3 ms·
Sure but that might be an unknowable problem. What if the difference in likelihood were 60/40. You could go down all sorts of rabbit holes and none of them wo
by throwawaymaths 2y ago
Sure but that might be an unknowable problem. What if the difference in likelihood were 60/40.
You could go down all sorts of rabbit holes and none of them would truly be falsifiable unless you observed an enantiomeric lifeform on some distant planet.