Comment by softwaredoug

These articles IMO suffer from good data. It seems we could probably at this point study open source projects longitudinally and see what we learn depending on their AI policies. Are there patterns in stability? Bugginess? Performance? Readability? Etc.