[mgaunard]

Why is the AI only able to reach 17%?

Surely it can just keep iterating until it implements the full test suite?

[hoppp]

Money probably. This is a cash burn project.

[ekinertac]

it's not that expensive actually. been working on this about a month with my 20x Max Claude subscription while I'm working on other projects as well.

[dzhiurgis]

I suspect it cost less than $100 using chinese models.

[ekinertac]

its still iterating, 17% is just where the counter is today. three weeks ago it was at 10%, two weeks ago 13.8%. i didnt post this as a final result, i posted becuase wp-admin rendering surprised me.

but no, it cant reach 100%. around 55-60% of the suite tests C extensions, gd, curl, soap, intl, mysqli, ffi, sockets etc. passing those would mean writing all those extensions from scratch too (libcurl, ICU, an image library...) which is a completely different project. the realistic ceiling for a from scratch engine is around 40-45% and thats the number im climbing towards.

[mgaunard]

most of those are just bindings, trivial work.