>>283564598
>>283563471
It kind of works if like a quarter of the clones are all trying something a little different instead of all doing the exact same thing, so you got 4 squads of 250clones each testing different methods until one set gets something right and makes some progress
The 1/1000th of the time seems a little egregious