Testing ChatGPT Operator, Tasks, and Deep Research

Over the last few weeks, OpenAI has done the previously unthinkable: it has consistently shipped interesting new user-facing products. First there was Tasks, a way to engage ChatGPT in helping you get things done. Then there was Operator, a way for the chatbot to actually do things for you. And finally there was “deep research,” an extremely imperfect but still very interesting tool for generating deep dives.

For now, Operator and deep research are both gated behind ChatGPT’s most expensive subscription, the $200-a-month Pro tier. (Tasks is available on the $20 Plus plan.) So on this episode of The Vergecast, we paid up and got to testing. The Verge’s Kylie Robison joins the show to talk about her experience with the shiniest things about ChatGPT – the good, the bad, the ugly, and the really, really, impossibly slow.

After that, The Verge’s Liz Lopatto joins us for an update on Elon Musk and the DOGE takeover of the US government. Liz explains where things stand now, why this is all such a big deal, and where this crisis is really headed. Musk has long assumed, often correctly, that the rules simply don’t apply to him. He’s testing that theory more aggressively than ever, and the consequences will be dire no matter where it ends.

Finally, we answer a question on the Vergecast Hotline (call 866-VERGE11, or email [email protected]!) about Super Bowl streaming. Nilay joins with some thoughts about the new scorebug, Tubi’s 4K streaming performance, and what we hope gets better before Super Bowl LX.

If you want to know more about everything we discuss in this episode, here are some links to get you started: