Ok, found an explanation here https://note.nkmk.me/en/python-numpy-dtype-astype/ Scroll down to "The number of characters in strings"

numpy creates the array with the maximum number of characters allowed in an element = the longest element. In my case this is 3. That is what dtype='<U3' indicates when you print the array. To allow additional characters in an element specify the datatype as dtype = 'U#' where # is the number of characters desired. I want to substitute "high" for "low" so dtype = 'U4' will work as indicated in the code below:

import numpy as np

X = np.array(["low"]*10, dtype = 'U4')
X

This returns: array(['low', 'low', 'high', 'low', ...])