Addy Osmani, a Google engineer, writes on AI-assisted coding, agentic workflows, and software engineering practice.
Commenters often cite his 'The 70% Problem' to argue that LLMs get you only part way on coding tasks, and link his newer posts.
> But that doesn't mean that I have to lay any more credence to the guys who have, since time immemorial, said "this new thing makes something easier; it will make us worse because we do not slog as much".
I don't use LLMs myself, but the anecdotes about skill atrophy seem credible. From the OP: "Even for me it shows. I tried to write some test code recently and I absolutely forgot how to write table tests because I generate all that. And it’s frightening."
An article about the topic that tries to stay positive: https://addyo.substack.com/p/avoiding-skill-atrophy-in-the-age
Yes, that's all true. Even so, vibe coding empowers anyone who can write clear instructions to build software, but the limits of the technology get hit pretty quickly by non-developers and they have little recourse. This blog post https://addyo.substack.com/p/the-70-problem-hard-truths-about is a great overview.
The tech will get better and better (I couldn't imagine we'd be doing this a year ago) but to be truly useful it has to reliably produce reasonably well engineered code, and effective debugging is a key piece of that.
Yep a good article about this: https://addyo.substack.com/p/the-70-problem-hard-truths-about
LLMs are good for the first 70% (I'd argue probably 50%).
Original posts:
https://twitter.com/addyosmani/status/1739052802314539371
https://addyo.substack.com/p/first-do-it-then-do-it-right-then
Which has been proved again and again to only get you 70% there [1].
[1] https://addyo.substack.com/p/the-70-problem-hard-truths-about
Addy Osmani also wrote today about this https://addyo.substack.com/p/conductors-to-orchestrators-the-future