Antonio Linares wrote:
Anthropic's Lead Engineer Boris Cherny stated in a May 2025 podcast interview that about 80% of the code for Claude Code—a command-line tool powered by Claude 3.7 Sonnet for tasks like codebase analysis and multi-file edits—was generated by the AI itself, under human direction and review.
Lo realmente curioso es que hasta ahora, o eso parece, nadie se ha atrevido a hacer ninguna modificacion/ mejora al core de Harbour con programacion IA. Tan poco interés hay en mejorar el core de Harbour ? Tan grande es el descrédito de decir que la IA ha generado el codigo ? Asi parece que, salvo la modificacion de Antonio en hbmk2, podemos afirmar sin equivocacion que por algun extraño sortilegio Harbour no se beneficiará del avance de la IAs. Seamos más puristas que nadie y volvamos al ensamblador.
BTW. También lo sugerà en el foro harbour. Ni un sólo comentario al respecto.
Here's the English translation:
What's really curious is that up until now, or so it seems, no one has dared to make any modification/improvement to Harbour's core with AI programming. Is there so little interest in improving Harbour's core? Is the discredit of saying that AI has generated the code so great? So it appears that except for Antonio's modification in hbmk2, we can affirm without error that by some strange spell Harbour will not benefit from the advancement of AIs.
BTW. I suggested it in the Harbour forum. Not a single comment about it.