News for hackers. Stuff that matters.
Re:At Amazon, Some Coders Say Their Jobs Have Begun to Resem... by redtaperat on Sunday May 25, 2025 @11:08PM (#44093608)
This is one of several shortcomings I’ve encountered in all major LLMs. The llm has consumed multiple versions of SDKs from the same manufacturer and cannot tell them apart. Mixing up apis, methods, macros, etc. Worse is that for more esoteric code with fewer samples, or more wrong than right answers in the corpus means always getting broken code. I had this issue working on some NXP embedded code.
Re:At Amazon, Some Coders Say Their Jobs Have Begun to Resem... by dpkirchner on Monday May 26, 2025 @12:16AM (#44093926)
Human sites are also really bad about mixing content from old and new versions. SO to this day still does not have a version field you can use to filter results or more precisely target questions.
Re:At Amazon, Some Coders Say Their Jobs Have Begun to Resem... by lodovic on Monday May 26, 2025 @01:45AM (#44094331)
We have MCP servers to solve that (and Context7)
Re:At Amazon, Some Coders Say Their Jobs Have Begun to Resem... by redtaperat on Monday May 26, 2025 @08:31AM (#44096833)
I see those as carefully applied bandaids. But maybe that’s how we need to use AI for now. I mean we’re burning a lot of tokens to undo mistakes in the weights. That can’t be the right solution because it doesn’t scale. IMO.