@jaredwhite I'm a bit confused how a client can demand "it [Claude Code] to be integral to all code going forward" to a developer. Are they demanding that you use Claude Code to write all of your code? But then, why should they care about the process as long as you can get the job done?
@jaredwhite @GZGavinZhao what a weird demand. Like why do they care? It makes no difference besides them believing that you used it.
I think I may ha e literally said „yes sure whatever you say, I am definitely using the stochastic parrot to write your code“ and then just not do it lol
@GZGavinZhao @jaredwhite how would they ever verify it, and what difference would it possibly make?
@nachtfunke @GZGavinZhao well I advertise my ethical qualms on my consultancy website 😅
Might need a slight revamp, but the sad reality is since I wrote this almost nothing has changed for the better. 🫠
@nachtfunke @GZGavinZhao @jaredwhite I heard a similar story from the opposite perspective recently, and AIUI, the reasoning was that LLMs make programmers (or even non-programmers) much more efficient, so resisting their usage means to choose being highly inefficient. I think similar things happen with hyped technologies generally, with those often promising productivity boosts, but apparently it happens more than usual with LLMs. Though it sounds like at least in some cases they do actually provide such boosts (which probably leads to the increased enthusiasm): I hear of people composing working projects quickly and without familiarity with the used technologies. Even if those come with technical debt and maintainability issues, it may be viewed as a desirable trade-off.
@GZGavinZhao @jaredwhite From what I've seen, "getting the job done" usually isn't the goal (anymore) for managers. It's ideological, and the goal has become "adopting AI".
@GZGavinZhao
> Are they demanding that you use Claude Code to write all of your code?
Yes, that's exactly it. 🙄