Could this finally be Microsoft's moment?
Short summary
Microsoft's recent Copilot announcements represent a shift from AI as answer-box to AI as persistent worker—the real competitive advantage lies not in the model itself but in the "harness" (interface, memory, tools, permissions, execution environment) that lets businesses actually deploy and manage AI agents as ongoing coworkers. The author argues Microsoft is better positioned than labs focused purely on model capability to win this infrastructure battle.
- The competition has moved from "can AI answer questions?" to "can we delegate real work to an AI agent and have it keep working across days?" This requires persistent context, cloud execution, and integration with business systems—not just better models.
- Microsoft's September 2025 announcements bundle Copilot Home, Code, and Autopilot with model choice flexibility, meaning enterprises aren't locked into whichever AI lab is currently ahead. This flexibility is the actual moat.
- The article maps five generations of AI interfaces (answer box → custom knowledge → tool-using assistant → coworker → persistent agent), and argues we're now at the point where the system architecture around the model matters more than incremental model improvements for real work deployment.
Full summary
The article argues that Microsoft's recent AI announcements—particularly around Copilot with Home, Code, and Autopilot—represent something genuinely different from earlier hype, even though early Copilot experiences poisoned the well. The author frames this through a taxonomy of five generations of AI work, progressing from simple answer boxes (ChatGPT's original form) to persistent agents that carry context and responsibilities across conversations. The crucial shift isn't raw model capability anymore—that's improving predictably. It's whether you can actually hand meaningful work to an AI system and have it keep working on that job while you're not watching. That requires infrastructure: persistent storage, execution environments, proper memory, tools, and permissions. Microsoft's advantage isn't necessarily smarter models; it's that they're building the *harness*—the full system around the intelligence—rather than just supplying raw capability like the labs do.
The article surveys how other companies (Anthropic, OpenAI, Meta, and others) are converging on similar architecture: agents with dedicated cloud computers, shared workspaces where results accumulate, and the ability for work to continue after you close your laptop. This matters because there's a real difference between asking an assistant a question (where you stay present for the answer) and delegating an ongoing task (where it should keep working whether you're there or not). The author notes that Anthropic's recent move to merge Cowork and chat into a single Claude interface makes sense—users shouldn't have to choose which mode of the product fits their need. Local execution has limits; depending on your laptop staying awake is a weak foundation for a persistent worker.
What Microsoft has going for it is optionality on the model layer. You can use their working environment without locking into whichever AI lab happens to be ahead at any given moment—you can swap models as better ones emerge. That's a real business advantage if the harness is solid. The remaining question, which the article flags but doesn't fully answer, is whether Microsoft can actually execute the full stack: distribution, support, pricing that makes sense, and genuine control over what agents do. The test isn't impressive demos of complex tasks anymore; it's whether ordinary work—supplier reviews, routine coordination, document assembly—actually works for regular employees without becoming an administrative nightmare.