2 comments

  • randomNumber7 17 hours ago

    This explains the quality of their products.

    But without joking, I cannot imagine this to be true, cause imo current LLM's coding abilities are very limited. It definitely makes me more productive to use it as a tool, but I use it mainly for boilerplate and short examples (where I had to read some library documentation before).

    Whenever the problem requires thinking, it horribly fails because it cannot reason. So unless this is also true for google devs, I cannot see that 25% number.

    • eternityforest 17 hours ago

      It feels like AI writes about 25% of my code. Libraries and frameworks make most of the code obvious, most of the time I'm not creating anything, I'm figuring out what the library author or language designers wants me to do.

      Over time you learn to work with the style of LLMs and architect stuff in an even more obvious predictable way, which seems to reduce bugs by a lot and make adding features easy, since the AI is trained on established codebases and knows what I might need later, rather than just making tiny spike solutions I have to rewrite.