The End of the Chatbot Era
The vibe shift in AI is officially here. At OpenAI’s annual DevDay yesterday, the company unveiled “Dots,” a new class of proactive AI agents designed to handle tasks autonomously rather than just waiting for you to type a prompt. CEO Sam Altman says these agents are meant to stop your brain from feeling “polluted” by constant information, essentially acting as a personal assistant that lives in your phone.
Cute But Calculated
If you’ve noticed that AI is looking a lot friendlier lately, you’re not tripping. Whether it's Meta’s Muse or OpenAI’s two-eyed puffball, developers are leaning hard into “cute” aesthetics to make users feel more comfortable—or perhaps just less suspicious—about letting software access their personal lives. These agents use emoji reactions and text-like interfaces to feel more like a group chat buddy than a cold piece of corporate tech.
The “Always-On” Trade-off
OpenAI’s Dots, powered by the GPT-6 Astra model, are currently only available to Pro subscribers paying $100 a month. While Altman promises a mass-market version is in the works, for now, it’s a premium flex. The utility is definitely real—my own Dot, “Toolie,” was surprisingly good at drafting emails and planning travel—but it requires deep access to your Gmail and Google Drive.
Why it matters
Real talk: AI agents are getting way better at actually finishing tasks compared to the glitchy tools we saw a year ago. But before you hand over the keys to your digital life, remember that “always-on” also means “always watching.” While these agents are marketed as helpful sidekicks, the privacy trade-off is massive. It’s giving… a lot of power to a tiny pink frog.
