ChatGPT dots look like magic in the demos. Then you read the fine print. 3 thing
By Joanna Lambadjieva · October 13, 2026 · Curated by George's Blog
ChatGPT dots look like magic in the demos.
Then you read the fine print.
3 things nobody is telling you.
Quick context:
A dot is OpenAI's new always-on assistant.
You hand it a responsibility, not a prompt.
It keeps working after you close the tab.
1. It doesn't cost $500.
Pro 100 gets you the exact same dot.
Higher tiers add more usage.
Only Pro 500 adds the faster speed mode.
2. Most of Europe can't get one on a personal plan.
EU, UK, Switzerland: no dot on Pro. At any tier.
The way in: Business Premium seats. (or VPN)
3. "Always-on" isn't unlimited.
Talking to your dot doesn't count toward your limits.
Every task it runs in Work or Codex does.
The breakthrough?
The brief, not the tool.
A prompt asks for one output.
A dot brief assigns a job:
→ Sources of truth
→ Cadence
→ Boundaries
→ A review owner
→ An end date
Skip that and you get confident output nobody checked.
I put everything in one guide:
🔹 Who can get a dot and what it costs
🔹 6 ecommerce workflows worth delegating
🔹 Copy-paste brief template + 3 examples
🔹 A safe 2-week pilot scorecard
I am giving it away for free but please let me know if you want it and what are your ideas for using a Dot? What work would you delegate?