From magical to frustrating in months. Why AI coding agents feel like dial-up internet and what ultra-fast inference could unlock for developer productivity.
The message seems to be “it works for me and if it doesn’t work for you it’s your fault and you’ll be left behind” from someone whose programming method is to juggle multiple instances of Claude simultaneously, getting them to write all the code. I remain skeptical and quite capable of writing my own code.
And note how these people always seem to be working on some new app, writing all new code, and not maintaining production systems. I’ve worked with enough people like that, and I don’t trust them. They impress execs with their flashy demos, and leave everyone else to pick up the pieces. Anyone can rattle off new code, but to design and build maintainable systems you need to know what you’re doing.
The message seems to be “it works for me and if it doesn’t work for you it’s your fault and you’ll be left behind” from someone whose programming method is to juggle multiple instances of Claude simultaneously, getting them to write all the code. I remain skeptical and quite capable of writing my own code.
And note how these people always seem to be working on some new app, writing all new code, and not maintaining production systems. I’ve worked with enough people like that, and I don’t trust them. They impress execs with their flashy demos, and leave everyone else to pick up the pieces. Anyone can rattle off new code, but to design and build maintainable systems you need to know what you’re doing.