正在加载视频...
视频加载失败
I tested a new coding assistant on 34,568 lines of code. (Cursor could not achieve what this tool did.) AI models are improving every day: - Faster inference - Longer context lengths But are AI coding assistants advancing at the same rate? Can they truly understand your entire project?... show more
11 条评论

Try Augment Code now:

Coral AI is the most powerful AI for documents. See the difference yourself:

Augment Code is amazing!! Even I tested it recently and it worked flawlessly.

AI coding assistants are indeed evolving, but the challenge lies in their ability to truly grasp the nuances of human coding logic. As we push for more advanced AI, we must also ensure they can adapt contextually. It's a race, and the finish line keeps moving.

There are plenty of approaches to giving LLM context. 34k lines of code, that's very small and I'm surprised how you got that result. I have a little over 400k, no problems withcontext. documents, docstringz, cursor rules solve most problems.

@threadreaderapp unroll

@akshay_pachaar @GabrielRymberg Namaste, please find the unroll here: Enjoy :) 🤖

AI coding assistants must evolve at the pace of our ambitions. Faster inference and longer context lengths are only the beginning. True understanding requires deep learning and adaptability.

Hi Akshay, very interesting! I've tried installing this and it crashes constantly. Lots of comments on the PyCharm plugin are the same. I don't suppose you have any tips?

Yo @akshay_pachaar, sounds like you put that new coding assistant through the wringer with a whopping 34k+ lines of code! Impressive stuff. AI models are definitely leveling up faster than my grandma's knitting skills.

Looks good


