Two days with OpenAI Dots: proactive, calibrated, and a strong orchestrator
After testing OpenAI's new personal assistant Dots for two days, the author says it balances proactivity with restraint, can orchestrate Codex and ChatGPT threads, and has best-in-class computer use.
Been testing dots, @OpenAI’s new personal assistant, for a couple days and I’m bullish: - really great onboarding experience when you already have many plugins connected - amazingly proactive and calibrated to not be annoying (difficult to nail) - can orchestrate other codex and ChatGPT threads (feels like the great unifier) - slack integration is very well done - best in class computer use in their sandbox *and* your computer https://openai.com/index/introducing-dots/