delivering the benefits.
But here's what I keep coming back to: none of this addresses the trust problem Warren identified. Developers don't just want new features. They want to know that Microsoft won't abandon technologies they've invested in. They want stable APIs. They want documentation that isn't six months out of date. They want GitHub to stop feeling like it's being strip-mined for AI training data.
Will Build address any of that? It remains unclear.
The developer relationship is existential for Microsoft. This isn't hyperbole. Windows matters because developers build for it. Azure matters because developers deploy to it. GitHub matters because developers trust it with their code. If that trust erodes far enough, everything else becomes harder.
I've talked to developers over the past few months who are genuinely reconsidering their Microsoft dependencies. Some are looking at Linux more seriously for desktop development. Others are diversifying their cloud deployments away from Azure. A few have moved their repositories off GitHub entirely, citing concerns about Copilot's training practices.
These aren't radical open-source ideologues. They're pragmatic people who've used Microsoft tools for years and are starting to feel like the company sees them as data sources rather than customers.
Microsoft's challenge at Build isn't just to announce impressive AI capabilities. It's to demonstrate that the company still understands what developers actually need day-to-day. Flashy keynote demos are easy. Rebuilding trust is hard.
Concrete commitments on API stability. If Microsoft announces new developer tools without addressing the churn that's burned people in the past, that's a red flag.
Clarity on data practices. GitHub Copilot's training data controversy hasn't gone away. Any new AI announcements need to address how Microsoft is handling code ownership and attribution. Vague reassurances won't cut it.
Actual developer voices. The best Build presentations have always been the ones where real developers show real work. If this year is all Microsoft executives talking about AI strategy, that tells you something about priorities.
What they don't announce. Sometimes the gaps are more revealing than the headlines. If Windows gets minimal attention while AI dominates, that's a signal about where Microsoft sees its future.
The keynote streams live at 12:30PM ET on June 2nd. I'll be watching, probably with more skepticism than I'd like. Microsoft has done impressive things with AI over the past few years, but impressive isn't the same as trustworthy. And right now, I think trust is what they need most.
