Most people use ChatGPT the same way they use a search engine: ask a question, get an answer, close the tab. Dots is built for something different — an AI that keeps working after you walk away.
What Dots Actually Is
Your ChatGPT account is probably a mess of unrelated chats. Maybe you’ve started using Projects to group some of them, which helps, but you still have to remember which project holds which context and jump between conversations constantly.
Dots sits above all of that. Instead of managing a dozen threads, you get one persistent conversation that can reach into your entire account — your chats, your projects, your scheduled tasks — and act on any of it. Think of it less like a chatbot and more like a chief of staff who already knows everything you’ve been working on.
It also comes with its own computer, running 24/7. That’s the part worth paying attention to.
The Shift: From On-Demand to Always-On
The real value isn’t the unified inbox. It’s that you can hand Dots a repeating task and it will actually execute it on a schedule, without you being present.
Here’s a concrete example. Say you’re apartment hunting in a competitive city. You could ask Dots to search current listings that match your criteria — price range, number of bedrooms, must-have amenities — and then continue checking every 30 minutes and alert you the moment something new appears. You refine the criteria mid-conversation (say, you decide you want outdoor space), and the scanner updates automatically. A scheduled task appears in your account with the full prompt already written.
That’s not a feature most people associate with ChatGPT. It’s closer to hiring a junior researcher whose only job is to watch a feed and report back.
Practical Ways to Use It Right Now
The apartment scanner is just one pattern. The same logic applies anywhere you’d otherwise check something manually and repeatedly:
- Flight and hotel prices. Set a price threshold for a trip you’re planning. Dots monitors and notifies you when fares drop instead of you refreshing Google Flights every day.
- Marketplace hunting. Looking for a used camera lens, a specific piece of furniture, or a vintage guitar? Point it at the right marketplaces and let it surface matches as they’re listed.
- Job listings. Connect it to LinkedIn or another job board, describe the role you want, and get alerts on new postings so you can respond fast — before the listing gets buried.
- Competitor or industry monitoring. If you want to track when a competitor updates their pricing page or when a publication covers a topic relevant to your work, a scheduled scan handles that without a separate tool.
In each case, the workflow is the same: describe what you want once, let Dots set up the task, and stay in the conversation to refine it as results come in.
How the Conversation Model Changes Things
One underrated aspect of Dots is that the feedback loop stays in one place. When results come back, you don’t open a new chat and re-explain your situation. You just reply. “Add a minimum square footage requirement” or “ignore anything more than 10 miles from downtown” — and the running task updates.
This matters because context is usually the bottleneck with AI tools. You spend half your time re-explaining what you need. Dots eliminates that by treating your ongoing relationship with it as the primary interface, not individual conversations.
You can also interact by voice. For a lot of people, the fastest way to start the day isn’t typing a prompt — it’s tapping a button and talking for 30 seconds. Dots is built for that.
Availability and What to Expect
Right now, Dots is rolling out on higher-tier ChatGPT plans. OpenAI has indicated it will reach more plans over time, so if you don’t see it yet, it’s coming. When you do get access, the best first move is a low-stakes test: set up a scanner for something you’re already tracking manually. That hands-on experience will show you the model faster than any explanation.
The underlying shift here is significant. AI tools are moving from reactive (you ask, it answers) to proactive (it watches, acts, and reports). Dots is one of the cleaner implementations of that idea available right now — and unlike custom setups that require separate automation tools stitched together, it lives entirely inside an interface most people already use every day.
Start with one scheduled task. See what comes back. Then add another.