Is Notion Trying to Be One Bucket for Everything?
How Notion's “AI workspace” pitch collides with the rise of AI agents
Notion’s latest message “Notion is not an AI chatbot” lands right in the middle of a bigger shift: AI workspaces vs AI agents.
Notion wants to be the place where all your work lives i.e. wiki, docs, projects, and now automation, rather than just another box you chat with. The question is whether that inevitably turns it into one bucket for everything, and whether that’s a tradeoff you actually want to make?
I have to say though, Notion’s reintroduction campaign is smart positioning. Instead of fighting purely over chat UX, it’s bidding to own the workspace layer.
the system of record for knowledge, tasks and projects, with AI acting inside that context. But under all this marketing, what people hire Notion for, hasn’t really changed.
When Notion first showed up, it was much closer to a note‑taking app with benefits. Early versions in the mid‑2010s were pitched as a clean, flexible space for notes, simple tasks and lightweight wikis an alternative to Evernote or Google Docs. As users pushed it harder, the team layered on databases, views (tables, boards, calendars) and collaboration, and the narrative shifted to an “all‑in‑one workspace for your notes, tasks, wikis and databases.” Over the last couple of years, the language has evolved again: the homepage now calls it “the AI workspace that works for you,” promising one workspace where your teams and AI agents capture knowledge, find answers and automate projects.
Teams use it as a central wiki for knowledge, SOPs and meeting notes, plus flexible databases for projects, tasks, content calendars and lightweight CRMs, all sliced into boards, tables and calendars.
Now what they are promising is consolidation: replace Docs + Sheets + Trello + Confluence with one configurable surface, and let AI draft, summarize and automate on top.
On the data side, you have to separate signal from story. Notion and its ecosystem push lots of logos, case studies and numbers about tools consolidated and hours saved. But most of those stats are produced either by Notion itself or by partners like Enterpret and specialist consultancies whose business depends on Notion’s success. That’s fine as directional evidence that customers feel value, but it’s not neutral, large‑N research on what lovers actually love it for or how it compares to simpler alternatives.
The “one bucket for everything” push is real.
Official messaging talks about “the AI workspace that works for you” and “bringing all your tools and teams under one roof,” and help docs describe the workspace as the home for all your content. Community culture doubles down with “Life OS” templates and YouTube tours of people “running their entire life in Notion.” For the vendor, this blended corpus of home projects, side projects, sales pipelines, and R&Dis a goldmine of usage and training data. For users, it can quietly blur boundaries they might prefer to keep, serious work becomes just another database, and personal life becomes another project view. confusing much.
A simple way to decide whether you need a workspace OS at all is a quick self‑diagnosis:
Do you regularly have more than 15–20 active projects in flight?
Do 5+ people need a shared source of truth for those projects?
Do you juggle multiple trackers across tools and still lose track of who’s doing what?
Do you spend more time re‑explaining context than doing the work?
If you’re answering “yes” to most of that, Notion‑class tools or their competitors can genuinely earn their keep. They give large, cross‑functional teams a shared language for wiki + docs + projects, and in that world an embedded AI layer can reduce coordination tax. In that scenario, “one bucket for everything” is less a bug and more a deliberate strategy.
If not, the overhead of designing and maintaining a workspace OS is likely pure friction.
You’re usually better served by tools that are opinionated about one job and let an AI agent glue them together.
For thinking and knowledge, Obsidian or Logseq give you local‑first Markdown graphs;
Anytype or AppFlowy offer open‑source, privacy‑first workspaces.
For projects and tasks, ClickUp, Trello or Linear optimise for shipping rather than general‑purpose wikis.
For docs and data, Google Docs/Drive or Coda/Airtable cover shared writing and tables without pretending to be your entire life OS.
And In most of these setups, AI like Claude sits above the stack summarizing docs, turning notes into tasks, synthesizing status without forcing everything into a single monolith.
If you read through the replies to Notion’s X post, the reaction is mixed and revealing. Some long‑time fans lean in: “Te amo Notion,” “hard to reintroduce yourself to people who’ve never stopped using you,” and praise for the video itself. Others are clearly frustrated: one user says they are “on the verge of stopping any use of Notion because of all the bloat” and just want it to “be a good note taker,” while another calls it “a worse, overpriced Obsidian with AI features nobody asked for.” A noticeable thread of confusion runs through the comments too, with people asking “what is Notion?” in one short sentence and joking that Notion is “not an AI chatbot, but it definitely thinks it is sometimes.” The gap between Notion’s attempt to clarify its story and users still not quite knowing what box to put it in is the whole problem in miniature.
Personally, I’m in that second camp. I tried Notion, didn’t feel the need for that much structure, and moved on. I’m more interested in the strategic question than the tool itself: in an era of powerful AI agents, who truly needs an all‑in‑one AI workspace? and who just needs a handful of boring, well‑chosen tools that talk to a very smart assistant?