Shared context is the real bottleneck in AI-assisted content work
If you run social for a living, you already know the dirty secret of the AI era: everyone on the team got faster, and the output still didn’t get better. Your copywriter is prompting one model, your designer is prompting another, your community manager has a third tab open, and none of them can see each other’s work. The result is five private chat histories and zero shared memory of what the brand actually sounds like, what shipped last week, or which hook already flopped. That’s the problem Spaces is trying to solve, and it’s worth paying attention to — not because it’s a social media tool, but because it’s a bet on where collaborative AI work is heading, and content teams are the most obvious early test case.
Spaces is a desktop app from a team posting on Product Hunt under the maker handle John. The pitch, in their own words, is that “each project gets one shared space” where people and their AI agents work together — shared chats, shared files, and specialist agents (a researcher, a copywriter, a launch lead) that anyone in the space can put to work. Routines handle recurring work like a morning brief or a Friday recap, and ping someone only when a decision is needed. The team frames the core frustration bluntly: “five people on a team, five private AI chat histories, and zero shared context.”
My take: that framing is correct, and it’s the most honest diagnosis of knowledge-work AI I’ve read on a launch page in a while. Whether Spaces is the right execution is a separate question. Let me walk through what it actually does, how it compares to what you’re probably using now, and where I think it falls short.
What problem this actually solves (and why it’s a content-team problem first)
Most social teams don’t have a tooling problem. They have a context problem. When I’ve scheduled 30 posts across five platforms in a month, the hard part was never pushing the button in Buffer or Later. The hard part was remembering that the TikTok hook that worked in March used a specific framing, that the LinkedIn audience hates the same joke the Instagram audience loves, and that the founder’s voice guidelines live in a Google Doc nobody opens. That institutional memory is exactly what gets shredded when each person works in their own private AI chat window.
Spaces attacks this by making the space the unit of work rather than the chat. Each project gets one container with shared chats, shared files, and shared agents. A specialist agent — say, a copywriter — carries its own instructions, memory, and tools, and anyone in the space can invoke it. That’s meaningfully different from how most teams use ChatGPT or Claude today, where the “agent” is whatever prompt one person happened to paste in that morning.
Why this matters more for social than for, say, engineering
Engineering teams already have shared context baked into their tooling — repos, issue trackers, CI logs. A content team’s “repo” is a pile of Drive folders, a Notion page, a Slack channel, and someone’s memory. So the marginal value of a shared AI workspace is higher for content than for almost any other function. If a routine can generate a morning brief from the same shared space where the copywriter agent lives, you’ve collapsed three tools and one meeting into one surface. That’s the promise, anyway.
The two genuinely differentiated claims
The team flags two things they think are different, and I agree both are worth noting.
First: provider flexibility. Spaces supports ChatGPT, Claude, Gemini, “or even local models running on your own machine.” The team explicitly says Spaces “isn’t a reseller and there’s no new token subscription. You’re buying the space, not the tokens.” That’s a real architectural stance, and it’s the opposite of how most AI SaaS monetizes. If you already pay for an LLM subscription, you’re not paying twice.
Second: local-first agents and keys. The team states that “your agents, API keys, and connected accounts stay on each person’s own computer. The cloud syncs only the space — chats, files, routines — nothing else.” For agencies handling client accounts, that’s a meaningful security posture. I’d want to verify it in practice, but as a claim it’s the kind of thing that gets a security reviewer to stop scrolling.
How it stacks up against what you’re probably using
Let me be concrete about the incumbents, because “shared AI workspace” is a crowded category once you squint.
If you’re a solo creator, you’re probably using ChatGPT or Claude with a folder of saved prompts, plus Notion AI if you’re organized, plus Canva and CapCut for production. Spaces doesn’t replace any of those. It replaces the coordination layer between them — the part where you paste a brief into a chat and hope the output matches your brand voice.
If you’re on a team, you’ve likely looked at Notion, Slack with an AI add-on, or a purpose-built tool like Jasper or Copy.ai. Jasper and Copy.ai are brand-voice and template engines — they solve “generate on-brand copy at volume.” Spaces is solving something upstream: “keep the whole team’s AI work in one shared, inspectable place.” Those are different jobs, and I’d argue Spaces’s job is the one that’s been underserved.
For scheduling and analytics specifically — Buffer, Hootsuite, Metricool, Sprout Social — Spaces is not a competitor. It has no publishing, no UTM tracking, no engagement-rate dashboards. Don’t buy it expecting that. Buy it if the thinking that feeds those tools is scattered across five people’s chat windows.
Where the math breaks
Here’s the part the launch page doesn’t address, and it’s the part I’d stress-test first: shared context is only valuable if people actually put things into the space. Every shared workspace tool — Notion, Confluence, whatever — dies the same death. The team that adopts it enthusiastically for two weeks, then drifts back to private chats because it’s faster in the moment. Spaces’s routines (morning brief, Friday recap) are a smart counter to that drift, because they create a recurring reason to open the space. But routines are only as good as the inputs. If your copywriter agent’s memory is empty because nobody fed it the last three campaigns, it’s just a chatbot with a nicer UI.
Why TikTok-first creators should care more than LinkedIn-first ones
This is a judgment call, but I’d bet the value skews toward high-volume, fast-iteration platforms. A TikTok or Instagram Reels creator ships dozens of concepts a week and needs to remember which hook, format, and sound worked. That’s a context problem that compounds fast. A LinkedIn-first operator posting three times a week can hold most of that in their head. So if you’re running a high-cadence short-form operation, the shared-memory pitch should land harder for you than for a B2B thought-leadership shop.
What creators and social teams can borrow from this — even without buying it
The most useful thing about a launch like this is the operating model it implies. You can steal the model without the tool.
1. Make the project, not the chat, the unit of work. In my own tests of similar setups, the single biggest quality jump came from giving each recurring client or content pillar its own persistent context — a doc with brand voice, past winners, banned phrases, and audience notes — and pasting that in at the start of every session. Spaces productizes that habit. You can replicate it today with a shared Notion page and a saved prompt.
2. Give your “agents” instructions, memory, and tools — explicitly. The specialist framing (researcher, copywriter, launch lead) is just good role definition. If you’ve ever written a creative brief, you already know how to do this. Write the brief once, reuse it.
3. Use routines to force the recap. A Friday recap that pulls from the week’s shared work is a lightweight version of a content retro. Most teams skip retros because they’re tedious. Automating the first draft removes the excuse.
4. Keep your keys and accounts local if you can. For agencies, the local-first claim is the most operationally interesting one. If it holds up, it changes the conversation with clients who are nervous about where their credentials live.
Where I think Spaces falls short, and who it’s not for
Transparency time. The launch page is thin on several things I’d want before recommending it.
- No integrations are named. The team mentions “connected accounts” and “tools” but doesn’t list which platforms or services. If you need it wired into your Metricool analytics or your Slack workflow, that’s an open question — not disclosed.
- No team-size or scale claims. No user counts, no performance benchmarks, no case studies. Not disclosed. That’s normal for a fresh launch, but it means you’re an early adopter, not a safe bet.
- Desktop-only. The team describes it as “a desktop app.” No mention of mobile or web. For a social media manager who lives on their phone between shoots, that’s a real limitation.
- The free tier is solo-only. The team says “the desktop app is free to work solo (no account needed), and a shared space is $10.99/year per person.” That’s aggressively cheap — which is great, but also means you should ask what the business model looks like at scale. Cheap can become expensive if the product doesn’t stick.
- It’s not a publishing or analytics tool. Repeat: no scheduling, no UTM tracking, no engagement dashboards, no API rate-limit handling. It sits upstream of your Buffer or Hootsuite stack, not in it.
Who it’s not for: solo creators who already have a prompt system that works; teams that live entirely inside one provider’s ecosystem (all-in on ChatGPT, say) and don’t feel the fragmentation pain; anyone who needs mobile-first access; and anyone looking for a scheduling or analytics replacement.
What I’d watch / test next
If this resonates, here’s what I’d actually do this week — not buy, test.
First, run the free solo tier on one real project. Pick your messiest recurring client or content pillar, and see whether the shared-space model changes how you work. You’ll learn in an afternoon whether “project as unit of work” fits your brain.
Second, ask the maker directly about integrations. The launch page invites feedback — “What would you want a shared AI workspace to do?” — so ask which platforms and tools are on the roadmap. Their answer tells you more than the feature list.
Third, pressure-test the local-first claim with your security or ops person before you put client credentials anywhere near it. “API keys stay on each person’s own computer” is a strong claim; verify it.
Fourth, watch the pricing. $10.99/year per person is a land-grab number. If the product sticks, expect that to change — and plan your stack accordingly.
And finally, steal the model regardless. The real lesson from Spaces isn’t the app. It’s that your team’s AI work is leaking context every single day, and the fix is organizational before it’s technological. Fix the habit first. Then decide if you need the tool.
