Claude Code worked for 20 or 30 minutes in total, and produced a Z80 emulator that was able to pass ZEXDOC and ZEXALL, in 1200 lines of very readable and well commented C code (1800 lines with comments and blank spaces). The agent was prompted zero times during the implementation, it acted absolutely alone. It never accessed the internet, and the process it used to implement the emulator was of continuous testing, interacting with the CP/M binaries implementing the ZEXDOC and ZEXALL, writing just the CP/M syscalls needed to produce the output on the screen. Multiple times it also used the Spectrum ROM and other binaries that were available, or binaries it created from scratch to see if the emulator was working correctly. In short: the implementation was performed in a very similar way to how a human programmer would do it, and not outputting a complete implementation from scratch “uncompressing” it from the weights. Instead, different classes of instructions were implemented incrementally, and there were bugs that were fixed via integration tests, debugging sessions, dumps, printf calls, and so forth.
"A lot of things have to be scrapped completely and rebuilt from the ground up," she tells the BBC. "In fact, if political constraints did not matter, the best thing to do would be to scrap PDVSA, but that isn't going to happen.
，这一点在91视频中也有详细论述
二、Cursor：“全员猎人”的人才密度实验Cursor是另一个极端样本。这家AI编程独角兽，内部运作完全不走寻常路。
一台原本定价3000元的中端手机，若存储成本从300元涨到540元，仅这一项就吞噬了240元的毛利空间。若终端不涨价，整机毛利可能直接归零甚至亏损。面对如此剧烈的成本冲击，手机厂商不得不做出艰难抉择。
「過去人們認為，如果你告訴AI它是一位數學教授，例如，它在回答數學問題時實際上會更準確。」桑德·舒爾霍夫（Sander Schulhoff）說。他是一位企業家和研究員，也是「提示工程」理念的推廣者。但舒爾霍夫和其他人表示，當你尋找資訊或提出只有一個正確答案的問題時，角色扮演反而會降低AI模型的準確性。