2026-08-13T00:00:00.000Z Microsoft https://microsoft.github.io/nfw-reader/ https://microsoft.github.io/nfw-reader/posts/rhythms-of-work 2026-08-13T00:00:00.000Z 2026-08-13T00:00:00.000Z

AI agents can read a task and still miss the larger patterns of a person’s work. Researchers organized 667 million workplace events at four temporal scales and found that the resulting “grammar of work” structure carried predictive signal beyond recent actions.

https://microsoft.github.io/nfw-reader/posts/seven-predictions 2026-08-06T00:00:00.000Z 2026-08-06T00:00:00.000Z

Jaime Teevan makes seven predictions for a world where knowledge is abundant, but attention is scarce: reading becomes asking, audiences start to participate, meetings turn into places you return to, and organizations finally become legible to themselves.

https://microsoft.github.io/nfw-reader/posts/beyond-the-org-chart 2026-07-29T00:00:00.000Z 2026-07-29T00:00:00.000Z

Interviews with 24 professionals show how AI is changing the communication, mentoring, feedback, and relationship-building work that formal job descriptions rarely capture.

https://microsoft.github.io/nfw-reader/posts/context-ai-doesnt-see 2026-07-22T00:00:00.000Z 2026-07-22T00:00:00.000Z

A prompt can sound complete while leaving out situational context that shapes what an email needs to do. A study of 92 Microsoft professionals traces what participants knew, what reached the AI, and how they responded when drafts missed the mark.

https://microsoft.github.io/nfw-reader/posts/the-agentic-economy 2026-07-06T00:00:00.000Z 2026-07-06T00:00:00.000Z

A Microsoft Research team argues the biggest economic effect of AI agents may not be faster workers, but rather the collapse of the cost of buyers and sellers reaching and transacting with each other, and the redrawing of markets that follows.

https://microsoft.github.io/nfw-reader/posts/beyond-the-ask 2026-06-22T00:00:00.000Z 2026-06-22T00:00:00.000Z

When knowledge workers use Copilot, the AI performs an average of 3.26 work activities per conversation — nearly double the 1.68 they explicitly ask for. A reading on the gap between what is asked and what is delivered, and why the conversation, not the task, is becoming the unit of accountability.

https://microsoft.github.io/nfw-reader/posts/causal-impact-copilot 2026-06-18T00:00:00.000Z 2026-06-18T00:00:00.000Z

Using data from more than 72,000 Word users, researchers compared sustained Copilot adopters with similar coworkers who adopted later, and found the pace of work in Word changed markedly after adoption. The study introduces a new way to measure AI impact as products and people evolve together.

https://microsoft.github.io/nfw-reader/posts/ai-isnt-human 2026-05-15T00:00:00.000Z 2026-05-15T00:00:00.000Z

Early lessons from our first close encounter with a very different intelligence. Five ways AI is different from humans, and why those differences matter.

https://microsoft.github.io/nfw-reader/posts/five-million-conversations 2026-05-14T00:00:00.000Z 2026-05-14T00:00:00.000Z

Findings from an analysis of five million M365 Copilot conversations — what people actually do with AI at work, and what the patterns mean for organizations deploying it.

https://microsoft.github.io/nfw-reader/posts/shifting-work-patterns 2026-05-14T00:00:00.000Z 2026-05-14T00:00:00.000Z

A six-month randomized trial across 66 large firms and 7,137 workers watched what happens when people actually get an integrated AI assistant for their email, meetings, and documents. This is the story of which patterns moved, which held still, and why the answer matters for how organizations should think about deploying AI.

https://microsoft.github.io/nfw-reader/posts/ai-mindset-experiment 2026-04-16T00:00:00.000Z 2026-04-16T00:00:00.000Z

A field experiment with retail employees at Gap Inc. testing how mindset interventions and behavioral scaffolding shape AI adoption and task performance.