I suppose what’s impressive is that (with the author’s help) it did ultimately get the port to work, in spite of all the caveats described by the author that make Claude sound like a really bad programmer. The code is likely terrible, and the 3.5x speedup way low compared to what it could be, but I guess these days we’re supposed to be impressed by quantity rather than quality.
We must have a different definition of arbitrary. OP ran 2.3 million tests comparing random battles against the original implementation? Which is probably what you or I would do if we were given this task without an LLM.
Well I cloned the repo and cannot generate this
battle test by following the instructions. It appears a file called dex.js that is required is not present among other things as well as other suspicious wrong things for what appears to be on the surface a well organized project.
I'm very suspicious of such projects so take it for what you will, but I don't have time to debug some toy project so if it was presented as complete but the instructions don't work it's a red flag for the increasingly AI slop internet to me.
I suppose what’s impressive is that (with the author’s help) it did ultimately get the port to work, in spite of all the caveats described by the author that make Claude sound like a really bad programmer. The code is likely terrible, and the 3.5x speedup way low compared to what it could be, but I guess these days we’re supposed to be impressed by quantity rather than quality.
one possible argument against the productivity is if the mirgration introduced too many bugs to be useable.
In which case the code produced has zero value, resulting in a wasted month.
Its not. The project does not work or actually implement anything. It just compiles and passes some arbitrary tests the author wrote.
We must have a different definition of arbitrary. OP ran 2.3 million tests comparing random battles against the original implementation? Which is probably what you or I would do if we were given this task without an LLM.
Well I cloned the repo and cannot generate this battle test by following the instructions. It appears a file called dex.js that is required is not present among other things as well as other suspicious wrong things for what appears to be on the surface a well organized project.
I'm very suspicious of such projects so take it for what you will, but I don't have time to debug some toy project so if it was presented as complete but the instructions don't work it's a red flag for the increasingly AI slop internet to me.