anthropics
[MARKETING SCAM] This issue is for the Club of Coders who Lost their Time and Money because of Claude Code
If like me, you saw the positive marketing of Claude Code and decided to release some bucks and some weekends to give it a try, but all you got was a long and desperate loop of Claude Code gaslighting you into believing that the feature is "perfectly implemented" only to discover that he made you a pile of mashed potatoes while forgetting the original goal after several automatic compaction, despite intense prompt engineering and numerous documentation files, ... then join the Club of Coders who Lost their Time and Money because of Claude Code, and comment here what is your experience with Claude Code on projects bigger than 3 files.
Though Anthropic makes it clear that Claude Code is in beta, the whole marketing is made to imply that Claude Code is extraordinarily autonomous. Some videos suggest that one can write instructions to Claude Code, let it work for a few hours, and get back for the result.
This is a lie.
Claude Code only works for very small projects, it will impress beginners, for sure, but as soon as the project size goes from very small to small, it loses track, constantly pretending that "BINGO ! I found the issue !" in an endless loop of try and errors, until the coder faints of fatigue.
Anthropic not being clear and honest about Claude Code capabilities, pushing coders to subscribe to significant subscriptions is clearly a scam, a money grab.
No, even if it is cutting edge technology, it is not in a usable state that justifies spending such amounts of money and time. You cannot work for real, on real projects with Claude Code.
I wanted to support Anthropic even if they don't have all the gimmicks (voice, image generation etc) because of their ethical mission, but my confidence in them brutally decreased after discovering the hard way what they have done with Claude Code. I thought that the switching to Claude 4 would address those issues, but the coder experience is the same exact extreme frustration.
I think this is a serious subject : it shows how augmenting intelligence can also augment gaslighting and frustration, which is an issue to address imperatively. I'm in total fear of imagining what will happen when such problems will happen with robots.
Some of the projects I lost my money and weekends to :
- a form in angular
- a clustering in python
Please share what sort of projects you lost your money and weekends to.
