What made you start taking LLMs seriously?
热度趋势
百分比基于当前可用热度信号，而非评论数或独立用户人数。
一位开发者最初对AI辅助编码持高度怀疑态度，原因在于他们观察到许多“AI糟粕”和难以维护的代码库。然而，在短短五天内构建了一个全栈应用程序后，他们的看法发生了转变。这项工作在初创公司通常需要至少一个月的时间，主要涉及CRUD功能。尽管开发者为了架构和代码质量付出了额外努力，但他们指出，一个粗略版本可以在一天内完成。…
I used to be highly skeptical of AI-assisted coding because of all the AI slop and unmaintainable codebases I'm seeing. However, building a full-stack app in five days that would have taken at least a month at a startup I worked at changed my mind.
Nothing too fancy, mostly CRUD features. I could've shipped a rough version in a day, but I went the extra mile for architecture and code quality.
Now I see that AI is not just useful for prototypes; it really depends on the engineer. I already know what good software looks like, and put a lot of effort into learning how to work effectively with coding agents. AI sure is also capable of transforming -1x into -10x.
Same goes for writing prose. AI slop may be good enough for some, but there are people who use proper prompts, iterate over the first draft and apply personal touches. Anyways...
What was the turning point for you?