“the most common software packages for fMRI analysis (SPM, FSL, AFNI) can result in false-positive rates of up to 70%”
“The death in May of Joshua Brown, 40, of Canton, Ohio, was the first known fatality in a vehicle being operated by computer systems.”
( from NYT)
It’s not that buggy software is a reality that strikes me, it’s that people think that either software is magic (for users) and therefore requires no attention, or that it will get patched soon™ enough (for “powerusers” and devs). The problem is that beta-testing, which is the new 1.0, shouldn’t be optional or amateur.