I actually opened HN to ask something similar. Thank you for putting this out there. Sadly, people who haven't delivered anything complex genuinely believe this is the end of the programmer role. I'm 43 and went through depression about my place in the industry. It was scary.

Then I decided to build something complex using Claude, and within a week I realized that whoever claims "90% of code is written by LLMs" is not being totally honest, the parts left out from such posts tell a different story: programming is going to get harder, not easier.

The project started great but turned into a large ball of spaghetti. It became really hard to extend, every feature you want to add requires Claude to rearrange large portions of the codebase. Debugging and reading logs are also very expensive tasks. If you don't have a mental model of the codebase, you have to rely on the LLM to read logs and figure things out for you.

Overall, my impression is that we need to use this as just another tool and get proficient at it, instead of thinking it will do everything.

Also, the recent Anthropic partnership with Accenture suggests otherwise [0]. If AI could do it all, why train humans?

So please don't leave the industry. I think it will get worse before it gets better. We need to stick around longer and plan for all this hype period.

[0] https://www.anthropic.com/news/anthropic-accenture-partnersh...

One of my friends and I started building a PaaS for a niche tech stack, believing that we could use Claude for all sorts of code generation activities. We thought, if Anthropic and OpenAI are claiming that most of the code is written by LLMs in new product launches, we could start using it too.

Unsurprisingly, we were able to build a demo platform within a few days. But when we started building the actual platform, we realized that the code generated by Claude is hard to extend, and a lot of replanning and reworking needs to be done every time you try to add a major feature.

This brought our confidence level down. We still want to believe that Claude will help in generating code. But I no longer believe that Claude will be able to write complex software on its own.

Now we are treating Claude as a junior person on the team and give it well-defined, specific tasks to complete.

They are trying to fix it using this comment but cancelled mid way. Not sure why.

https://github.com/github/gh-aw/pull/14548