> The “OpenAI model” sets up a crazy race among humans to digest and explain a messy AI proof (work that could easily be some combination of thankless, barely-credited, competitive, and unfun)
Sadly, this is also what day-to-day work looks like for a lot of software engineers in industry right now. I spend my time reading and verifying thousands of lines of messy AI-written code. Compared to actually producing something, it's thankless, barely-credited, and unfun work.
Is anybody getting good code out of these things reliably without a bunch of additional legwork? Like, what a 2020 mid-level software engineer would write? I know it's old fashioned to read code these days.
To be clear, SotA models/harnesses are really good at making things that work one-shot, and their code golf and debugging game is insane.
When I go to implement, even with a good spec, I end up with code that's 80% of the way there in a fractal manner. The modular decomposition is 80% of the way to good code. The function decomposition is 80% of the way there. The computation structure and variable naming within functions is 80% of the way there. I can walk the AI through it and address each level of issues, but it's tedious as hell, and not clearly faster than doing it myself in some cases.
This is with omp/claude/codex, with Fable 5.1/Opus 5.5/6 Astra. The Chinese models do better at staying coherent, but they're a little less smart IME. I've tried many permutations of "fable planning, opus sub-agent implementation"; "pass a branch back and forth between opus and astra, as reviewers and feedback implementers". I haven't tried the "software factory with an architect and 2 juniors" thing. I haven't tried AGENTS.md beyond the /i-have-adhd, iso-24495 (lol at the Opus 5 induced PTSD), "No cleft constructions or Latinate absolutes" and general project orientation. Model character seems to change too fast to make agents worth it, and I see stuff about skills and excessive AGENTS reducing model capabilities.
Great work :) If you're interested in Korean programming languages, there's a functional one called 'Nuri': https://github.com/suhdonghwi/nuri/
Rather than just translating keywords, it lets you write code that actually uses Korean grammar. For example, "10을 5로 나누고 출력하다" (literally "10 by 5 divide and print") outputs "2".
That is a fair feedback and I have known those languages which are very reasonable and fairly designed language. But I wanted to more focused on translated into rust for english speakers first, which would make bigger user for this language. Thanks for your feedback!
Yes, countries in the Sinosphere have historically used Chinese characters to write their languages. That's why Korean "yaksok" and Japanese "yakusoku" sound so similar. Both words are written with the same Chinese characters, "約束". The characters were borrowed from Chinese, but each language adapted them to its own pronunciation system.
For example, "library" is pronounced "tu-shu-guan" in Chinese, "do-seo-gwan" in Korean, and "to-sho-kan" in Japanese. All three can be written with the same characters, "圖書館". In modern Korea, though, people use Hangul, so very few Koreans actually know how to write "library" in Chinese characters. In Japan, Chinese characters are still heavily used, but for difficult ones, they often write kana alongside them as a reading aid.
It's very much like how Latin "universitas" became "university" in English, "universidad" in Spanish, and "università" in Italian.
There's a significant amount of Japanese loanwords in modern Korean due to Japanese annexation(1910-1945/1965), as well as in modern Chinese to much lesser extent.
These aren't an indication of a shared vocabulary or ancestry, just loanwords for concepts that were novel and scientific by victorian standards.
Sadly, this is also what day-to-day work looks like for a lot of software engineers in industry right now. I spend my time reading and verifying thousands of lines of messy AI-written code. Compared to actually producing something, it's thankless, barely-credited, and unfun work.
reply