Well, to be fair, in a pilot you don't generally see a full implementation of whatever is being piloted. If it works on the most vulnerable population, maybe the program will grow.
As long as you're prepared for the political backlash when someone is unhappy with the selection.
For example, if you're randomly selecting 100 people the odds of getting someone in the 1% are ~2/3. Making UBI payments to someone in the 1% would likely kill this experiment.