Kasra Rahjerdi
@jc4p
Quick comparison of OpenAI and Google's new automated coders: https://kasra.blog/2025/05/20/day-1-comparison-of-codex-jules/ Jules is GOATed but constantly out of capacity, Codex kinda lame af but probably useful for very specific use cases. Chats + PRs included in post.
2 replies
1 recast
17 reactions
ruburi
@ruburi
lol just remembered this — imagine agents that roam around updating all your repos, keeping your code up-to-date and shiny. like every agents job is to clean up and prune your little techno garden. kinda beautiful 😢
0 reply
0 recast
1 reaction
ruburi
@ruburi
all chatgpt models became lames in coding recently. dont know why though, maybe it seems that way only to me
0 reply
0 recast
1 reaction