AI can already write the first draft, so why is Microsoft bringing Office into the chat window?
On October 9, Microsoft announced that it was bringing Word, Excel, and PowerPoint editing and collaboration capabilities to Copilot. The rollout is starting with Chat and Cowork for Frontier customers—not all users have access to the features yet.
The change worth examining comes after the first draft. Files stay in OneDrive, with their formatting, comments, sharing permissions, and version history intact. Colleagues can make changes to the same file without repeatedly exporting and piecing together multiple versions.
My view is that competition in AI-powered office tools is extending to the “final delivery.” A model can generate a spreadsheet quickly, but teams still need to check formulas, adjust layouts, leave comments, and confirm who changed what. Preserving this continuous workflow may be what keeps users coming back to a product.
When evaluating software companies, look at whether these capabilities improve task completion rates, team usage frequency, and renewal intent. A new entry point should not be counted as new orders, and certainly not every generation should be treated as monetizable demand.
For AI-related assets like TAO, FET, and RENDER, it’s also important to assess separately how demand for models, applications, and computing power translates into each asset’s business. Microsoft’s product progress won’t automatically translate into revenue for them.
The first draft grabs attention; whether it can be delivered smoothly determines if users will come back next time.
$TAO $FET $RENDER
Tap my avatar to view my live trading account
On October 9, Microsoft announced that it was bringing Word, Excel, and PowerPoint editing and collaboration capabilities to Copilot. The rollout is starting with Chat and Cowork for Frontier customers—not all users have access to the features yet.
The change worth examining comes after the first draft. Files stay in OneDrive, with their formatting, comments, sharing permissions, and version history intact. Colleagues can make changes to the same file without repeatedly exporting and piecing together multiple versions.
My view is that competition in AI-powered office tools is extending to the “final delivery.” A model can generate a spreadsheet quickly, but teams still need to check formulas, adjust layouts, leave comments, and confirm who changed what. Preserving this continuous workflow may be what keeps users coming back to a product.
When evaluating software companies, look at whether these capabilities improve task completion rates, team usage frequency, and renewal intent. A new entry point should not be counted as new orders, and certainly not every generation should be treated as monetizable demand.
For AI-related assets like TAO, FET, and RENDER, it’s also important to assess separately how demand for models, applications, and computing power translates into each asset’s business. Microsoft’s product progress won’t automatically translate into revenue for them.
The first draft grabs attention; whether it can be delivered smoothly determines if users will come back next time.
$TAO $FET $RENDER
Tap my avatar to view my live trading account