toothpaste_ostrich@feddit.nl to Programmer Humor@programming.devEnglish · 1 年前Uh...feddit.nlimagemessage-square50linkfedilinkarrow-up1308arrow-down117
arrow-up1291arrow-down1imageUh...feddit.nltoothpaste_ostrich@feddit.nl to Programmer Humor@programming.devEnglish · 1 年前message-square50linkfedilink
minus-squarejanAkalilinkfedilinkarrow-up155·1 年前Vibecoding: go to chatgpt ask it to make an app ask it to fix errors (ad infinitum) ??? sell app (optional) get sued and ruin your reputation (hopefully)
minus-squareGissaMittJobb@lemmy.mllinkfedilinkarrow-up10·1 年前Largely accurate apart from special-purpose tools usually being employed instead of ChatGPT
minus-squarejanAkalilinkfedilinkarrow-up2·1 年前Same thing. Coding agents just automate the third step and waste even more money and energy.
minus-squareTORFdot0@lemmy.worldlinkfedilinkEnglisharrow-up6arrow-down1·1 年前The ??? isn’t a reference to underpants gnomes. It’s just that no one has ever gotten LLMs to actually fix its own bugs to find out what that step is
minus-squaredangling_cat@piefed.blahaj.zonelinkfedilinkEnglisharrow-up1·1 年前I think it has to use one of the specialized AI coding agents like Claud Code. This way, it can burn more tokens and ruin your code base at scale.
Vibecoding:
Largely accurate apart from special-purpose tools usually being employed instead of ChatGPT
Same thing. Coding agents just automate the third step and waste even more money and energy.
The ??? isn’t a reference to underpants gnomes. It’s just that no one has ever gotten LLMs to actually fix its own bugs to find out what that step is
I think it has to use one of the specialized AI coding agents like Claud Code. This way, it can burn more tokens and ruin your code base at scale.