OpenAI launched ChatGPT Agent on Thursday, its latest effort in the industry-wide pursuit to turn AI into a profitable enterprise—not just one that eats investors’ billions. In its announcement blog, OpenAI says its Agent “can now do work for you using its own computer,” but CEO Sam Altman warns that the rollout presents unpredictable risks.
[…]
OpenAI research lead Lisa Fulford told Wired that she used Agent to order “a lot of cupcakes,” which took the tool about an hour, because she was very specific about the cupcakes.
I’m still wondering. Like did it call up a bakery and place an order? Or go online? I know it didn’t actually make the cupcakes itself.
But I’m not sure that spending an hour trying to wrangle ChatGPT into getting your cupcakes is any faster or easier than placing the order yourself.
The article also noticeably omits what happened after. Were the cupcakes made, and did they match what she wanted?
The AI willed those cupcakes into existence, why don’t you trust them?
It’s like the metaverse and NFT, you’re not supposed to think about how it works. Instead you just need to believe reality will magically reorganize to make it work.
Also did baking the cupcakes use more or less energy than ChatGPT used to order them?
I’m guessing it’s the AI agent stuff. Which at the moment is literally just automating browsing through a website.
Apparently there will be APIs to do this in the future. Ironically, AI wouldn’t even be needed for that to be useful.