Been using Claude Sonnet 4 for a web app that's using React, Mantine, Vite, and a NoSql db.
Frankly, it's been incredible. The app is not complicated by any means. It's a really simple CRUD app for tracking animals (what farm animals are on, breeds, medical history, birthing schedule, etc.), but when I've provided mockups, it's done a surprisingly good job to at least get me 80%-90% of the way there.
I think it has to do with my tech stack honestly. Well built components can get an individual developer with poor FE experience quite far. I think it's the same for AI as well at this point.
Been using Claude Sonnet 4 for a web app that's using React, Mantine, Vite, and a NoSql db.
Frankly, it's been incredible. The app is not complicated by any means. It's a really simple CRUD app for tracking animals (what farm animals are on, breeds, medical history, birthing schedule, etc.), but when I've provided mockups, it's done a surprisingly good job to at least get me 80%-90% of the way there.
I think it has to do with my tech stack honestly. Well built components can get an individual developer with poor FE experience quite far. I think it's the same for AI as well at this point.
Yes I’ve grumpily wasted far too much time on debugging reasonably basic layout things with AI and getting nowhere.
AI is really bad at fe.
Claude is exceptionally bad at FE; it has hardly improved since Claude 3 while everything else has made strides forward.
GPT seems legit multimodal, as in images and text in the same space, and it can maneuver FE at the level of an experienced engineer.
Ai is ok at fe but to me seems best at raw html and ironically can really struggle with css and some frameworks and design systems
Im guess it has the most training samples on raw html followed by react, etc
I had much better success with FE than with BE.
Does it mean AI is really bad at FE and even worse at BE?