Apple, Notion, and a Meta Refugee Are All Coming for Office Apps

The productivity suite wars just got a third front, and it opened on the same day Apple quietly became a serious competitor in the same fight.

Notion, Apple, and a Meta Refugee All Want to Replace Your Office Apps

A former Meta AI director named Rohit Patel is leaving to build a startup called Ndoc, which is taking direct aim at Microsoft Office and Google Workspace. Patel spent eight years at Meta working on AI evaluations and reinforcement learning — essentially, the systems that measure whether AI is actually doing what it's supposed to do and then train it to do better. The hardest part of AI-assisted writing isn't generating text, it's knowing when the AI got it wrong — and that's exactly what his background is built around. Ndoc enters a brutal market, but so did Notion, and Notion won millions of users by being simply less annoying than the incumbents.

https://techcrunch.com/2026/10/27/meta-ai-director-rohit-patel-leaves-to-take-on-microsoft-office-and-google-workspace/

Apple Is Late to Office AI and That Might Not Matter

Apple is adding an AI assistant to Pages, Keynote, and Numbers — its word processor, presentation tool, and spreadsheet app — that can actually understand what's in your document before acting on it. For Keynote specifically, it can design slides, create animations, and refine existing work based on context, not just commands. Apple is using a combination of its own AI models and OpenAI's, which means it's built on something with a track record, unlike the first-party AI features that have quietly underperformed across most of the industry. The suite is free on every Mac and iPhone. If you're paying for Google Workspace primarily to use Slides, that's worth pressure-testing right now.

https://www.theverge.com/tech/1019373/apple-pages-keynote-numbers-ai-chatbot-update

Opera Wants AI to Run in Your Browser Without Sending Your Data Anywhere

Opera, the web browser with a small but loyal following, is proposing a new technical standard called WebAI that would let AI models run directly inside your browser on your own device, rather than sending your data to a company's servers to be processed. Think of the difference between a calculator app that works offline versus a web form that sends your numbers somewhere else and returns an answer. Running AI locally means faster responses, no internet required, and nothing leaving your computer. Whether other browsers adopt WebAI matters enormously — Opera alone has maybe 3% of the browser market — but the fact that someone is pushing for this standard is a signal that the "AI that doesn't surveil you" market has real demand behind it.

https://techcrunch.com/2026/10/27/opera-introduces-a-new-web-standard-to-let-users-deploy-ai-locally-in-their-browser/

The Factory Floor Is Where AI Gets Real

MIT Technology Review has a piece on what they're calling "industrial intelligence" — the combination of documented processes and the unwritten knowledge that experienced factory workers carry in their heads, like knowing how a specific material behaves before a machine restart, not just what the manual says to do. The argument is that AI's next significant commercial opportunity isn't in offices but in manufacturing, capturing and applying that tacit, hard-to-transfer expertise before veteran workers retire. For anyone in operations, supply chain, or industrial sectors, this framing is useful because it reframes the AI conversation away from text generation and toward something more durable: institutional memory that doesn't walk out the door.

https://www.technologyreview.com/2026/10/27/1147142/industrial-intelligence-the-next-growth-opportunity-for-manufacturing/

A $100M Bet That Nvidia Isn't the Only Game in Town

General Compute just raised $100 million in Series A funding — a first major outside investment round — and it's building cloud computing infrastructure for AI that deliberately avoids Nvidia chips, betting instead on inference chips from other manufacturers. "Inference" means running an AI model to get answers, as opposed to training it from scratch — it's what happens every time you use ChatGPT or Copilot. Nvidia has dominated AI hardware so thoroughly that alternatives are often dismissed, but the demand for compute is so high that investors are clearly willing to fund experiments in the alternative supply chain. If General Compute works, it puts downward pressure on the cost of running AI, which eventually matters to every business paying for AI tools by the query.

https://techcrunch.com/2026/10/27/general-compute-raises-100-million-series-a-countering-the-neocloud-disaster-narrative/

When three different battles over your document software are happening simultaneously — a startup, Apple, and Microsoft all moving at once — the winner probably isn't going to be whoever has the best AI. It's going to be whoever makes switching feel like the least amount of work.

← More daily briefs